CASE Network Reports 39 - The Episodes of Currency Crisis in Latin American and Asian Economies
1 like416 views
This document, edited by Marek Dąbrowski, analyzes the causes and developments of currency crises in Latin American and Asian economies, particularly focusing on episodes in Mexico, Argentina, Thailand, and Malaysia. It presents various factors contributing to these crises, including macroeconomic, microeconomic, and political elements, and discusses crisis management strategies and post-crisis outcomes. The publication aims to draw conclusions relevant for Poland and other transitioning countries based on the findings from these case studies.
![9
The Episodes of Currency Crisis in Latin...
Part I.
The Mexican Peso Crisis 1994–1995
by Wojciech Paczyñski
1.1. History of the Crisis
In order to put into context the developments that final-ly
led to the currency crisis of 1994/1995, it is useful to go
back as far as mid-1980's. In December 1987, a set of
reform policies was initiated aimed at "remaking the Mexi-can
economy" [Lustig, 1992]. The shift in policies was acce-lerated
after 1988 as they gained support from the newly
elected president Carlos Salinas de Gortari [DeLong et al.,
1996]. The reform package was successful and brought
macroeconomic stabilisation. Inflation was reduced from
nearly 160% in 1987 to the range of 18% – 30% in
1989–1991 and further down to less than 12% in 1992 and
8.3% in 1993. At the same time economic growth resumed
reaching 3.5–4.5% pa. in the period 1989–1992. This result
was quite remarkable given the record of failed reform
attempts in previous years [Blejer and del Castillo, 1996].
The 1989 foreign debt restructuring left Mexico with
relatively low and mostly long-term foreign debt (it
accounted for some 19% of GDP at the end of 1993) [Sachs
et al., 1995]. Public debt was substantially reduced from
67% of GDP in 1989 to 30% in 1993. From December
1990 onwards, foreigners were allowed to purchase short-term
government peso-denominated debt instruments [Gil-
Diaz, 1998]. After the restructuring, Mexico once again
gained access to international financial markets. Private ca-pital
inflows surged to an average of above 6% of GDP in
the period 1990–1993 [IMF, 1995].
Economic policies during the period 1990–1993 result-ed
in the implementation of important structural reforms in
various fields. The authorities undertook major domestic
financial sector reform and capital account liberalisation
[Otker and Pazarbasioglu, 1995] and privatisation. One
should also note the improvement of the regulatory frame-work
governing economic activity in many sectors, e.g. in
tourism, means of transport, petrochemicals, electricity,
telecommunications, etc. [WTO, 1997]. Another important
factor was trade liberalisation. This process had started
CASE Reports No. 39
much earlier. In 1985 Mexico formally joined the General
Agreement on Tariffs and Trade (GATT). The next major
step was the signing of the North American Free Trade
Agreement (NAFTA) in 1992 that stipulated the reduction
in non-tariff barriers, liberalisation of investment laws,
changes in competition law, etc. The NAFTA finally took
effect in January 1994.
The government followed a path of budgetary discipline.
The operational budget of the public sector [1] was in sur-plus
in the range of 2–3% of GDP in the early 1990's. Sev-eral
social pacts were concluded between the government
and labour organisations as well as business representatives
[Blejer and del Castillo, 1996]. Among the issues agreed
upon was the exchange rate policy that became the central
anti-inflationary instrument. The question whether the
exchange rate policy was appropriate and whether it result-ed
in an overvaluation of the peso is one of the major issues
raised in all analyses of the currency crisis of 1994/1995.
This problem will be discussed later.
In 1993, the overall economic situation deteriorated
slightly. GDP growth slowed to only 0.6% and private
consumption and investment fell in real terms. These
developments are mostly attributed to the ongoing
restructuring in the manufacturing sector, a tightening of
credit conditions by monetary authorities, and a credit
squeeze resulting from the deterioration in the quality of
banks' loan portfolio [IMF, 1995]. There was also some
uncertainty about the approval of NAFTA, which was final-ly
resolved in November.
Despite these setbacks, until 1994 Mexico was widely
regarded as an example of a successful economic reform
story. Some other views [Dornbusch and Werner, 1994]
appeared among economists, but were not picked up nor
were they considered important by investors. Suddenly, in
the course of 1994, several events took place that turned
out to be of considerable importance for Mexico's econo-mic
situation.
January witnessed the peasant rebellion in Chiapas – the
first one of political events of 1994 that later turned out to
have a significant impact on financial stability of the country.
[1] Operational balance is defined as primary balance plus the real portion of the interest paid on public debt.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-9-320.jpg)
![10
Marek D¹browski (ed.)
CASE Reports No. 39
In February, U.S. interest rates started to rise. In March the
candidate in presidential elections Luis D. Colosio was
assassinated. This came as a shock to investors and led to
severe financial turbulence. The peso exchange rate
increased from the bottom of the intervention band [2],
where it stayed before, to the ceiling of the band, which
constituted a nominal devaluation of ca. 10%. This was
accompanied by a decrease in Central Bank reserves of
around 9 billion USD. The monetary authorities followed a
path of a rather loose monetary policy, boosting credit to
the economy in order to prevent interest rate increase and
to support weak commercial banks. Also, in the run-up to
the presidential elections (the output slowdown could had
been another factor) fiscal policy became more expansion-ary.
The actions involved some tax cuts and increases in
social spending [IMF, 1995].
In August presidential elections took place that gave a
victory to Ernesto Zedilo. His victory, with a higher than
expected margin, was considered a positive event from the
point of view of foreign investors even though the elections
were not carried out in a perfect way. In September, the
secretary general of the ruling party was assassinated.
Higher domestic interest rates (around 16% pa. from
April until July as opposed to around 10% pa. in the first
quarter) and the approval of a 6.75 billion USD short term
credit line from NAFTA partners helped to ease the pres-sures
from financial markets. The peso exchange rate stayed
near the ceiling of the band, outflow of capital was stopped
and reserves remained relatively stable from April until
October. After July, interest rates began even to decline.
Another policy action implemented by the authorities in
order to increase the credibility of maintaining the exchange
rate rule and to prevent increases in interest rates was sub-stituting
short term peso-denominated government debt
(Cetes) with dollar-indexed (but payable in pesos) short
term bonds (Tesobonos). This started after the March
events and continued in the following months. The out-standing
stock of Tesobonos increased significantly – from
14 billion pesos in March 1994 to 63.6 billion in November.
The whole operation within a very short period dramatical-ly
changed the composition of short-term debt held by the
private sector. While in the first quarter the share of
Tesobonos in total Cetes and Tesobonos stock did not
exceeded 10%, it reached almost 60% in July.
The current account continued to deteriorate in the
third quarter of 1994 reaching a record level deficit of 7.9
billion USD. In addition, both the stock of Tesobonos and its
share in total short-term debt increased further. Heightened
concerns about the sustainability of Mexico's external posi-tion
led to intensified capital outflows. The reserves
declined by 4.7 billion USD between October and Novem-ber
and further 2.5 billion USD to 10 billion USD in mid-
December. On 1 December president Zedillo took office
and two days later the unrest in Chiapas intensified. Given
the current situation, the authorities decided on 20 Decem-ber
to widen the exchange rate band by 15%. This move
was accompanied by the announcement of the authorities
to support the peso at a rate of around 4 pesos to the U.S.
dollar. This announcement was, however, not perceived as
credible by investors, who put further pressure on the
exchange rate. The Bank of Mexico lost around 4 billion
USD within two days and was forced to freely float the peso
on 22 December.
Inflation was certainly one of the most important prob-lems
that the authorities had to tackle after the devaluation.
It jumped to the level of around 8% monthly, but in the sec-ond
half of the year was reduced to the range 2–4% month-ly.
The peak of 12-month inflation was recorded in Decem-ber
1995, when it stood at 51.97%. During the first quarter
of 1995, the peso depreciated at a rather high rate reaching
6.82 in the end of March. It then regained some strength
fluctuating between 5.8 and 6.4 pesos to the dollar until
September, to fell further in the last quarter to 7.64 in the
end of December.
The crisis also resulted in a severe recession with GDP
falling by 9.2% YoY in the second quarter and respectively
by 8.0% and 7.0% in the third and fourth quarter of 1995
[INEGI, 2000]. Industrial production dropped sharply and
the unemployment rate increased. In the second part of the
year, the first signs of economic recovery became visible.
These trends intensified in the last quarter, and since the
second quarter of 1996 the Mexican economy returned to a
path of fast growth (YoY rate of GDP growth reached 7.2%
in the second quarter of 1996). The severity of the 1995
recession was caused by several factors, the most important
probably being very high interest rates (lending rate stayed
close to 70% in the first half of the year) that were used as
an anti-inflationary measure. Other factors included a signi-ficant
drop in capital inflows, and sudden reduction in cred-it
in the economy. Domestic consumption was further
repressed due to debt overhang and possibly substantial
negative income and wealth effects resulting from the deva-luation
[SHCP, 1995]. Gruben et al. (1997) point at sectoral
fragmentation of severity of recession and the timing and
strength of a rebound.
During 1995, significant adjustment took place in exter-nal
position of Mexico. Exports surged by 30% in compari-son
to 1994 and imports contracted by around 9%. As a
result, the trade balance improved from a deficit of 18.5 bil-lion
USD in 1994 to a surplus of 7 billion in 1995. The cur-rent
account deficit contracted from nearly 30 billion USD
in 1994 to only 1.5 billion USD. A very important achieve-
[2] Since November 1991 Mexico operated a moving band exchange rate system. This is discussed in more detail in section 1.2.4.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-10-320.jpg)
![11
The Episodes of Currency Crisis in Latin...
ment of the authorities was the elimination of the short-term
debt overhang and consequently regaining access to
international capital markets. In particular Tesobonos were
practically eliminated from the short-term debt stock during
1995. An access to credits from the foreign financial support
package played an important role in managing the debt
problem. Mexico used close to 12 billion USD of IMF cre-dits
in 1995 in addition to around 14 billion of other excep-tional
financing. A much-improved economic condition
allowed Mexico to pay back these credits, in some instances
ahead of schedule. Since 1996 Mexico has experienced re-latively
stable economic growth.
1.2. In Search of the Causes of the Crisis:
Macroeconomic Factors
1.2.1. Fiscal Policy
The role of fiscal policy in the peso crisis has not been
emphasised in most of the studies. This is because in the
early 1990's Mexico achieved remarkable successes in near-ly
balancing the public finances. The general public sector
deficit declined from around 16% of GDP in 1986 to about
2% in 1993. In 1994 the result was not much worse – the
deficit reached around 3.9% of GDP [3]. This fiscal perfor-mance
was to a large extent due to reduced interest pay-ments
during the period. One important observation is that
fiscal policy was not tightened, and thus was not used as a
tool for dealing with negative shocks of 1994. On the con-trary,
fiscal policy was rather looser in the election year.
The role of quasi-fiscal operations via development bank
credits in 1994 is not very clear. Very soon after the crisis,
some authors presented the view that fiscal expansion
CASE Reports No. 39
through this channel could had played some role in the mix
of bad policies that were implemented in 1994 [World Bank,
1995]. Most of the analyses show however, that develop-ment
banks' credit was not an important factor. Sachs et al.
(1995) argue that most of the activities of these banks do
not belong in an economically meaningful definition of a
budget deficit.
An innovative way of looking at the role of fiscal policy
in explaining the crisis is proposed by Kalter and Ribas
(1999). They point out the role of the increasing magni-tude
of go-vernment operations, rather than the size of
government deficit, in affecting the relative price of traded
to non-traded goods (i.e. the real exchange rate), the
financial condition of the traded sector, and interest rates.
They argue that the significant rise in government non-oil
revenue collections measured in U.S. dollars or in terms of
traded goods prices has had an effect on the tradable sec-tor
analogous to that of a surge in export commodity
prices (Dutch disease). The resulting deterioration of
finances of traded goods sector was then passed to com-mercial
banks' finances. While the arguments used by
Kalter and Ribas (1999) are interesting and certainly add
another dimension to the understanding of fundamental
reasons behind the crisis, they do not provide an explana-tion
for sudden events of December 1994.
1.2.2. Savings and Investment Balance
In the period 1988–1994 Mexico witnessed a notice-able
growth in investment and a decline in savings (see
Table 1-2). Overall investment grew from 20.4% of GDP
in 1988 to 23.6% in 1994. Interestingly, public sector
investments remained relatively stable and were even
reduced, while the growth was due to the private invest-
Table 1-1. Public sector balances 1986–1994 (in percent of GDP)
Financial balance Primary balance Operational balance
1986 -16.1 3.7 -2.4
1990 -3.3 7.6 1.8
1991 -1.5 5.3 2.9
1992 0.5 6.6 2.9
1993 -2.1 3.6 2.1
1994 -3.9 2.3 0.5
Notes: Financial Balance includes all public sector borrowing requirements.
Primary Balance is defined as Financial Balance less interest paid on public debt.
Operational Balance is defined as Primary Balance plus the real portion of the interest paid on public debt.
Other sources provide slightly different data.
Source: Sachs et al. (1995).
[3] There is no consensus about the size of public sector surplus or deficit. Different authors use distinct measures. For example, Kalter and Ribas
(1999) estimate the overall public sector deficit close to 1% in 1992–1993 and close to 2.5% in 1994.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-11-320.jpg)
![12
Marek D¹browski (ed.)
Current
account
1988 1.4 17.6 19 5 1 .4 520.4 -3.6 2.2 -1.4
1989 3.1 15.6 18.7 4.8 16.5 21.3 -1.7 -0.9 -2.6
1990 6.7 12.5 19.2 4.9 17 21.9 1.8 -4.5 -2.7
1991 7.5 10.3 17.8 4.6 17.8 22.4 2.9 -7.5 -4.6
1992 7.1 9.5 16.6 4.2 19.1 23.3 2.9 -9.6 -6.7
1993 6.3 8.9 15.2 4.2 17.8 22 2.1-8. 9 -6.8
1994 5 10.7 15.7 4.5 19.1 23.6 0.5 -8.4 -7.9
Source: Sachs et al. (1995)
CASE Reports No. 39
Table 1-2. Saving and investment levels 1988–1994 (in percent of GDP)
Saving Investment Net saving
Public Private Total Public Private Total Public Private
ment boom. This was accompanied by an even more
apparent reduction in propensity to save. Total savings fell
from 19% of GDP in 1988 to only 15.2% of GDP in 1993
and 15.7% in 1994. Private savings declined from 17.6%
of GDP in 1988 to 8.9% in 1993 before starting to grow,
albeit modestly, in 1994. This leads to the conclusion that
the deterioration in the current account – the deficit
reached 6.8% of GDP in 1993 and 7.9% in 1994 – was
primarily caused by the level of private savings not match-ing
the level of private investment [4].
1.2.3. Private and Public Debt
The role of Mexico's indebtedness in provoking the
financial crisis deserves a more detailed analysis. First, it
should be noted that Mexico had a history of problems asso-ciated
with its foreign debt, including the crisis of 1982. In
contrast to the past, the beginning of the 1990's was marked
by a very significant improvement in this sphere. Public debt
was reduced from some 64% of GDP in 1989 to 35% of
GDP in 1993. Of this, 23% of GDP was foreign debt that as
a result of 1989 restructuring had a favourable maturity
structure (long term liabilities prevailed). Domestic debt
accounted for 12% of GDP [IMF, 1995] [5]. These numbers
were low in comparison to other developing and developed
countries. Moreover, standard debt indicators such as the
ratio of total debt to GDP or to exports or the ratio of inte-rests
on debt to GDP or exports were improving in the
early 90's. This clearly shows that the overall level of public
debt did not play a big role in the loss of investors' confi-dence
in 1994.
What did matter, however, was the maturity and cur-rency
structure of the domestic debt and the level of pri-vate
borrowing. From 1989 to 1992, net credit to the
private sector from the financial system expanded at an
average annual rate of 66% in nominal terms, offsetting
the decline in borrowing of the public sector resulting
from the substantially improved fiscal position [IMF,
1995]. In 1993 net domestic credit of the banking system
continued to expand at an annual rate of around 20%.
Interestingly, this decomposes into a substantial reduc-tion
in credit to the public sector (around 30%) and an
expansion of credit to the private sector of the same
magnitude (see table 1-3). This trend continued through
the first half of 1994, while in the second half public sec-tor
borrowing also started to rise, bringing the 12-month
rate of growth of net domestic credit from the banking
system to around 30%. It is also worth noting the faster
expansion of credit to private sector from development
banks than from commercial banks in 1994. Other inter-esting
statistics are presen-ted in Gil-Diaz (1998) [6]
which indicate that in the period from December 1988
to November 1994 credit card liabilities rose at an aver-age
rate of 31% per year, direct credit for consumer
durables rose at a yearly rate of 67% and mortgage loans
at an annual rate of 47%, all in real terms.
The above numbers, along with numbers cited in sec-tion
1.2.2 with regard to private investment and savings,
show that it was mostly private sector borrowing that
brought the current account deficit to the levels it reached
in 1993 and 1994 (more than 6% of GDP). Such a level of
the deficit seems quite high but was nevertheless easily
financed in 1993, and from that perspective there were
[4] Sources differ in calculations of the current account deficit in relation to GDP. For example Gurrha (2000) citing official Mexican data estimates
the deficit to account for 5.8% of GDP in 1993 and 7.0% in 1994.
[5] Also in this case different numbers are cited by other authors. For example Sachs et al. (1995) estimate the public debt at 67% of GDP in 1989
and 30% in 1993. According to this source this last number can be broken down to 19% of GDP of foreign debt and 11% of domestic debt. The ave-rage
maturity of domestic debt was around 200 days.
[6] The statistics were provided to the author (a former Vice Chancellor of the Bank of Mexico) by the Economic Research Department of Bank
of Mexico.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-12-320.jpg)
![13
The Episodes of Currency Crisis in Latin...
Table 1-3. Monetary sector – the expansion of credit 1990–1994 (twelve-month rates of growth, end of period)
1990 1991 1992 1993 1994
Broad money (M4) 46.4 30.9 19.9 25.0 17.1
Net domestic assets of the financial
system
22.6 31.6 20.8 15.5 32.0
Net credit to public sector 3.9 -1.6 -31.7 -46.2 25.3
Net credit to private sector 63.2 53.3 57.126. 4 31.9
Net domestic credit of commercial
49.4 48.6 20.9 24.3 …
banks
Net domestic credit of development
banks
-10.0 18.8 23.4 47.4 42.2
Source: IMF (1995).
reasons to believe that this could had happen again in
1994. Firstly, one should note that large capital inflows that
resulted from markedly improved perception of the econ-omy
by foreign investors and that were possible thanks to
capital account liberalisation of 1990 were transferred to a
significant increase of short term debt. The inflow was
sterilised by issuing short-term government debt instru-ments
(Cetes). As a result short-term indebtedness
increased significantly with short term debt to total debt
stock ratio rising from 21.9 in 1992 to 28.1 in 1994 [World
Bank, 2000]. At the end of 1993 the value of Cetes alone
reached 22.9 billion USD, i.e. it was very close to net
international reserves of the Bank of Mexico (24.9 billion
USD) (see Figure 1-1). This placed Mexico in a potentially
vulnerable position. On top of that, during 1994 the cur-rency
structure of short-term debt underwent a substan-tial
change.
After the March assassination and resulting turbulence
in the financial markets, the authorities started exchanging
peso-denominated bonds (Cetes) with dollar-indexed debt
instruments (Tesobonos). This action was aimed at uphold-ing
the investors confidence in the exchange rate regime
after the peso depreciated by around 10% reaching the
ceiling of the intervention band. It was also used as a tool
to avoid further increases in interest rates. Werner (1996)
estimates that the substitution from Cetes to Tesobonos
was equivalent to an interest rates increase of around 8 to
11 percentage points. He argues that accounting for the
currency composition of government debt gives more
appropriate measures of currency risk premium in the
period before the crisis. The scale of substitution from
Cetes to Tesobonos was huge. By June 1994, the amount
of Tesobonos and Cetes were roughly equal and in Decem-ber
the amount of Tesobonos was around 5.5 times that of
Figure 1-1. The composition of short term government debt (USD million)
30000
25000
20000
15000
10000
5000
0
1993M12
1994M6
1994M5
CASE Reports No. 39
1994M8
1994M7
1994M11
1994M9
1994M10
1994M12
1995M2
1995M1
1995M4
1995M3
Cetes
Tesobonos
BoM reserves
Notes: Cetes – three month peso-denominated government debt.
Tesobonos – three month dollar-indexed government debt.
Source: own calculations based on IMF, IFS and World Bank (1995) data.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-13-320.jpg)
![14
Marek D¹browski (ed.)
1,8
1,6
1,4
1,2
1
0,8
0,6
0,4
0,2
Cetes. Another way to look at the process is to note that
around 15 billion USD of private sector holdings in Cetes
were swapped for Tesobonos from March till November.
After the devaluation on 20 December, government
borrowing was clearly not sustainable. Investors rushed to
withdraw their investments and the government found
itself unable to cover short term liabilities that led to a
panic and the severe currency devaluation. Eventually, only
a huge international support package helped to solve the
problem. The shift to dollar denominated short-term pub-lic
debt certainly contributed to the whole set of factors
that provoked the crisis. Interestingly, however, risks asso-ciated
with rapidly growing short-term dollar indexed debt
of Mexico seem to had been underestimated, not to say
unnoticed, by the international financial community until
the devaluation took place. Sachs et al. (1996) present
puzzling statistics on international press coverage of Mexi-co.
The issue of Tesobonos was completely ignored by
leading international financial papers with only one article
mentioning it being published before December 1994 [7].
The problem of accumulated dollar denominated debt
accompanied by depleted foreign reserves constituted an
important factor in provoking the panic after the
announcement of devaluation on 20 December (Sachs et
al., 1995, see also section 1.2.5).
1.2.4. Exchange Rate Policy
In the last decades, Mexico altered its exchange rate po-licy
several times and had a history of several episodes of sig-nificant
devaluations (e.g. 1976, 1982, and 1985). A fixed
exchange rate regime that was introduced in 1988 and later
corrected on several occasions played a major role in the
anti-inflationary strategy of Mexican authorities. From Janu-ary
1989 until November 1991, a preannounced crawling
peg was in operation (with two reductions of the rate of
crawl), and from 11 November 1991 an exchange rate
intervention band was used with several changes in the rate
of crawl of both upper and lower bands. From 1991 until
November 1994, the peso steadily and very slowly depre-ciated
in nominal terms. Yet since the level of inflation was
much higher than in the U.S., in real terms the peso appre-ciated
by around 15% if consumer price indexes for Mexi-co
and the U.S. are applied or close to 21% if the compari-son
is based on wholesale price indexes [8].
The question whether the peso was overvalued at that
time brought much attention, especially after the December
crisis. One should note, however, that there were voices
pointing to an overvaluation of the peso and the possible
risks that it posed already in early 1994, the best known
being Dornbusch and Werner (1994). In most analyses of the
[7] The authors surveyed the Financial Times, the New York Times, and the Wall Street Journal. The number of such articles jumped to 6 in Decem-ber
1994 and 46 in January 1995.
[8] In a new study Dabos and Juan-Ramon (2000) estimate the model of the real exchange rate in Mexico. Their results suggest that on the eve of
the crisis the peso was overvalued by a number in the range of 12 to 25 percent. This result is consistent with the majority of previous studies.
CASE Reports No. 39
Figure 1-2. Real exchange rate movements 1975–1995
0
RER_cons RER_prod
1975M1
1976M1
1977M1
1978M1
1979M1
1980M1
1981M1
1982M1
1983M1
1984M1
1985M1
1986M1
1987M1
1988M1
1989M1
1990M1
1991M1
1992M1
1993M1
1994M1
1995M1
Note: Rer_cons index is obtained using consumer price indexes in Mexico and the U.S., while Rer_prod is based on producer price indexes; 1 is
an average value of respective indexes over the period.
Source: Author's calculation based on IMF, IFS data](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-14-320.jpg)
![15
The Episodes of Currency Crisis in Latin...
crisis that appeared after 1994, the view that the peso has
indeed been overvalued seems to gain rather strong support
[World Bank, 1995]. The standard reasoning points to the
fact that the exchange rate-based stabilisation under capital
mobility has led to a large current account deficit and real
appreciation of the peso that at some point became unsus-tainable
and the correction of real exchange rate was need-ed
[IMF, 1995]. It is, however, not clear whether this has
played an important role in determining the crisis. In partic-ular,
some authors concluded that the peso overvaluation is
not at all useful in explaining the crisis [Gil-Diaz, 1998]. Also,
as Sachs et al. (1996) point out, significant reduction in infla-tion
in 1994 and 10% nominal depreciation from March to
April 1994 certainly diminished the overvaluation problem.
On 20 December, the upper limit of the intervention
corridor was widened by 15%, but at that time the devalu-ation
of that scale was widely regarded as insufficient. On
the other hand, it undermined the confidence in the will and
ability of the Mexican authorities to uphold the announced
exchange rate policy. The continued pressure and a lack of
possibilities to support the peso forced the authorities to let
the peso flow freely on 22 December. In later months, the
peso depreciated sharply hitting the rate of 6.82 pesos to
the dollar in the end of March 1995, i.e. nearly twice as
much as before 20 December. It recovered slightly in the
next few months.
1.2.5. Monetary Policy
During the whole of 1993, disinflation continued and
interest rates on short term government papers were on a
downward trend. Interest rates declined from around 17%
at the beginning of 1993 to 13–14% in November. The
approval of NAFTA in that month allowed for a further sig-nificant
decrease below 10% in February and March 1994.
Political turbulence at the end of March resulted in a surge
in interest rates that stayed in the 16%–17% range from
April until July. From August onwards, interest rates started
to fall again and remained stable at around 13.7% from Sep-tember
until November. Exactly the same pattern was fol-lowed
by the real interest rates differential (i.e. real rate on
Mexican papers compared to real rates on American Trea-sury
bills).
As Sachs et al. (1996) point out, such a behaviour of
interest rates is markedly different from the one predicted
by standard first generation crisis models (e.g. Krugman,
1979). This fact is presented as a main argument against the
hypothesis that a speculative attack can be a mechanism
used to describe the peso crisis. This point is perhaps some-what
weakened when one takes into account interest rates
differential adjusted using measures of currency structure of
short term public debt [Werner, 1996]. One of the primary
motives for substituting Cetes with Tesobonos was to avoid
an adjustment via higher interest rates. This policy proved
to be rather short sighted as dollar-indexed short-term debt
very quickly reached high levels (see section 1.2.3).
The policy of keeping interest rates low had imme-diate
implications for the level of foreign exchange
reserves in 1994. This basic yet important point is
stressed by Sachs et al. (1995). Until March 1994, the
Mexican private sector was selling securities to foreign
investors at a rate that can roughly be estimated at
around 20 billion USD yearly. This capital inflow financed
Figure 1-3. Interest rates and inflation 1993–1995
80
70
60
50
40
30
20
10
0
-10
-20
CASE Reports No. 39
Cetes rate Real interest rate differential (Cetes vs. US T-Bills) Inflation (12month)
1993M1
1993M3
1993M6
1993M9
1993M12
1994M3
1994M6
1994M9
1994M12
1995M3
1995M6
1995M9
1995M12
Note: all numbers are percent per annum.
Source: author's calculations based on IMF, IFS data.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-15-320.jpg)
![16
Marek D¹browski (ed.)
CASE Reports No. 39
Figure 1-4. Components of monetary base January 1993 – May 1995 (millions of pesos)
120000
100000
80000
60000
40000
20000
0
-20000
-40000
the current account deficit. After March, interest rates
demanded by foreign investors increased significantly yet
the monetary authorities responded by trying to fix inter-est
rates using credit expansion. The Central Bank simply
offered to buy securities accepting low interest rates.
This shows up in the Bank of Mexico accounts as domes-tic
credit expansion to both private sector (mainly banks)
and the government (mainly Tesobonos purchased from
private investors). Such a behaviour did not provide any
incentive to reduce the current account deficit and left no
other way but to finance it from foreign reserves. What
actually happened was that credits (issued in pesos) were
converted into dollars to cover the trade deficit at the
fixed exchange rate.
Another way of looking at the mechanism is to note the
identity decomposing the change in monetary base into the
change of domestic credit and the change in reserves. With
the monetary base being relatively constant, the expansion
of domestic credit was mirrored by declining reserves (see
Figure 1-4).
The most common way of defending the policies of the
central bank was to say that without providing credit,
interest rates would have risen to levels that would se-riously
affect the economy, and that the Central Bank was
forced to act as a lender of last resort to commercial banks
[Sachs et al., 1995; Gil-Diaz, 1998]. Carstens and Werner
(1999) argue that, "in the case of Mexico during 1994 mo-netary
policy had to defend the predetermined exchange
rate, without affecting a weak banking system". Sachs et al.
(1995) recommend, that the credit should be expanded
moderately, while indeed the exchange rate should be
allowed to depreciate. Still, some interest rates hike with
all the adverse effects on economic growth seems to had
been necessary anyway (and such a solution would proba-bly
be less painful than the adjustment through a crisis).
1.2.6. Foreign trade
Mexico experienced a substantial increase in private
spending and trade deficits in 1988–1994. One should note
that this kind of experience is similar to the one of many other
countries that have undertaken exchange rate based stabilisa-tion
programs. The trade deficit almost reached 16 billion
USD in 1992, was somewhat reduced in 1993 and again rose
to 18.4 billion USD in 1994. A deficit of that magnitude did
not reflect weak performance of Mexican exports, which
were growing at an average annual rate of above 10%
between 1990 and 1994. The prospects for Mexican exports
seemed to be very promising, especially after the final
approval of the NAFTA in November 1993. One should note
at that point that the U.S. was by far the most important tra-ding
partner, accounting for more than 81% of exports and
more than 71% of Mexican imports already in 1992. These
shares have increased yet further in later years.
1.2.7. Balance of Payments
The current account balance that was in surplus in 1987
soon turned to negative numbers. The size of deficits
increased significantly after 1990. In 1991 it accounted for
4.6% of GDP, in 1992 and 1993 stayed at about 6.5% to
widen still further to around 8% of GDP in 1994. As shown in
section 2.2 this was primarily caused by private sector invest-ment
exceeding its savings rather than imprudent fiscal poli-
-60000
Monetary base
International reserve
Net domestic credit
1993M3
1993M5
1993M7
1993M9
1993M11
1994M1
1994M3
1994M5
1994M7
1994M9
1994M11
1995M1
1995M3
1995M5
Source: World Bank (1995)](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-16-320.jpg)
![17
The Episodes of Currency Crisis in Latin...
cies. The deficit was financed by high inflows of foreign capital
to the private sector, majority of which was portfolio invest-ment.
With capital inflows higher than the level of the current
account deficit central bank's foreign currency reserves were
gradually increasing from 6 billion USD in 1988 to 25.4 billion
USD in 1993. In order to sterilise capital inflows the govern-ment
issued large amounts of short-term peso- denominated
treasury bills (Cetes) (see also section 1.2.3).
The situation changed markedly after March 1994. The
inflow of foreign capital fell abruptly. The capital account
position from the balance of payment deteriorated from
11.8 billion USD in the first quarter to 3.7 billion in the sec-ond
quarter. In turn, central bank's reserves fell from 29.3
billion USD at the end of February to 17.7 billion at the end
of April, i.e. by 11.6 billion USD. The reserves remained
rather stable at that level until November, when the next
wave of reserve erosion took place. At the end of Novem-ber
they stayed at only 12.9 billion USD. This provoked the
final speculative attack against the peso.
1.3. In Search for the Causes of the
Crisis: Microeconomic Factors
In the years leading to the crisis major positive changes
took place in the real sector environment. The economic
program that was implemented starting in the late 1980s
included several structural reforms. Substantial deregulation
and privatisation took place along with trade liberalisation
[Martinez, 1998]. The Mexican privatisation program was
one of the most comprehensive in the world in terms of
both the size and the number of companies privatised [La
Porta and Lopez-de-Silane, 1999]. Privatisation was most
intensive in the period 1989–1992 and by 1992 the govern-ment
had withdrawn from most sectors of the economy
with the exception of oil, petrochemicals and the provision
of key infrastructure services. This constituted a major
change to the situation from the early 1980s when the state
was intensely involved in the economy through more than a
thousand state-owned enterprises.
The financial system, that until late 1988 was highly regu-lated,
also underwent a quick and substantial liberalisation
[Gelos and Werner, 1999]. All these factors contributed to a
major improvement in perceived prospects of the Mexican
economy and consequently, given the situation in world finan-cial
markets, resulted in large capital inflows to Mexico in the
period 1990–1993. There is no general consensus concerning
the role of financial and real sector weaknesses in the peso
crisis. Also, this channel is not very often thoroughly analysed,
perhaps due to limited access to relevant data.
It is clear that Mexico experienced a rapid expansion of
credit to the private sector (see section 1.2.3). It is likely
CASE Reports No. 39
that this was associated with poor screening of borrowers,
and consequently, declining quality of credit [cf. Edwards,
1999]. Such a process is not unique to Mexico. Lidgren et al.
(1996) highlight the problem of a lack of necessary credit
evaluation skills in formerly regulated banks that are there-fore
unable to use newly available resources more efficient-ly.
Also, the notion that banks problems often precedes the
financial crisis (devaluation) has strong support from other
cross-country analyses [e.g. Kaminsky and Reinhart, 1996;
Lidgren et al., 1996]. Gil-Diaz (1998) points to several caus-es
of the rapid debt increase, the speed of which over-whelmed
supervisors, e.g.:
– speedy and not always well prepared privatisation of
banks, sometimes with no respect to "fit and proper" criteria,
either in the selection of new shareholders or top officers,
– lack of proper capitalisation of some privatised banks
and involvement in reciprocal leverage schemes,
– lack of capitalisation rules based on market risk; this
encouraged asset-liability mismatches that in turn led to a
highly liquid liability structure,
– loss of human capital in banks during the years when
they were under the government; banking supervision
capacity not meeting the requirements of increases in
banks' portfolios.
1.4. Political Situation
The Mexican peso crisis provides a clear and very inte-resting
example of how political factors can contribute to a
financial turbulence. The series of unexpected events in po-litics
had a visible influence on the behaviour of economic
aggregates and certainly played an important, through hard
to measure, role in triggering the December crisis.
The first of the series of events started on New Year's
Day 1994 when peasants in the southern Mexican state of
Chiapas began a rebellion by taking over six towns. Even
though the uprising was rather quickly suppressed, it
remained an issue in internal political life and occasionally
flared up again. Especially before the August elections, the
rebellion was again discussed and was recalled to question
the extent of popular support for the government's eco-nomic
program. On 23 March, Luis Donaldo Colosio, the
presidential candidate of the ruling party, was assassinated.
The causes of this murder were never actually revealed, and
there were signs that it might had been associated with ten-sions
within the ruling Institutional Revolutionary Party (PRI)
[World Bank, 1995]. It should be noted that the PRI gov-erned
Mexico since 1929 as a party organised around well
connected political families. The assassination brought se-rious
turbulence to the financial markets with the peso at
the ceiling of the intervention band and sharply higher inte-](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-17-320.jpg)
![18
Marek D¹browski (ed.)
CASE Reports No. 39
rest rates. Improper response of authorities to the turmoil
in the end of March and in April set the stage for the
December crisis [Sachs et al., 1995].
The situation seemed to have calmed down when
another candidate of the PRI, Ernesto Zedillo won the pre-sidential
elections on 21 August. He received above 50% of
votes that came as a surprise to many observers. It is indeed
hard to verify whether the elections were free of fraud. On
the other hand, 1994 elections were perhaps more demo-cratic
and fair than many previous elections in Mexico. The
outcome of the elections was generally considered positive
from the financial stability point of view but may also have
caused the authorities to believe that the worst of the insta-bility
was over. One month after the elections, on 28 Sep-tember,
PRI leader Jose Francisco Ruiz Massieu was assassi-nated.
This murder remained a mystery too with several
high officials of PRI possibly somehow implicated.
All these developments certainly changed the position of
Mexico's traditional ruling party. The process of reform in
the political scene actually began slightly earlier, and, as
Dornbusch and Werner (1994) point out, the rapid embrace
of greater openness has shattered the PRI coherence, so
that the old corporatism became unmanageable. It is thus
clear that during the whole 1994 there were serious ten-sions
within the ruling elite and the uncertainty about Me-xico's
political future was a factor in the foreign perception
of the country's financial stability.
It should also be noted that the years of presidential
elections have traditionally been associated with financial
turbulence. This fact was recalled in many analyses of the
2000 presidential elections. Also, as many authors agree the
long period between the voting and taking the office by the
president elect has a negative impact on the quality of go-verning
the country. President Zedillo took office on 1
December and it was followed by the intensified unrest in
Chiapas.
1.5. Crisis Management
While the policy mistakes of most of 1994 played an
important role in triggering the December crisis, improper
handling of the initial devaluation on 20 December probably
exacerbated the crisis.
The devaluation on 20 December was announced after
weeks of assurances that the government was committed to
the previous exchange rate system. The announcement was
made by the Finance Minister on radio and television rather
than through an official channel. As the World Bank (1995)
stresses, such a way of publicising such a major policy
change angered investors. Also, it turned out that business
leaders were consulted before the devaluation, thus having
the opportunity to make profits at the expense of unin-formed
foreign investors [Krugman, 1997]. The devaluation
was widely considered insufficient and the exchange rate
immediately depreciated to the ceiling of the band, i.e. by
15%. The authorities had sacrificed the credibility without
satisfying market expectations. The rush out of the country
continued on the next day with Central Bank's reserves
reportedly falling below 6 billion USD [World Bank, 1995].
President Zedillo affirmed the commitment to the new
band, but on the next morning the government let the peso
float. During the day it depreciated by a further 15%. On 26
December the planned press conference by the Finance
Minister on the government anti-crisis plan was cancelled at
the last moment. On the next day the peso depreciated to
5.45 pesos to the dollar. The auction of dollar denominated
government bonds attracted almost no bids. Increasing
prices made labour leaders to demand wage negotiations.
On 29 December a new Finance Minister, Guillermo Ortiz
Martinez was appointed, who within a few days announced
a new economic program.
On 3 January, Stanley Fisher, Acting Managing Director
of the International Monetary Fund, made a statement
expressing the Fund's support for this program and
announcing the establishment of the Exchange Stabilisation
Fund of 18 billion USD with contributions under the
NAFTA from the monetary authorities of other major
countries as well as from private investors. The talks on a
stand-by credit from the IMF started a few days later and
an 18-month credit of 17.8 billion USD was finally
approved on 1 February.
The Mexican authorities' program constituted of three
main components: minimising the inflationary pressures of
the devaluation, pushing forward structural reforms to sup-port
and promote competitiveness of the private sector, and
to address short term concerns of investors and establish a
coherent floating exchange regime. To the end of the first
objective, a National Accord was set among workers, busi-ness
and government to prevent wages and prices hikes, the
government spending were to be reduced by 1.3% of GDP,
and cuts in credits from state development banks were to
be implemented.
In terms of structural reforms President Zedillo pledged
to propose amendments to the constitution allowing for pri-vate
investment in railroads and satellites, to open the
telecommunication sector to competition, and to increase
foreign participation in the banking sector. As the third
objective is concerned, the co-operation with investment
banks in order to address the issue of Tesobonos was
announced, as well as creating a futures market in pesos,
and commitment to a tight monetary policy. In the beginning
of March, the finance minister announced a package of fur-ther
measures aimed at strengthening the program. These
included substantial increases in prices charged by public
enterprises, VAT rate hike, and public expenditure reduc-tions,
and were designed to allow the public sector to stay](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-18-320.jpg)
![19
The Episodes of Currency Crisis in Latin...
in surplus in 1995. Also, a further reduction of development
banks was declared.
The Mexican Rescue Plan was supported with what was
then the biggest ever financial support package. The initial
amount of 18 billion USD announced by Fisher at the begin-ning
of January after further intense discussions rose to the
range of 25–40 billion USD in mid- January and finally
around 52 billion USD of loan guarantees and credits at the
end of February. This package included 20 billion USD of
loan guarantees from the U.S. government, 17.8 billion USD
stand-by credit from the IMF (by-then the largest ever
financial package approved by the IMF for a member coun-try
both in terms of the amount and the overall percentage
of quota, 688.4%), 10 billion USD from central banks via
the BIS, and several billion dollars from other American gov-ernments
[World Bank, 1995].
The unprecedented size of the support package brought
about several controversies, especially in the U.S. The Clin-ton
Administration was criticised heavily both for lack of
action before the crisis, and for too much engagement in
co-ordinating the support package. On 29 March 1995,
Undersecretary of the Treasury, Lawrence Summers
defended in closed hearings in the Senate the Administra-tion's
failure to publicise a warnings on the situation in Me-xico.
He admitted that the U.S Treasury lost the confidence
in the peso before the dramatic devaluation took place but
did not want to set off a market run on Mexico by making a
public statement about the situation [Burkart, 1995]. The
main arguments backing the support package pointed to the
fact that Mexican economy was illiquid rather than funda-mentally
insolvent [DeLong et al., 1996]. In retrospect it
seems that this view was indeed right, even though the U.S.
engagement in the package was to a large extent motivated
politically, i.e. by fears of possible political destabilisation in
the neighbouring country [Krugman, 1997].
1.6. Conclusions
The above presentation of several key factors and their
possible role in explaining the crisis shows that there is still
no clear consensus on the issue. Several aspects did play a
role and only their joint impact led to the abrupt events of
end of December 1994. Various models were used to
describe the crisis. These were both models of the second
generation type, pointing to the role of self fulfilling expec-tations
and the political and economic constraints faced by
the authorities, as well as modified first generation models
stressing the importance of economic fundamentals. With
respect to the causes of the crisis following general points
can be made:
– Private sector savings did not match the level of invest-ment.
Mexico had easy access to credit as a result of the si-tuation
CASE Reports No. 39
in the world financial markets, and liberalisation of
the economy. Resulting credit expansion was not accompa-nied
by proper credit screening.
– Mexico was relying too heavily on foreign borrowing
in 1993 and early 1994, having no easy escape route in the
case this inflow would stop.
– The exchange rate rule was perhaps not quite consis-tent
with the developments in other spheres (overindul-gence
of credit, excess of funds in international financial
markets, fast growth of short-term debt, financial liberalisa-tion).
The real overvaluation of the peso might also have
played some role.
– Mexico experienced a series of unexpected negative
shocks during 1994 – the rise in U.S. interest rates coincid-ing
with political tensions in the country.
– Mexican politics in 1994 did play an important role in
the crisis.
– Lack of availability of timely and accurate information
on the economic situation in the country might have played
some role in the abrupt change of investors' attitude
towards Mexico.
– The policy response to the shocks of early 1994 was
certainly inappropriate (this is an ex post diagnosis).
– Neither fiscal nor monetary policy tools were used to
adjust the economy to a worsening situation during 1994.
– Allowing for the erosion of foreign reserves of that
extent while building a large and rapidly growing stock of
dollar indexed short term debt in the period March-
December 1994 was an extremely risky strategy that did
not work. This set the stage for the December crisis and
then led to very high interest rates and, consequently, harsh
consequences for the real sector.
– Perceived risk of financial collapse played a role in both
causing the collapse and making it very severe.
– Inappropriate management of the devaluation and
improper steps taken in the days following it led to a com-plete
loss of confidence in Mexican policies and conse-quently
to more severe consequences.
An interesting feature of the Mexican crisis is the seve-rity
of the recession that was caused by it. On the other
hand, the crisis was relatively quickly overcome and the
economy seems to have overcome its underlying causes.
One of the possible explanations of such developments
might be that the private sector was indeed heavily depen-dent
on external financing. Then again, a relatively quick
rebound of the economy could suggest that it was funda-mentally
sound, and the crisis exposed it to the liquidity
trap. In other words, given the abundance of credit, the pri-vate
sector was using it heavily and possibly sometimes
unwisely, but exposed to the dramatic change in the exter-nal
environment was still able to become competitive again.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-19-320.jpg)
![23
The Episodes of Currency Crisis in Latin...
Part II.
The 1995 Currency Crisis in Argentina
by Ma³gorzata Jakubiak
2.1. Introduction
This paper presents the economic developments that
took place in Argentina at the time of the 1995 currency cri-sis.
This financial turbulence resulted from the contagion of
the Tequila crisis of the late 1994. And although the country
maintained its commitment to a peg under the currency
board arrangement, the reserves of the central bank were
severely depleted and the consequences for the economy
manifest.
The study starts from a description of the economic
reforms of the early 1990s that set the framework for the
monetary and fiscal policies in place when the crisis hit.
There is then a discussion on the macro- and microeconom-ic
climate, followed by the description of policy responses to
the crisis. The paper concludes with the assessment of
whether the core factors that drove the crisis have been
properly addressed by looking at the post-crisis situation.
2.2. Overview of Economic Situation
Before and During the Crisis
2.2.1. Reforms of the Early 1990s and Crisis
Developments
The 1980s were marked by a series of economic and
financial problems. One of them was chronic inflation and
periodic hyperinflation which led to widespread dollariza-tion
of the economy. Moreover, prevailing public sector
deficits were crowding out private sector credit (Garcia-
Herrero, 1997). A banking crisis erupted in 1980, and then
developed into a severe currency crisis a year later. The
next crisis started in 1985 and ended in 1987. The subse-quent
economic plan aimed at lowering inflation resulted in
its outburst in 1989 and the complete dollarization of the
economy as investors' confidence weakened. Two huge
devaluations took place in the late 1989 and in 1990. After
another bout of financial turbulence, one more stabilization
CASE Reports No. 39
plan failed. The exchange rate, which under earlier plans
had been pegged, was allowed to fluctuate, and all price
controls were removed (Choueiri and Kaminsky, 1999).
The difficulties of the period 1989–1990 – mainly as a result
of hyperinflation – allowed for the general recognition of
the need for reforms.
The changes, that put the country on a sustained growth
path began with the Convertibility Plan of 1991. The Convert-ibility
Law, which is still in operation, established the curren-cy
board arrangement. Financial sector reforms followed.
Argentina also eliminated, by 1993, restrictions on capital
flows, relaxed or abolished barriers on imports and exports
and deregulated trade and some professional services. Until
1994, roughly 90% of all state-owned enterprises was pri-vatized,
bringing considerable gains to economic efficiency
[IMF, 1998]. In 1994 Argentina recorded economic growth
of 8%, managed to lower inflation to 4.2% (from over 10%
in 1993 and 25% in 1992), and kept the budget deficit of the
federal government at the level of 0.5% of GDP. The ban-king
sector recorded growth in credits and deposits. From
the early 1990s, capital started to flow in, as a result of
investors' more favourable perception of the region and rel-atively
high interest rates.
At the time, interest rates in the United States were low,
and net flows of portfolio capital to Argentina amounted to
33.7 billion dollars in 1993 and to 8.4 billion in 1994. These
huge capital inflows led to an explosion of domestic credit,
consumption, real estate and stock market booms, and lack
of diversification of bank portfolios [Choueiri and Kaminsky,
1999]. The current account deficit deteriorated as a result
of real exchange rate appreciation. As the U.S. and world
interest rates soared in 1994, the Mexican peso was deva-lued
in December 1994 and the capital outflows brought
about balance of payment pressures, the rumors about
abandoning the currency board spread out. The Argentine
banking system suffered from a run on deposits, and the
credit crunch followed. Between December 1994 and
March 1995, the central bank (BCRA) lost 41 percent of its
international reserves, defending the peso-dollar peg. The
banking system lost 18% of its deposits in five months
which generated liquidity problems. A number of prudential
regulations were introduced in early 1995 in order to](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-22-320.jpg)
![24
Marek D¹browski (ed.)
CASE Reports No. 39
restore confidence in the banking system. By the end of the
1995, the deposits went nearly back to their pre-crisis le-vels
and the monetary authorities managed to restore its
gross international reserves [IMF, 1999]. The currency
board was defended. However some smaller banks conti-nued
to experience problems, and the crisis resulted in a
severe recession. In 1995, real GDP went down by 4%.
Among the core factors that allowed the transfer of
external shocks to the real economy and leveraged the cri-sis
were weak links with world financial markets, relatively
underdeveloped domestic financial markets, labor market
rigidities, systematic crowding out, and real exchange rate
inflexibility imposed by the currency board regime
[Caballero, 2000]. All these factors are addressed in the fol-lowing
sections of the paper.
2.2.2. Monetary and Exchange Rate Policy
From 1991 Argentina has followed a currency board
exchange rate arrangement. The currency board is – after
the classical monetary union – the second most rigid form of
the exchange rate regime. In this orthodox form of a fixed
exchange rate regime the role of monetary authorities is
reduced to issuing notes and coins that are fully backed by a
foreign reserve currency on demand at a fixed exchange
rate. There is a minimum of 100% backing of reserve
money by net foreign assets of the central bank, and cur-rency
boards often hold excess reserves to offset against
asset valuation changes. These excess reserves are related
to the net worth of a currency board (Pautola and Backé,
1998), because seigniorage can be earned only from interest
on reserves.
The 1991 Convertibility Law and the 1992 Central Bank
Charter created the basis for the functioning of the curren-cy
board in Argentina. The exchange rate of the Argentine
peso was fixed at one against U.S. dollar, and the central
bank was required to keep 100% of its monetary base in
international reserves. However, since 1995 1/3 of it may be
kept in the safe dollar-denominated government bonds
[Hanke, Schuler, 1999] while holdings of these securities
cannot grow by more than 10% per year. These re-gula-tions
eliminated the possibility of inflationary financing of the
government deficit. Moreover, the charter restrains the
central bank from financing provincial or municipal govern-ments,
public firms, or private non-financial sector [Pou,
2000]. The central bank became fully independent from the
legislative and executive branches of government, and set its
principle goal at maintaining the value of domestic currency.
As can be seen from the above description, the Argen-tine
currency board is not the strictest form of a currency
board where monetary authorities cannot intervene in the
market, cannot act as a lender of last resort, and where
interest rats are solely market-determined. Indeed, the
Central Bank of Argentine Republic (BCRA) has some room
for discretionary monetary policy, because its international
reserves are not fully backed by the domestic currency and
because it can set reserve requirements for commercial
banks. This ability to retain some flexibility was used during
the 1995 financial crisis.
The international reserves of BCRA started to shrink
quickly in January 1995 when the Argentine peso came
under a pressure, and when the bank tried to rescue trou-bled
commercial banks (see Figure 2-1). The reserves hit
the lowest level in March 1995, which was around 2/3 of the
monetary base, the minimum coverage requirement. BCRA
Figure 2-1. International reserves and monetary base coverage, 1994–1996
17000
16000
15000
14000
13000
12000
11000
10000
9000
8000
7000
1.4
1.2
1
0.8
0.6
0.4
0.2
0
millions of USD
Total reserves minus gold
(foreign exchange+SDPs)
Reserves/Reserve Money
1994M1
1994M3
1994M5
1994M7
1994M9
1994M11
1995M1
1995M3
1995M5
1995M7
1995M9
1995M11
1996M1
1996M3
1996M5
Source: own calculations on the basis of IFS data](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-23-320.jpg)
![25
The Episodes of Currency Crisis in Latin...
Table 2-1. Interest rates, 1994–1995
1994 1995
10 11 12 1 2 3 4 5 6 7 8 9 10
Prime deposit rate 8.27 8.72 9.55 10.65 11.64 19.38 19.07 15.54 10.83 10.24 9.17 9.21 8.92
Lending rate 9.83 10.00 13.56 18.06 19.06 34.05 26.45 22.13 16.19 14.57 13.29 13.26 12.55
Real interest rate
1.94 1.99 2.08 2.17 3.30 11.65 11.36 8.40 4.29 4.07 3.31 3.86 3.73
differential
Source: IFS, own calculations
Note: real interest rate differential is calculated as the difference between real Argentine deposit rates and real U.S. deposit rates
was then buying dollar-denominated Treasury bonds, in
order to mitigate the effects of the credit crunch. The cen-tral
bank's holding of these government notes increased by
25% from 1994 to 1995, and declined sharply afterwards
[Caballero, 2000]. This decline is reflected in the Figure 2-1
as the high indicators of foreign exchange reserves in 1996.
As the central bank was depleting its reserves, interest
rates rose sharply, reflecting a domestic liquidity squeeze
and rise in the country risk premium. The real interest rate
differential vs. U.S. deposit rate reached above 11 percent-age
points, while during 1994, its value was close to 2. High
interest rates induced the private sector to lower demand
for credit [1] and reduce expenditures. Banks cut their cred-it
lines and refinancing facilities. All these factors contributed
to a decline in the economic activity in 1995 and to higher
unemployment [Catao, 1997].
2.2.3. Fiscal Policy
Through the early Convertibility Plan years, the govern-ment
was running budget deficits. The central government
budget deficit averaged 0.5% of GDP during the years
1991–1994 – with a surplus only in 1993. The deficit of the
consolidated public sector (including federal budget, provin-cial
government budgets, and off-budgetary funds and pro-grams)
averaged 1.8% of GDP, with 2.5% of GDP in 1994.
The crisis year of 1995 was marked with the 1.5% deficit of
central government budget, and 4.3% deficit of the whole
public sector.
According to the IMF calculations, fiscal policies were
pro-cyclical in the early 1990s, with public sector deficits ris-ing
faster than the cyclically adjusted public sector balance
[IMF, 1999]. During this period (1991–1994), when output
was growing above potential and fiscal impulse was expan-sionary,
the lowering of inflation has been achieved by the
nominal exchange rate anchor and the supply-side oriented
reforms, such as change in the tax system, and elimination of
distortionary tariffs. Only from 1995 onwards, as the coun-try
slipped into recession, did the fiscal impulse start to have
a negative impact on demand, thus contributing to the
CASE Reports No. 39
reduction, and finally to the elimination of inflation [IMF,
1999].
Tax reform aimed at eliminating some taxes and shift-ing
the relative tax incidence from production to con-sumption
and incomes was implemented in the early
1990s. Many distortionary taxes, such as those on
exports, bank debits and assets, with a yield about 3% of
GDP, were removed. Some exemptions, mainly from VAT,
as well as subsidies, were abolished. To improve labor
market flexibility, the government significantly reduced
the employer payroll tax in some sectors (reversed for a
couple of months during the 1995 crisis). There was also
a significant decline in the number of workers employed
by the state as efforts to improve efficiency in the public
sector were undertaken (the provinces started to be
responsible for the health and education services [IMF,
1999]). New and stronger laws increased the govern-ment's
ability to control tax evasion.
Pension reform, aimed at shifting from the pay-as-you-go
publicly funded system to a system combining public
transfers and private capitalization, started in mid-1994.
The reform resulted in the reduction of future liabilities of
the public sector. However, the immediate costs for the
budget are estimated to be around one percent of GDP
per year, as the government pays the contribution to the
private system for those who voluntarily opted to shift
away from the pay-as-you-go scheme. In any event, these
costs appeared first in the consolidated budget in 1996,
exactly one year after the currency crisis, and thus the
consequences of this reform for the fiscal sector are not
explored further.
2.2.4. Private and Public Debt
The reforms of the early 1990s allowed Argentina to
return to the voluntarily financing of its external debt, which
was rescheduled under the Brady Plan [Pou, 2000]. The
developments in borrowing from the international capital
markets, both public and private, have moved in the direc-tion
of declining spreads, the lengthening of maturity, and
[1] There were also other factors contributing to the decline of the private sector demand for credit, such as crowding out by government bor-rowing,
which turned to domestic banks for financing its monetary interventions in 1995, such as providing troubled banks with fresh credit.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-24-320.jpg)
![27
The Episodes of Currency Crisis in Latin...
European countries in the mid-1990s, or than the invest-ment
ratios of Southern European countries in the 1980s.
It can be seen from the comparison of capital flows
that the majority of this domestic resource gap was
financed through foreign direct investment.
2.2.6. Foreign Trade
One of the reforms of the Convertibility Plan involved
the elimination of all tariffs on exports and the majority of
non-tariff barriers on imports. Imports tariffs have been
cut form over 40% average rate in 1989 [IMF, 1998] to
8.4% at the end of 1994, with a zero rate on capital goods
and raw materials. This eliminated distortions in foreign
trade.
Argentina is an exporter of raw materials and some
lightly processed primarily products. These primary and
agro-industrial products account for 70% of all exports,
and are subject to high fluctuations in world prices. Since
1990, there has been a quick rise of manufacturing exports
to Brazil, following the creation of MERCOSUR (South
American customs union). But still, in 1994 this number
was below 25% of total value of exports. Generally,
because of high commodity concentration, Argentine
exports have been highly volatile.
Following the removal of restrictions and high eco-nomic
growth, Argentine imports between 1991 and 1997
was growing more than four times as fast as real GDP.
Catao and Falcetti (1999) estimated that long-run income
elasticity of imports was above 2 during this period, and
that this elasticity showed some pro-cyclical behavior. It
should be noted that between 1991 and 1995, the real
exchange rate kept appreciating, which also explains the
rapid growth of imports at this time, but not as much as
booming economic activity. Over 50% of imports in 1994
came from the US, the EC and Japan.
Trade deficits in place since the early 1990s were main-ly
driven by high dependency on world prices for exports,
soaring domestic demand and incomes, together with real
appreciation toyical or most emerging markets. Trade sur-pluses
of 1995–1996 were taking place because world
prices for traditional Argentine exports increased, the
economy stepped into one-year long recession and there
was a fall in the real effective exchange rate during the
Tequila crisis.
2.2.7. Balance of Payments
Argentina has been systematically running current
account deficits in the 1990s, which is a typical feature of
an emerging economy. Even in the absence of trade
deficits, as in 1995 when imports fell because of the reces-sion,
the negative current account balance was caused by
the outflows of the investment incomes. As 1995 showed,
the current account deficit had to be financed by foreign
borrowing, following changes in the investors' preferences
and the sudden outflow of short-term capital.
It should be noted that there have been large inflows
of portfolio capital during the two years preceding the cri-sis.
Net inflows of short-term capital in 1993 amounted to
Figure 2-2. Foreign trade, 1993-1997
9000
8000
7000
6000
5000
4000
3000
2000
1000
0
CASE Reports No. 39
Imports Exports
1993Q1
1993Q2
1993Q3
1993Q4
1994Q1
1994Q2
1994Q3
1994Q4
1995Q1
1995Q2
1995Q3
1994Q4
1996Q1
1996Q2
1996Q3
1996Q4
1997Q1
1997Q2
1997Q3
1997Q4
Source: IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-26-320.jpg)
![28
Marek D¹browski (ed.)
around 15% of Argentine GDP, which was at that time 16
times more than the net inflows of long-term investment.
However, Foreign Direct Investment inflows have been
growing systematically during the analyzed period,
notwithstanding the 1995 decline in real output. The pri-mary
reasons being prevailing international conditions, the
implementation of structural reforms by Argentina, priva-tization,
and the elimination of restrictions on foreign
investors, as well as the removal of restrictions on capital
transactions in general [IMF, 1998]. Since the beginning of
the 1990s, there have been no capital controls on financial
or commercial operations between residents and nonresi-dents
[BCRA, 2000].
The significant rise in FDI flows in 1995 and in 1996 can
be attributed mainly to the creation of the private pension
funds, and to the sales of private firms to foreign investors
CASE Reports No. 39
Table 2-5. Balance of payments (mil. USD), 1992–1997
1992 1993 1994 1995 1996 1997
Current account balance -5 521 -8 030 -10 992 -4 985 -6 521 -11 954
Trade balance -1 396 -2 364 -4 139 2 357 1 760 -2 123
Exports of goods 12 399 13 269 16 023 21 161 24 043 26 431
Imports of goods -13 795 -15 633 -20 162 -18 804 -22 283 -28 554
Non-factor services (net) -2 463 -3 221 -3 692 -3 326 -3 366 -4 178
Investment income -2 393 -2 931 -3 567 -4 529 -5 331 -6 089
Current transfers (net) 731 486 406 513 416 436
Capital and financial account 7 350 20 328 11 155 4 623 11 175 16 826
Direct investment (net) 3 218 2 059 2 477 3 818 4 922 5 099
Portfolio investment (net) 4 513 33 731 8 389 1 864 9 727 11 087
Other investment (net) -381 -15 462 289 -1 059 -3 474 640
General government -1 343 -10 196 969 1 197 -199 136
Bank 76 -570 761 2 570 -2 744 -1 615
Other sectors 739 -697 -1 423 -4 832 -567 2 130
Net error and omissions 54 -1 173 -872 -1 853 -1 316 -1 498
Overall balance 1 883 11 125 -709 -2 215 3 338 3 374
Financing -1 883 -11 125 709 2 215 -3 338 -3 374
Reserve assets -3 264 -4 279 -685 82 -3 875 -3 293
Use of IMF credits -73 1 211 455 1 924 367 -38
Exceptional financing 1 454 -8 057 938 209 170 -43
Source: IMF IFS
Figure 2-3. Private and public net capital flows to non-financial sector, 1992–1997
4. 0%
3. 0%
2. 0%
1. 0%
0. 0%
-1 .0%
-2 .0%
1992 1993 1994 1995 1996 1997
% of GDP
Public non-financial sector Private non-financial sector
Source: own calculations based on the data from Argentine Ministry of Economy and IMF IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-27-320.jpg)
![29
The Episodes of Currency Crisis in Latin...
[IMF, 1998]. Net FDI flows accounted for around 1% of
GDP in 1992–1994 and for 1.7% on average, during the
period 1995–1997. However, there was almost no change
in the aggregate existing foreign investment stock, which
averaged around $3.5 billion per year during the period
1992–1995. Its structure, though, has been changing, indi-cating
faster growth of FDI non-related to the privatiza-tion
opportunities. FDI grew most rapidly in the commu-nications
and manufacturing sectors. Significant privatiza-tion-
related investment flows were recorded also for the
electricity, gas and water and for the petroleum industries.
Around 40% of all FDI coming to Argentina during
1992–1995 originated in the USA.
Despite the important role of huge inflows of portfolio
investment in 1993, capital inflows (both private and pub-lic)
constituted a relatively small fraction of GDP for a fast
growing economy. During 1992–1994, while GDP was
increasing by 7.8% per year, overall capital flows [2]
amounted to 4.9% of GDP. This suggests that the link of
Argentine financial market with international financial mar-kets
was weak. As it was visible in 1995, this significantly
constrained government ability to use international financ-ing
when there was an external shock. Official capital
flows depicted on the Figure 2-3, rose in 1995 supported
by loans from IADB and the World Bank.
2.2.8. Real and Nominal Rigidities
As there was little room for monetary policy, the bur-den
of adjustment during 1995 fell on wages and prices.
Argentina has a European-style labor market with centra-lized
bargaining, high severance costs, and still high – on
average – wage taxes. These rigidities – both nominal and
real – amplified external financial shock by forcing a larger
share of adjustment on output and employment. This was
costly, since unemployment has stayed well above 10%,
even in 1999.
In addition, real exchange rate inflexibility, brought
about the convertibility regime, combined with labor mar-ket
rigidities and limited access to financial markets made
the 1995 fall in output worse [Caballero, 2000].
2.2.9. Banking System
The introduction of the Convertibility Plan markedly
changed the Argentine banking sector. High inflation and
macroeconomic volatility of the 1980s, together with large
capital flight, caused the demand for domestic money to
decline heavily. The ratio of broad money (M3) bottomed
at 6 percent of the GDP in 1990, the overall sum of
deposits of the banking system was low, and real interest
rates on deposits were negative. Thus, the banking system
reforms of the early 1990s focused primarily on the
strengthening of the whole system and on removing obsta-cles
to financial intermediation [IMF, 1998].
The introduction of the simplified system of reserve
requirements began in 1991. The rates on both foreign
and domestic currency transactions were unified, and the
adherence to these requirements was made more effec-tive.
New capital adequacy requirements, incorporating
lending interest rate risk factor, were put in place, and
they have been gradually tightened. The loan classification,
portfolio risk rules and provisioning rules were introduced
in 1994. The supervisory role of the financial superinten-dence
was reinforced in order to verify compliance with
all the prudential standards. The overall result of these
regulations was visible in the fall of outstanding central
bank credits to financial institutions [IMF, 1998]. The risk-weighted
capital to assets ratio rose to 11.5% in January
1995, well above the 8% Basle standard [Pou, 2000]. The
reserve requirements averaged 17.5% of deposits [IMF,
1998].
The reforms of the early 1990s removed the barriers
of entry and increased competition between banks. They
were aimed at ensuring the safety of individual banks and
the whole banking system. They also intended to reduce
moral hazard. These issues were very important, since
under the currency board, the BCRA role as the lender of
last resort was very limited. Basically, the central bank
was not allowed to provide liquidity to the banks in finan-cial
trouble. The BCRA was able to conduct intervention
through repo operations, but this was limited to the
smoothing of fluctuations in the interbank market. The
Convertibility Law allowed extending credit to financial
institutions in the emergency situation only [IMF, 1998].
2.2.9.1. Developments during December 1994-
March 1995
As the fears about the possible abandoning of the peg,
triggered by the Mexican devaluation, intensified, and as
Argentine debt prices kept falling, banks experienced a
run on deposits. This situation is shown on the Figure 4,
describing effective growth rates of loans and deposits
[3]. As the depositors started to withdraw their funds, it
initially affected wholesale banks, whose loan portfolios
were composed mainly of government bonds, then
spread to the entire banking system. In 1995, the BCRA
[2] Together with banking sector flows.
[3] Time series of deposits and loan growth rates, each less their respective real interest rate
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-28-320.jpg)
![30
Marek D¹browski (ed.)
Figure 2-4. Effective growth rate of deposits and loans, 1994–1997
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
-5.0%
-10.0%
-15.0%
-20.0%
1994M6
1994M9
1994M12
1995M3
1995M6
1995M9
1995M12
started to use its reserves to provide funds to banks. The
central bank lost nearly 20% of its monetary base cove-rage
[IMF, 1998] [4], which was the limit of the interven-tion
allowed as a response to crisis situation. Despite
these efforts, the loss of deposits resulted in a huge cre-dit
crunch.
It should be noted that an important feature of the
Argentine banking system at the time were the problem
loans, accounting for more than 10% of the total loan
portfolio in 1994. Moreover, they have not been uni-formly
distributed [IMF, 1998]. This factor was responsi-ble
for some of the changes in the banks' assets that hap-pened
in the early 1995. Another important factor was
the lack of official deposits insurance, which, coupled
with the restricted role of the BCRA as a lender of last
resort, intensified the perception of deposit risk.
During the early phase of the run of deposits, there
has been a visible shift to quality. First of all, this meant
that depositors started to convert peso deposits to dollar
deposits, expecting the devaluation. This is shown on Fi-gure
5. While the peso deposits of the whole banking sys-tem
have been falling since December 1994, the dollar
deposits were still growing in February 1995. Then the
fall in the dollar deposits was less pronounced, and after
mid-1995 they quickly started to build up again. At the
same time, peso deposits stayed at a relatively unchanged
level until the end of the year. We can see that the reco-very
of deposits in the Argentine banking system towards
1996M3
1996M6
1996M9
1996M12
1997M3
the end of 1995 was attributable mainly to the increase in
the amount of foreign currency deposits.
Secondly, the non-uniform distribution of the problem
loans meant that small public provincial banks had a dis-proportionately
bigger share of non-performing loans
than the larger banks. When the run on deposits started,
these small banks suffered more, as their depositors start-ed
to move funds to the larger banks. Public provincial
banks lost their market share (from 12.8% of total assets
in 1994 to 9.6% in 1995), while private banks gained.
While 56.7% of total assets belonged to the private banks
in 1994, the share increased to 58.6% at the end of 1995.
The market share of the large national banks remained
relatively unchanged at around 30% of total assets [Bur-disso
et. al, 1998]. It is claimed that although part of the
small banks market share was lost as a result of privatiza-tion
following the crisis, an important fraction was gone
due to a change in the market perception of their credit
risk.
Another way of looking at the situation of the small
banks during the Tequila crisis is to examine their indica-tors
of profitability. Return on total assets of the 20 largest
banks remained relatively stable and generally positive
during the years 1994–1997, while the profitability of total
retail banking sector was much more volatile, often nega-tive,
and fell sharply during the first months of 1995 [Bur-disso
and D'Amato, 1999]. This situation increased the
contagion, as some banks have virtually found themselves
CASE Reports No. 39
[4] Around 30% of usable foreign exchange reserves according to the author's calculations (see Figure 2-1).
1997M6
1997M9
1997M12
Loans Deposits
Note: The term "effective" refers to the annual growth rates of credit and deposits less the real lending and the real deposit rates, respectively
(after Caballero, 2000). PPI year-over-year inflation was used in the construction of real interest rates.
Source: own calculations based on IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-29-320.jpg)
![31
The Episodes of Currency Crisis in Latin...
Figure 2-5. Peso and dollar deposits of the deposit money banks, 1994–1997
70000
60000
50000
40000
30000
20000
10000
0
short of liquid assets what affected even a number of
apparently solvent institutions [IMF, 1998]. The full-scale
run on banks started in February 1995.
2.2.9.2. Role of Private, Public and Foreign
Ownership
Although the privatization process started in early
1993, following the remonetization of the Argentine econ-omy
and the redefinition of the role of the public sector,
only after the Tequila crisis did this process gain momen-tum.
Until early 1995, only 3 banks had been privatized
[Burdisso et. al, 1998]. At the end of the 1994 there were
135 private banks – both foreign and domestically owned
– in Argentina. They accounted for over 50% of the whole
banking system assets. 33 public banks were on average
larger: they had 30% (national) and 13% (provincial) of
the market share [Burdisso et. al, 1998].
Following the removal of restrictions on foreign
investment and capital flows of the early 1990s, the num-ber
of foreign banks in Argentina increased. However, in
1994 they still accounted for under 20% of system assets
only [Goldberg et. al, 2000]. When comparing their loan
portfolios to those of the state-owned banks, it shows
that the foreign banks had lower mortgage shares and
higher shares of commercial, government, private and
other lending. Foreign banks were similar under this
respect to the domestic private banks. However, the for-eign
banks were perceived as generally safer and health-ier.
The loan growth rates of the foreign banks were sub-stantially
higher than the respective rates of the domes-tic
banks in 1994, and they continued to grow faster even
during the crisis period [Goldberg et. al, 2000].
According to the research of D'Amato et. al (1997),
bank "fundamentals" – such as their profitability and level
of interest rates – as well as the overall macroeconomic
situation were very important in driving the dynamics of
deposits. However, there is evidence of contagion effects
in the group of small and medium-sized banks on which
public information was poorer.
2.2.10. Domestic Financial Market
The financial markets in 1994–1995 had tenuous links
with the international markets. As already mentioned,
the relatively small inflows of capital relative to the size
of the economy bear this out. Another argument in favor
of the weak link with the world markets is the fact that
foreign capital focused mainly on large enterprises, and
that the smaller companies had difficult access to the
international markets. During the early months of 1995,
the volatility of the stock index for prime companies and
the overall stock index substantially differed. The volatil-ity
of stock index for prime companies (MERVAL)
increased significantly, while the volatility of the total
market index remained relatively unchanged [Caballero,
2000]. Notwithstanding volatility, values of both MERVAL
(see Figure 2-6) and total stock market index have been
relatively low from December 1994 until the end of
1995.
There was also a large spread-premium on Argentine
sovereign bonds relative to the U.S. throughout this time,
which further confirms the country risk-premium and its
weak ties with international financial markets. Moreover,
Argentine markets are still underdeveloped by interna-
CASE Reports No. 39
peso deposits dolar deposits
1994M1
1994M5
1994M9
1995M1
1995M5
1995M9
1996M1
1996M5
1996M9
1997M1
1997M5
1997M9
nillion of peso
Source: own calculations based on data provided by A. Ramos from the IMF, and on the IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-30-320.jpg)
![32
Marek D¹browski (ed.)
Figure 2-6. Stock market indicators, 1992–1998
30%
25%
20%
15%
10%
5%
0%
Dec
92
Dec
93
Dec
94
Dec
95
tional standards. The broad money, loans, and the stock
market capitalization (see Table 2-2 and Figure 2-6)
expressed as a fraction of GDP are relatively low
[Caballero, 2000].
Stock market capitalization reached over 50% of
GDP in Chile, more than 25% in Mexico, around 40% of
GDP in Spain and Portugal, and about 100% in the U.S.
in 1997. The figure in Argentina was well below 20% of
GDP during the period of 1993–1996. Broad money
monetization, which stayed during the years 1993–1996
at around 19% of GDP in Argentina, had on average val-ues
around 25% of GDP in Brazil and Mexico. Leaving
aside well developed, leading markets, Argentine per-formed
poorly even when compared to other countries
in the region.
The consequences of the weak financial links and the
sub-development of the domestic market became visible
during the 1995 crisis. When the country found itself near
the limit of access to international markets, this hampered
the swift allocation of resources.
2.2.11. Private and Public Sector
Prior to the crisis, the enterprise sector was almost
entirely private, as during 1991–1994 around 90% of all
state-owned enterprises were privatized [IMF, 1998]. This
move increased productivity, brought significant gains in
800
700
600
500
400
300
200
100
the reallocation of resources as the public sector shrank,
but also boosted unemployment.
Despite a large and expanding private sector, there were
considerable differences in the growth prospects between
larger enterprises and agricultural producers together with
small industrial enterprises located in the countryside. Diffi-cult
access to bank credit may serve as an example.
Although the size of the private sector has became more
and more important since the beginning of the 1990s, one
of the problems of the post-crisis period was the lack of
recovery of private sector credit. It is claimed that its
growth has been repressed by huge government borrow-ings
from the domestic market, which took place during
1995. As the government turned to domestic banks for
financing its monetary interventions during the Tequila crisis
(Figure 2-7) while facing external constraints, the private
sector credits have been crowded out [Caballero, 2000].
The fast consolidation process in the banking system also
enhanced this trend [5].
2.3. Political Situation and Management
of the Crisis
It is believed that the run on deposits of March 1995
was also caused by the bad perception of the current polit-
[5] As many of the local branches of the wholesale and cooperative banks disappeared after 1995, the information concerning their clients' credit-worthiness
was not available (there was no countrywide credit rating system). There is evidence that the surviving banks were unwilling to "screen"
their potential clients and did not want to lend to unknown borrowers from the countryside. Larger share of resources was then used to buy govern-ment
CASE Reports No. 39
bonds and improving liquidity position [Catao, 1997].
Dec
96
Dec
97
Dec
98
0
stock market capitalization
as % of GDP
MERVAL
(stock exchange intex
for prime companies)
Source: National Securities Commission](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-31-320.jpg)
![33
The Episodes of Currency Crisis in Latin...
Figure 2-7. Net public borrowing from domestic banks relative to private sector credit
35%
30%
25%
20%
15%
10%
5%
0%
1994Q1
1994Q2
1994Q3
1994Q4
1995Q1
1995Q2
1995Q3
1994Q4
1996Q1
1996Q2
1996Q3
1996Q4
1997Q1
1997Q2
1997Q3
1997Q4
1998Q1
1998Q2
1998Q3
1998Q4
ical and economic policy related situation. There were
uncertainties concerning short-term fiscal policy,
enhanced by the incoming presidential elections (on the
May 14th). As the election system changed, the incumbent
president needed over 50% of votes to avoid the second
round. There was no IMF program in place at this time
[D'Amato et. al, 1997].
Towards the end of April, the central bank charter has
been changed slightly to allow a more flexible use of redis-counts
in order to help banks. This move was misinter-preted
as a relaxation of the currency board regime.
Moreover, there have been spreading rumors about the
possibility of suspending convertibility of bank deposits
[D'Amato et. al, 1997].
The monetary operations of the BCRA linked with the
announcement of the new fiscal package after an agree-ment
with the IMF, stopped the fall of deposits during the
period between March and May 1995. The IMF, as well as
other international institutions promised a significant
amount of financing [6] [D'Amato et. al, 1997]. And
indeed, there has been net inflow of $1.9 billion until the
end of the year.
The dynamics of the recovery of deposits after May
1995 varied according to their types. First of all, this
recovery can be attributable mainly to the dollar deposits,
as the time was needed to restore confidence in the
domestic currency. Secondly, the quickest response came
from the deposits of the large national public banks, and
then from the foreign banks. These groups of banks were
the first to start regaining their deposits. The cooperative
banks and interior banks still suffered the largest fall in
deposits in mid-1995.
The immediate response of the monetary authorities
to the crisis situation was primarily directed towards
improving the liquidity of the banking system. The
authorities lowered reserve requirements and provided
troubled banks with fresh credits (thus depleting the
reserves of the BCRA in April-May 1995) through swaps
and rediscounts of prolonged maturity. This was allowed
due to the already mentioned modification of the central
bank charter. The BCRA acknowledged rediscounts
beyond 30-day window in case of systemic liquidity prob-lems
[IMF, 1998]. Two Fiduciary Funds were created: one
to facilitate mergers and acquisitions within the private
banking sector, and the other to foster the privatization
of both provincial banks and firms [Burdisso et. al, 1998].
The temporary safety net redistributing the liquidity
within the system and controlled by the largest national
banks was created, as well as a privately managed
deposit insurance scheme. This deposit insurance system
was founded with compulsory contributions of financial
institutions as a surcharge on deposits. The scheme has
CASE Reports No. 39
Net claims on government in % of private sector credit
Source: own calculations based on IFS data
[6] On April 6, 1995, the IMF approved the fourth year extension of the extended Fund facility (EFF) for Argentina. The three-year EFF, approved
initially in 1992, was supposed to support Argentina's medium-term economic and financial program. With the extended program, in April 1995, about
US$1.6 million was immediately available to Argentina, and the rest (US$1.2 million) have been disbursed in three quarterly installments. One year later,
the IMF approved a stand-by credit for Argentina of about US$1 million over the next 21 months, in support of the government 1996–1997 econom-ic
and financial program [IMF, 1995 and 1996].](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-32-320.jpg)
![34
Marek D¹browski (ed.)
an upper limit per depositor in order to hamper moral
hazard [IMF, 1998].
As the confidence in the banking system was restored,
deposits kept mounting, and the interest rate spreads
started to decline, the government commenced the intro-duction
of new prudential measures. Generally, these
measures were aimed at further raising the liquidity of
financial institutions, capital to assets ratios, addressed still
existing information asymmetries in the credit market,
and improved the existing payment system [IMF, 1998].
In 1995, the reserve requirements were still being
replaced by liquidity requirements, with rates depending
of the residual times of maturity [7] [BCRA, 2000]. These
requirements have been gradually tightened over time by
increasing their rates [8] and extending their applicability
to other types of bank liabilities. The purpose of this move
was to improve the public perception of individual banks'
liquidity position and limit imprudent lending policies as
there is evidence that the public discriminated against
"good" and "bad" banks on the basis of their perceived liq-uidity
position during the Tequila crisis.
Between 1996 and 1997, the authorities further
strengthened banks' capacity to withstand liquidity short-ages
by the creation of a contingent repo facility between
the BCRA and a group of 13 major international banks.
This agreement allows to swap a collateral – Argentine
government securities owned by the central bank or by
domestic financial institutions – for up to $ 7.3 billion in
cash [IMF, 1998]. Capital to assets ratio was further
increased by the incorporation of a new weighting system
that takes into account market risk factors, as well as by
the increased role of the regulator of banks. There were
also steps towards the improvements in the information
about debtors available to financial institutions (addressing
adverse selection problem), as well as about individual
banks. There were also significant improvements in the
functioning of the payment system, which today consists
of a real-time gross settlement scheme and three auto-mated
clearinghouses [Pou, 2000].
It is claimed that the "second set" of prudential measures
significantly improved the liquidity of the system, as there
was no system-large run on banks during the Brazilian crisis.
2.4. Post-Crisis Developments
Domestic financial markets have grown visibly since
1994. Some indicators, such as monetization or stock mar-ket
capitalization, show gradual improvement. Neverthe-less,
the rise of private sector credit has been constrained
for a long time. Credit to the private sector has basically
did not recover until only very recently.
The banking system increased significantly, and became
generally healthier. Regulations implemented after the
1995 reduced its exposure to external shocks. As a result
of consolidation within private banks, privatizations and
closures, the number of banks declined to 119 in 1999.
Several smaller, provincial banks, which were not trans-parent
and suffered during the 1995 run, have been priva-tized.
Non-performing assets in the banking sector
decreased significantly. They accounted for 8% of total
assets in private banks and for 13% in public banks in
1997.
There is a need for further fiscal adjustment. The sup-ply-
side rigidities should be addressed, so that external
shocks will not affect the real economy as quickly as in
1995.
The public sector is still running large deficits which is
a problem. It is argued that around 1% of GDP of these
deficits per year can be attributed to the pension reform,
effective since 1996, and thus should improve future effi-ciency.
Nevertheless, as the external conditions hardened
again in 1998–1999, public sector deficits went up to over
4% of Argentine's GDP.
The country risk still remains high and was increasing
during the Asian, Russian, and Brazilian crises. The ratio of
broad money to GDP – although growing over time – is
low by international standards. Similar to the Argentine
country risk, also interest rates were rising in response to
the recent currency crises. Although the increases, which
took place in 1998 and 1999, were significantly lower than
during 1995, the rates stayed high for a long period, and
finally soared in 2000.
The high interest rates in the period of 1999–2000
reflect significant decreases in consumer and business con-fidence,
and the progressive hardening of the borrowing
conditions on international financial markets [9] (and
hence the suppressed access to foreign borrowing for
Argentine investors). The slow recovery from recession
affecting Argentine economy since mid-1998, has also
been attributable to the impact of fiscal tightening on
domestic demand, the political uncertainties (new govern-ment
taking office in 1999) reflecting doubts about the
course of economic policy, and the downturn of trading
partner demand.
Argentine GDP fell by 3.4% in 1999, and according to
the preliminary data, by 0.2% in 2000 [IMF, 2001]. There
has been consumer price deflation, and the significant fall in
CASE Reports No. 39
[7] Higher for shorter residual time of maturity.
[8] To reach 20% of most banking liabilities in 1998; the rate substantially higher than in other countries of the region.
[9] High U.S. interest rates, and reduced access to international financial markets for emerging economies in general.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-33-320.jpg)
![35
The Episodes of Currency Crisis in Latin...
export volume. Trade balance improved in 2000, but main-ly
as a result of suppressed imports. Domestic investment
fell to 16% of GDP, and domestic savings – to less than
13% of domestic production. External public debt reached
32% of GDP in 2000, and public finances deteriorated. To
ease the government financing constraint, the authorities
have secured a financial support package of about US$39
billion, including an augmented stand-by agreement with
the IMF, credits from the Inter-American Development
Bank and the World Bank, and a loan from Spain. The
authorities plan to enforce a program aimed at promoting
private and public sector investment, ensuring fiscal sus-tainability,
and reduce the public debt burden in the medi-um
term. GDP is projected to grow above 2% in 2001.
On the positive side, there was no full-scale run on
banks during the recent episodes of financial turbulence in
the emerging markets. And, in aggregate, there was neither
a loss of deposits, nor a loss of credits when looking at the
annual data – although their growth has slowed markedly.
Even though these recent crises precipitated the recession,
there was neither international, nor domestic capital flight
from the banking system. The result of a financial stress test
indicates that Argentinean banks appear to be well insulat-ed
from the interest rate risk. In April 2000, they were
found able to withstand a flight of deposits of an amount
twice as large as in 1995 [IMF, 2000].
Coming back to the Tequila crisis, the main factors
responsible for spreading out the 1995 crisis seem to be,
in their majority, addressed. However, there is still a need
to further deepen financial markets, as well as to broaden
the role of the private sector. As the developments that
took place in 1999–2000 showed, fiscal-side reforms need
to be addressed, if the country wants to emerge on a sus-tained
growth path while maintaining the currency board
regime. Similarly, the labor market reforms should be
implemented to allow for more flexibility.
2.5. Conclusions
There is evidence that the 1995 Argentine financial crisis
coexisted with weak credibility of the currency board and
risk-averse investors. A calibration of a model of contagious
currency crisis to Argentine data done by Choueiri (1999)
shows that if we are to believe that investors were sufficient-ly
risk-averse, this financial turmoil could be attributed to the
Tequila effect from the Mexican devaluation alone. Moreover,
the economic fundamentals of Argentine economy did not
matter in triggering the crisis [Choueiri, 1999].
Although there were speculative attacks on the peso,
Argentina did not devalue the currency. Instead, the mon-etary
authorities depleted its exchange reserves, and real
interest rates rose. This situation matches the definition of
a currency crisis given by Eichengreen, Rose and Wyplosz
(1994). The currency crisis occurred, although there was
no change in the nominal exchange rate. As a result of cur-rency
pressures, the bank runs followed.
The characteristics of the 1995 Argentine crisis are
perfectly captured by the second-generation theoretical
models of the currency crises. Speculations about the pos-sible
devaluation of the Argentine peso increased the
probability of this devaluation. The authorities were then
facing a choice between short-term and long-term eco-nomic
goals. The government had reasons both to aban-don
the peg and to defend it, but the latter only at a cer-tain
cost. Finally, the currency pressure was not directly
related to economic fundamentals.
The interesting question is what would happen if the
monetary authorities decided to devalue the peso. First of
all, it should be noted that Argentina, with its high exter-nal
debt stock, was financially fragile. Large real deprecia-tion
would have surely risen country risk premium. From
this point of view, if the devaluation had taken place in
1995, it would have had destabilizing effects, to the finan-cial
system in particular.
On the other hand, as the real exchange rate was depre-ciating
gradually following the crisis, real interest rates
stayed high for long, thus adversely affecting investment and
output. It is then tempting to assume that if a country had
decided to devalue, the current output and employment
would not have fall as much as they did in 1995. However,
is it a possible outcome? Firstly, it should be remembered
that when the crisis hit, it aggravated the already existing
problems in the banking sector, and in the whole economy
in general. As the experience of some of the East Asian
countries indicate, devaluation does not have to be an
immediate remedy, when the real-economy problems lie in
the lack of transparency of the banking and enterprise sec-tor,
even when domestic financial market is relatively well
developed. There is also a problem of an initial overshoot-ing.
And output may fall as well in such situation. Secondly,
Argentine GDP started to rise during the year following the
crisis, and its growth rate has been impressive. It should be
remembered, that the Argentine economy has been grow-ing,
on average, by 4.7% during the period 1991–1999
notwithstanding two recessions [Pou, 2000].
To sum up, it is hard to believe that the Argentine eco-nomic
performance would have been better if the authori-ties
devalued the currency in 1995. The credibility of the
anti-inflationary policy would have been destroyed and the
country risk would have been much higher, indicating high-er
vulnerability to subsequent external shocks. Besides,
there were positive changes in the Argentine banking sys-tem
brought about by the crisis and by the decision to con-tinue
the existing exchange rate policy. It is also doubtful,
whether the real economy response would be much differ-ent
than it was in 1995.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-34-320.jpg)
![39
The Episodes of Currency Crisis in Latin...
Part III.
The 1997 Currency Crisis in Thailand
by Ma³gorzata Antczak
3.1. Introduction
The financial turmoil that erupted in Thailand in 1997
did not fit into any group of models of financial crises exist-ing
in the economic literature at that time. It is just recent-ly
when researches tried do develop the so-called third
generation models. The Thai experience is an example
which confirms that financial crises occur when macroeco-nomic
as well as microeconomic fundamentals experience
vulnerabilities.
This paper seeks to explore the country-specific factors
lying behind the Thai financial crisis. It provides analysis of
macroeconomic and microeconomic roots of the crisis. It
shows that while macroeconomic imbalances played an
important role, the close relationship among banks, corpo-rations
and the government created additional problems,
which resulted in many bankruptcies and led to a sharp and
unexpected economic downturn. The financial crisis con-tributed
to a sharp contraction in domestic demand and
activity. Also, the paper describes the sequence of the crisis
and its management.
Having relatively strong macroeconomic fundamentals
(excluding a deteriorating current account balance, falling
investments and some foreign exchange reserve indicators)
the Thai authorities did not face any dramatic external
shock as in the second-generation models. In Thailand, as in
all Asian countries there was a boom-bust cycle in segments
of the asset market (stocks, land prices, and real estate)
preceding the currency crisis.
Starting from the late eighties, prudent macroeconomic
policies have supported a period of rapid economic growth
and price stability in Thailand. However, in recent years the
combination of a fixed exchange rate (which was linked to a
basket of other currencies but with a strong dominance of the
U.S. dollar), an increasingly open capital account, and impres-sive
economic growth, attracted short-term capital inflows.
These inflows were often channeled to over-invested sectors
CASE Reports No. 39
due to weak prudential regulations in the banking sector
enhanced by risky investments and poor corporate gover-nance
[1]. The huge amount of short-term investments left
the economy vulnerable to sudden shifts and external shocks.
These began to materialize in 1996 as a sudden drop in
exports led to a high current account deficit. At the same
time, slowing economic activity and a weakening in the
financial position of banks and finance companies led to
debt-servicing difficulties and an increase in non-performing
loans (NPLs). Starting from May 1997, the Thai currency
market was destabilized by a series of currency attacks of
increasing intensity. The Thai authorities attempted to
defend the baht by increasing short-term interest rates and
intervention in the market. As a result, the Bank of Thailand
reserves were depleted, to significant extent, and the baht
depreciated sharply.
From the beginning of the crisis, economic policies
have been progressively strengthened through: suspension
of unlivable finance companies, expenditure cuts in the
central government budget, and an increase in the central
bank interest rates. Probably the most important action
was a change in the Thai exchange rate regime, effective
on July 2, 1997, from the so-called fixed but adjustable peg
to a managed float. Building on these steps, the govern-ment
developed a comprehensive medium-term econom-ic
policy package, which was implemented with the help of
the IMF. It was focused on the stabilization of the curren-cy
and strengthening of the financial system.
3.2. The Way to the Crisis
Simplifying the classification, the economic fundamentals
can be divided into two broad categories: macro and micro-economic.
At the onset of the crisis, macroeconomic fun-damentals
in Thailand remained relatively sound and did not
show many signs of vulnerability.
[1] Recent literature, which emphasizes weak banking and financial sectors as one factor in currency crisis, includes Chau-Lau and Chen (1998),
Chang and Velasco (1998), Krugman (1998), and Marshall (1998).](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-37-320.jpg)
![40
Marek D¹browski (ed.)
CASE Reports No. 39
Table 3-1. Basic macroeconomic indicators for Thailand
In the mid 1980s, Thailand's economy embarked on a
decade of rapid economic growth. From 1981 through
1986 growth had averaged 5.5 percent. But from 1987
through 1995 the growth rate almost doubled, averaging
close to 10 percent per annum. The acceleration of eco-nomic
growth was primarily investment-led and the Thai
economy experienced a significant shift in the composition
of production. Thailand became a more industrial econo-my
while the agricultural share of GDP fell by half from
1980 to 1996. Manufacturing production and non-tradable
sector of construction, finance and real estate offset this.
The expansion of investment provided the counterpart for
these changes. In late 1980s, investment growth rates
exceeded 20 percent per annum, almost doubling the
growth rate of the economy. In the 1990s investment
growth rates were rising more in line with overall GDP
rates of growth and the share of investment stabilized at
the level of 40 percent of GDP.
In the early 1990s, inflation measured by CPI was
under control and stabilized at the level around 5.8 per-cent
per annum. The price stabilization led to a gradual
decline in nominal interest rates. Demand for high-pow-ered
money in Thailand was relatively stable and broad
money monetization was increasing (Table 3-1), amounting
to 86 percent of GDP in 1997. The central budget indicat-ed
a surplus of 2.4 percent of GDP in 1996. The unem-ployment
rate was very low at 1.1 percent during
1993–97. Investment and saving rates were high, averaging
at the level above 30 percent of GDP. The exception to the
favorable economic outlook was a deteriorating current
account deficit, which rose to 7.8 percent of GDP in the
years 1995–96 and was mostly covered by short-term
portfolio investments.
The deficit reflected private-sector demand for foreign
capital. In the 1980s Thailand's priority was large net capital
inflow promotion, through tax and institutional reforms (see
below) while concurrently developing its financial markets
[2]. Large positive interest rate differentials and a pegged
exchange rate supported this policy. This regime provided a
guarantee to short term investors that they can make a
quick exit at little or no cost. Authorities' measures to
attract foreign capital included:
– Elimination of restrictions on foreign investments,
– Elimination of most barriers on foreign ownership of
export oriented industries [3],
– Granting tax incentives to foreign mutual funds and
investments in the stock market
– Creation of closed-end mutual funds,
[2] In 1992, the authorities approved the establishment of the Bangkok International Banking Facility (BIBF), which greatly eased access to foreign
financing and expanded short-term inflows.
[3] Some limitations on foreign ownership were retained in non-export oriented industries and on the maximum foreign ownership of companies
listed on the stock exchange.
1981-1994 1994 1995 1996 1997 1998 1999
Real GDP Growth 8.5 9.9 8.9 5.9 -1.8 -10.0 4.0
CPI Inflation average 3.8 5.1 5.8 5.8 5.9 8.5 5.8
Fiscal Balance to GDP Ratio* -0.6 1.9 3.0 2.4 -0.9 -2.4 -1.1
Private Sector Credit to GDP Ratio 90.9 97.5 100.0 122.5 115.1
Current Account to GDP Ratio 5.3 -5.5 -7.8 -7.8 -2.0 12.7 9.0
Financial Account to GDP Ratio 8.4 13.0 10.5 -11.3 -13.0
Gross National Savings to GDP Ratio 35.5 35.6 33.2 31.9 31.1 30.1
Gross Domestic Investments to GDP
Ratio
41.38 43.3 43.7 33.6 19.0 20.4
Broad Money Monetization**
(in percent)
71.0 72.2 75.4 85.9 99.0
Source: Own calculations on the basis of data from the IMF, The World Bank
* Central budget balance (percentage of fiscal-year GDP)
** Monetization of an economy is defined as a ratio of a measure of money to an annualized value of GDP at current prices.
Table 3-2. Net capital inflow to GDP Ratio (in percent) and nominal interest rate differential in Thailand
1992 1993 1994 1995 1996 1997 1998
Net Capital Inflow 8.5 8.4 8.4 13.0 10.5 -11.3 -13.0
Differential of Interest
Rate 5 5.5 4 4.5 4.8 3.4 -0.4
Source: Own calculations on the basis of data from IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-38-320.jpg)
![41
The Episodes of Currency Crisis in Latin...
– Establishing rules for foreign debenture issues by Thai
companies,
– Reduction of taxes on dividends remitted abroad.
The promotion of capital inflows combined with a rapid-ly
growing economy contributed to very substantial net cap-ital
inflow to Thailand in the range of 9–13 percent of GDP
between 1989 and 1995. Between 1991–1996, net capital
inflows amounted to 85 billion U.S. dollars.
The composition of capital inflows evolved, as a growing
proportion of the net inflows had short-term nature (port-folio
and other investments), reaching 95 percent of the
total in 1995 (Figure 3-1). Net direct investment inflows
played a bigger role at the beginning of 1990s reaching its
peak in 1993 at the level of 15 percent of total inflows.
However, net portfolio inflows became more important in
1994, as a result of the mentioned subsequent reforms of
the Thai stock markets and the large interest rate differen-tial.
The contribution of foreign direct investments to a total
capital inflow stabilized at the level of 5–7 percent in
1994–1996. The continuation of short-term capital inflow
kept the overall balance of payments in surplus helping to
fuel investments and economic growth. Net capital inflow
used to be partially sterilized. In practice, the sterilized
intervention maintained high domestic interest rates and a
large wedge between domestic and international interest
rates. It attracted foreign capital even more.
3.2.1. Macroeconomic Signs of Vulnerability
There were already signs of Thailand's vulnerability
before the crisis. The main macroeconomic indicators sig-naling
a crisis were: the level of official international
reserves, deterioration in investment, high current account
deficit and excessive credit expansion.
Although the official foreign exchange reserves increased
by 183 percent between 1990 and 1996, they were not suf-ficient
to protect against speculative attacks in the context
of an increasing current account deficit and short-term
external debt payments.
Traditionally, three months of imports' coverage is con-sidered
a minimum threshold of official foreign exchange
reserves. The East Asian countries, apart from Korea,
recorded relatively safe reserves to imports ratio in the
1990s (Table 3-3). This relates particularly to Thailand.
CASE Reports No. 39
Although the level of reserves did not indicate the danger of
a currency crisis, the reserves to short-term debt ratio was
much less favorable. In this respect, Thailand represented
one of the weakest records [Jakubiak, 2000], with interna-tional
reserves below the country's short-term debt obliga-tions.
Three months before the crisis, the ratio of reserves
to short-term external debt indicated the 1.1 coverage of
short-term obligations, two months before the crisis it fell
below 1, and in July 1997 it amounted only to 0.7. The value
of this indicator in Thailand showed that reserves did not
exceed official and officially guaranteed short-term debt in
the pre-crisis period.
Another important indicator is the ratio of international
reserves to base money. Thailand recorded relatively safe
levels of backing in the years prior to the crisis, but substan-tially
lower during the last months preceding the crisis. The
ratio of reserves to base money was falling from about 2.5
in July 1996 to 1.5 in the time of crisis in July 1997 what indi-cated
the increasing financial fragility of the economy.
Over-investment and excessive capital accumulation
accompanied the years of rapid growth in Thailand. In 1996,
investment growth slowed, falling to little less than 7 per-cent
compared with an average of more than 10 percent
growth per annum during the previous five years. Private
investment grew by only 3 percent, reflecting signs of excess
capacity and earlier over-investment. In 1997, the overall
investment to GDP ratio declined to 33 percent from 43
percent of GDP in the previous year. The declining invest-ment
contributed to the output contraction during the
onset of the crisis.
An excessive expansion of the non-tradable sector, par-ticularly
in the real estate and construction activities played
a crucial role in the investment break down. In the pre-cri-sis
period all sectors of the economy were growing rapidly.
Private investment in construction grew rapidly during
1990–94 ant it took up to 50 percent of total fixed invest-ments.
The public investment in construction grew at 25
percent on average, twice higher than the private invest-ment
in construction. Much of the office construction in the
commercial sector was built not by professional property
developers, but by companies itself and for their own use.
In the pre-crisis period, overall private credit was grow-ing
rapidly (Figure 3-4). In particular, this related to loans to
the real estate and housing projects carrying out by financial
companies. Many of these loans were turn to non-perform-
Table 3-3. Reserves in months of imports in Asian Countries
1992 1993 1994 1995 1996 1997 1998
Indonesia 3.3 3.0 2.4 2.6 3.1 5.5 5.4
Korea 1.9 2.0 1.7 2.0 2.4 3.1 4.1
Malaysia 3.3 3.4 4.3 3.3 3.3 4.3 3.7
Thailand 4.5 4.1 4.0 4.8 6.8 8.0 6.8
Source: Own calculations on the basis of data from IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-39-320.jpg)
![Table 3-5. Non-performing loans at domestic commercial banks 1995–99 (percent of total loans)
[4] These numbers exclude debt instruments such as bills of exchange and commercial papers, and are calculated using an end-1997 exchange rate
43
The Episodes of Currency Crisis in Latin...
the growth in asset values was not accompanied by equiva-lently
high growth of earnings. The return on assets fell by
roughly one-third from 1994 to 1996, and as a result stock
prices started to go down in the second half of 1996.
The consolidation of companies' ownership in Thailand
was very strong. In the ten largest non-financial private sec-tor
firms, the top three shareholders owned on average as
much as 45 percent of the outstanding shares. The desire of
the owners to retain control of their conglomerates led
them to use debts to finance their expansion. Large capital
account liberalization facilitated this expansion, by increas-ing
the supply of funds to corporations. As a result, by end-
1997 the corporate sector held debts of approximately 153
billion U.S. dollars (more than 150 percent of GDP), where
123 billion U.S. dollars was financed by domestic banking
system, and 30 billion U.S. dollars from abroad [4]. The
result of this debt-financed expansion was an increase of
debt-equity ratio of corporations from 1.5 in 1994 to more
than 2 in 1997.
Several years of strong economic growth – underwrit-ten
by rapid credit expansion and large capital inflows –
exposed underlying structural weakness of the banking sec-tor,
especially under a poor regulatory framework. The
financial sector grew rapidly in the 1990s, driven by expan-sion
of finance companies and the banking sector. The
investment-led growth of the Thai economy was largely
debt financed, which was reflected in the rapid growth of
banks' assets. Simultaneously, softening of licensing require-ments
for finance companies contributed to their expan-sion.
Finance companies tended to focus more on con-sumer
and real estate financing, while banks leaned more
toward investment financing, particularly in the manufactur-ing
sector. Banks recorded high profit and their share prices
boomed in the period of 1993–97. The interest rate spreads
averaged about 6.3 percentage points in this period, which
supported bank profitability with return on assets averaging
1.6 percent in this period.
However, belying this positive picture, indicators of
underlying weakness in the finance sector started to appear.
Both banks and finance companies were heavily exposed to
the property sector but the exposure was most acute in the
case of finance companies. In 1996, investments of finance
companies in real estate and construction amounted to 35
percent of the total credit, while commercial banks invest-ed
around 20 percent of their total credit in real estate and
construction. This was particularly worrisome in light of the
increasing evidence of over-investment in the property sec-tor.
Additionally, substantial share of finance companies'
credit was being channeled into the stock exchange, leading
to a rapid growth of risk.
Already in 1996, finance companies started to exhibit
liquidity problems, illustrated by increasing strains of
accrued interest. Although the level of overall non-per-forming
loans was relatively low (12 percent of total at the
end of 1996), accrued interests in several banks were high-er
than average and growing, suggesting that the true NPLs
were actually higher and increasing. The first clear sign of
trouble occurred in March 1997 when the Bank of Thailand
and the Ministry of Finance announced that ten, as yet
unknown, finance companies would need to raise capital.
In early 1997, more severe liquidity problems emerged
as the economy slowed. Public confidence in finance com-panies
eroded as solvency problems occurred. In May 1997,
the Bank of Thailand suspended 16 insolvent finance com-panies
and announced that its creditors are expected to
bear part of their losses. During the spring of 1997 the
finance sector began to experience a large-scale deposit
withdrawals, which lead to massive and secret liquidity sup-port
from the authorities to 66 finance companies. This sup-port
peaked in August 1997, reaching altogether about 10
billion of U.S. dollars (about 8 percent of 1997 GDP). The
deposit withdrawal represented a flight to quality by house-holds
and businesses moving their savings from finance
companies to large commercial banks.
The banking sector in Thailand also started to show
weaknesses. Bank capital was substantially overstated,
reflecting reliance on collateral of uncertain value. Anecdo-tal
evidence suggest that banking practices focused heavily
on "name" based lending, relying on personal guarantees
and collateral to secure loans. These transactions were
mostly valued not by independent appraisers what had its
picture in bank balance sheets and income. Indeed, while
reported NPLs of banks amounted to 11.6 percent of
assets, this figure largely included loans that had been non-performing
for one year and over, and did not capture the
most recent deterioration in asset quality. Many private
market analysts estimated NPLs to be at least 15 percent of
total banks' loans at that time.
The subsequent deterioration of the corporate balance
sheets and adverse effects of the depreciation led to a rapid
of 1 U.S. dollar = 47 Bahts
CASE Reports No. 39
1995 1996 1997 1998 1999Q2
NPLs 8 10 22 48 51
Source: Thailand: Selected Issues. IMF Staff Country Report No. 00/21](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-41-320.jpg)
![44
Marek D¹browski (ed.)
build-up in non-performing loans and decapitalization
throughout the whole fragile financial system.
3.3. The Crisis
In 1996 economic growth, exports and investment dete-riorated
in the face of an appreciating real exchange rate.
The current account was in deficit, interest rates were high,
and inflation was increasing. Moreover, serious weakness
appeared in the financial system due to exposures to the
property sector and inadequate loan provisioning. High
interest rates to counteract capital outflows aggravated the
solvency and liquidity position of many banks and finance
companies and resulted in intervention by the authorities to
support the financial system.
In early 1997, the baht came under pressure as traders
began to doubt the viability of its peg to the dollar. The Thai
currency was subject to several speculative attacks in the
first-half of 1997 and the central bank intervened actively on
foreign exchange markets and imposed capital controls in
May 1997. As a result, the official international reserves fell
by almost 13 billion U.S. dollars in the first eight months of
1997 (by 38 percent). The fall was continued and the low-est
level was recorded in February 1998 when reserves
reached the 1994 level.
On July 2, 1997, faced with a banking crisis, a run on the
currency, and large reserve losses, the Bank of Thailand
floated the baht. The currency fell 10 percent immediately
and then weakened further. The bath depreciated by an
additional 22 percent against the U.S. dollar during July.
On July 28, 1997 Thailand formally sought IMF assis-tance.
On August 20, 1997 the IMF announced an assistance
package of 4 billion U.S. dollars and established a list of
reforms that the country was obliged to implement [5].
3.3.1. Managing the Crisis. The IMF Intervention
in Thailand
On August 20, 1997, the IMF's Executive Board
approved a 34-month Stand-By Agreement with Thailand,
amounting to 4 billion U.S. dollars (equivalent of SDR 2,900
million or 505 percent of quota). The adjustment program
was aimed at stabilizing the exchange rate and reducing the
current account deficit through control of domestic credit,
and limiting the rise in inflation. Key elements of the policy
package included fiscal policy measures, and financial sector
restructuring, including closure of insolvent financial institu-tions,
consolidation of banks, and non-performing loan man-agement.
The program provided for three reviews to be complet-ed
during the first year (program targets for September
1997, December 1997 and June 1998). Thereafter, the pro-gram
was to be subject to two twice-yearly reviews (pro-gram
targets for end-December 1998, end-June 1999, and
end-December 1999). Upon approval of the program, Thai-land
drew 1.2 billion U.S. dollars from the IMF and received
a further 4 billion U.S. dollars from bilateral and multilateral
sources.
Additional financing was pledged by the World Bank and
the Asian Development Bank (2.7 billion U.S. dollars), which
also provided extensive technical assistance. Financial sup-port
by Japan and other interested countries (10 billion U.S.
dollars) was pledged at a meeting in August, hosted by
Japan. Bilateral financing has been disbursed in parallel with
the purchases from the IMF. Total official financing of the sta-bilization
program amounted to over 17 billion U.S. dollars
(Table 3-6).
In the second half of 1997, the baht continued to depre-ciate
as the contagion in Asia began. While macroeconomic
policies were on track and nominal interest rates were
raised, market confidence was adversely affected by delays
in the implementation of financial sector reforms, and poli-tical
uncertainty.
By the time of the review under the special emergency
procedures (on October 17, 1997), there were also signs
that the slowdown of economic activity would be more pro-nounced
than anticipated. And in fact it was. The further
depreciated exchange rate put pressure on increase in inter-est
rates, and it resulted in a much sharper decline in private
investment and consumption than originally anticipated.
A new government took office in mid-November 1997.
The new economic team headed by Prime Minister Chuan
reconfirmed the commitment to the adjustment program.
To help stabilize the foreign exchange market, the program
Billion U.S. dollars Percent of GDP
IMF 4.0 3.0
Asian Development Bank and World Bank 2.7 2.0
Other 10.5 7.0
Total package 17.2 12.0
Source: "IMF-Supported Programs in Indonesia, Korea, and Thailand. A Preliminary Assessment". IMF, Washington DC 1999
CASE Reports No. 39
Table 3-6. Official financing of stabilization program
[5] A detailed chronology of the crisis in financial sector is given in the Appendix 1](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-42-320.jpg)
![46
Marek D¹browski (ed.)
of eight percent of GDP in 1996 to a surplus of more than
12 percent of GDP in 1998. Due to the fall in Thailand's
terms of trade and falling dollar export prices, export vol-ume
growth exceeded 8 percent per annum in 1997–98.
Conversely, import volumes fell dramatically by more than
40 percent from 1996 to 1998, reflecting the weakness in
domestic demand and the relative price effect of the deval-uation.
Once the full extent of the weaknesses in Thailand's
economy became known, including the underlying problems
in the financial sector and the collapse of Thailand's interna-tional
position, financial market confidence vanished. Thai-land's
pre-crisis problem of persistent and excessive capital
inflows was transformed into one of managing major capital
outflows, with creditors refusing to rollover short-term
debt. As indicated before, investment was falling sharply [6]
in the first quarter of 1997 what influenced output contrac-tion.
In 1998, the recession widened and real GDP declined
by 10 percent. Private consumption also fell markedly. Con-sumer
durables were particularly hard hit, with car sales
falling to around one quarter of their pre-crisis levels. As a
result of high interest rates and the reduced availability of
credit, consumption fell by 13 percent in 1997 and 1998.
Another factor responsible for decline in consumption was
lowering personal incomes. With the general collapse in
domestic demand, unemployment increased and wages
declined.
Since January 1998, the rate of private credit growth
(adjusted after correcting for changes in valuation due to
exchange rate fluctuations) declined steadily (Figure 3-4).
Later the growth of credit continued to decrease and was
even negative at the rate of 10 percent per annum in Janu-ary
1999. In 1999 the situation started to improve. And in
the end of 1999, the credit started to grow at the rate of 5
percent per annum. However, the volume of credit was still
20 percent below the peak reached in late 1997. This sug-gests
two possibilities. First, that credit-intensity of firms has
fallen as firms started to rely increasingly on retained earn-ings
and other sources of non-bank financing or, second,
that there has been a shift in the allocation of credit [7].
After the sharp contraction in 1998, the recovery start-ed
in 1999. Manufacturing production, which had already
bottomed out by the middle of 1998, grew at double-digit
rates through much of 1999, and by September 1999 it had
surpassed its pre-crisis peak. On the demand side, lower
interest rates and improving recovery prospects have stim-ulated
private consumption. This trend was supported by a
temporary VAT reduction, which took effect in early 1999.
In 1999, Thailand's economy reached the growth rate of 4
percent, which was much more modest than in the past. In
2000, the recovery was strengthened and real GDP was
expected to increase by 4–5 percent. If 2000 trends con-tinue,
Thailand will recover pre-crisis levels of output and
consumption per-capita by the end of 2002 (the IMF esti-mates).
The major contributors to growth continued to be
exports (6.3 percentage points) and private consumption
(3.4 percentage points).
Total factor productivity (TFP), which began to decrease
well before the crisis, became negative in 1996, and bot-tomed
out in 1998. However, since 1999 TFP appeared to
start to grow again, helped by structural reforms and cycli-cal
bounce back, and was estimated to grow from 1 to 1.4
percent in 2000 (the IMF estimates).
Exports have done well and have been a key driver of
the recovery. In 1999, they grew by close to 9 percent, and
were set to grow by 6.8 percent in 2000 (IMF estimates).
The U.S. and EU markets contributed to the pick-up in Thai
exports. More recently, the exports to Japan and ASEAN
countries recovered, and accounted for over 30 percent of
total exports in 1999. Nonetheless, there were concerns
regarding the competitive weakness of the Thai industry.
Skill-intensive activities complained of shortages of high level
skilled manpower, and technology-intensive activities
remained largely confined to the final assembly stage of
operations. More recently, technology intensive exports
have increased. Recent data suggest that the growth in
Table 3-7. Contribution to economic growth in the year 2000 (percent)
Growth Contribution to Growth
Real GDP growth (%) 4.5 4.5
Private consumption 6.4 3.4
Public consumption 4.9 0.5
Private investment 11.0 1.2
Public investment 4.5 0.5
Exports 11.0 6.3
Imports 17.0 -7.5
Source: World Bank. Thailand Economic Monitor. June 2000
[6] While investment fell across the board, investment in construction was especially badly hit, its share fell to 35 percent of the total investment
CASE Reports No. 39
from 50 percent before the crisis.
[7] Some firms have suspended servicing their loans, thereby "obtaining credit" by generating NPLs.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-44-320.jpg)
![47
The Episodes of Currency Crisis in Latin...
Table 3-8. Contributions to GDP growth (in percent)
imports has started to slow down. Imports fell by 17 per-cent
between December 1999 and January 2000 (IMF esti-mates).
Given projected growth rates of imports and exports,
the current account balance was expected to generate a
surplus of 7.7 billion U.S. dollars in 2000 (5.3 billion US$ of
the trade balance surplus). On the capital account side,
repayments by the private sector were expected to fall
from 15.5 billion in 1999 to 9.5 billion U.S. dollars in 2000.
This created a cushion to support potential weakness in
portfolio flows and foreign direct investment. Gross official
reserves increased to 34 billion U.S. dollars at the end of
1999. But in the middle of 2000, the country's foreign
reserves stagnated at around 32 billion U.S. dollars [8]. This
level was sufficient to cover more than three times the cash
in circulation at that time (which approximated 9 billion U.S.
dollars and about 400 billion Thai bahts). This level of
reserves was an equivalent to 200 percent of debt maturing
in the next 12 months, and 6 months of imports.
A second key driver of recovery was private consump-tion.
In 1999, total private consumption grew by 3.5 per-cent,
recovering from sharp contraction in 1998 (-12.3 per-cent).
This growth was broadly consistent with a return of
consumer confidence, reflected in an increase in aggregate
disposable income resulting from rising wages and higher
levels of employment. Inflation remained under control in
1999 and did not create a risk for the economy. The gov-ernment
set an inflation ceiling of 3.5 percent for the year
2000, what was possible to reach.
However, employment data show that the recovery is
still fragile. The crisis did not appear to have affected the
trend in a significant way. After the onset of the crisis
employment was expanding gradually but the unemploy-ment
rate was falling at a very slow pace. The February
unemployment rate fell from 5.4 percent in 1999 to 4.8 per-cent
in 2000. While the unemployment rate appeared to be
modest when compared to European countries, Thailand
has no unemployment insurance system and welfare impact
can be severe. However, a review of the adjustments in the
labor market showed that wage reductions among less edu-cated
workers were less severe compared to the educated
workers, suggesting that labor markets protected the less
well off (WB Monitor).
Looking back over the two and half years under the
Fund-supported program, the successful implementation of
macroeconomic policies can be observed. All above-men-tioned
indicators show that the main objectives of the pro-gram
have been met. Over the medium term, the key chal-lenge
will be to sustain economic recovery in the context of
a heavily indebted corporate sector and continued weak-ness
of financial system. The results of corporate debt
restructuring are not satisfactory and this process still has
some way to go. In order to complete reforms in the finan-cial
sector, the speeding up of corporate debt restructuring
process is also necessary.
3.4. Conclusions
The reasons for the Thai financial crisis were almost
exclusively internal [9]. The Thai experience shows that
financial crises can erupt not only when macroeconomic but
also when microeconomic indicators express vulnerabilities.
The Thai crisis can be classified as representing a kind of
third generation model, which theoretical backgrounds is
still questionable [10]. In early 1997, Thailand faced a canon-ical
balance of payments crisis when a structural misbalance
between the deficit in current account and capital and finan-cial
account (sources of financing) occurred. In the second
quarter of 1997 the authorities defended the baht and inter-national
reserves diminished what together with microeco-nomic
problems led to a currency crisis. Once the full
CASE Reports No. 39
Capital
Quality adjusted
labor TFP GDP
1995 6.2 1.0 1.7 8.9
1996 5.4 1.8 -1.8 5.4
1997 3.4 2.7 -7.9 -1.7
1998 1.8 1.3 -13.0 -10.0
1999 1.8 1.4 0.8 4.2
2000 1.9 1.6 1.0 4.5
Medium term 2.0 1.6 1.4 5.0
Source: World Bank. Thailand Economic Monitor. June 2000
[8] Chase, International Fixed Income Today, 13 September 2000.
[9] The important external factor of the crisis was yen/U.S. dollar exchange rate developments.
[10] Antczak (2000)](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-45-320.jpg)
![55
The Episodes of Currency Crises in Latin...
Part IV.
The Malaysian Currency Crisis, 1997-1998
by Marcin Sasin
4.1. Overview
4.1.1. Introduction
Malaysia, a country in Southeast Asia situated in the
Malaysian Peninsula, gained its independence from Britain in
1957. Since then, the main political force in the country –
the multiethnic National Front (Barisan Nasional) – has won
all 10 elections. The key component of the National Front
is the United Malays National Organization led by Mahathir
Mohamad, who has also been a Prime Minister since 1981.
The population of Malaysia, standing at 23 million is 60%
Malay, 30% Chinese and 10% Hindu by origin. The econo-my
and politics of Malaysia operates principally along racial
lines. The Malays have monopolized the country's politics,
they occupy key posts in the administration, military, police,
15
10
5
0
-5
CASE Reports No. 39
constitutional bodies etc. Chinese descents have dominated
the country's economy and exercise control over 40% of all
the nation's economic wealth [1]. The Hindu population is
predominantly visible in the labor force. After some dra-matic
racial tensions in the 1960s, the authorities have
introduced so called New Economic Policy (NEP) – its pri-mary
goal has been "accelerating the process of restructur-ing
Malaysian society to correct economic imbalance so as
to ... eliminate the identification of race with economic
function... and the creation of a Malay commercial and
industrial community... Within two decades, at least 30 per-cent
of the total commercial and industrial activities ...
should have participation by Malays... in terms of ownership
and management" [2].
Uniform political leadership and social consensus grant-ed
Malaysia a stable environment and allowed it to enter the
track of rapid economic growth. The government has been
involved actively in large infrastructure projects and other
Figure 4-1. Malaysia: GDP growth (% p.a.)
-10
1991 1992 1993 1994 1995 1996 1997 1998 1999
Source: IMF, IFS
[1] Foreigners control over 30% - based on ownership of share capital of KLSE listed companies.
[2] At the times of the formulation of this strategy Malays controlled around 10% of national wealth. The strategy proved partly successful, in 1997
the Malays' share increased to around 20%.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-53-320.jpg)
![56
Marek D¹browski (ed.)
CASE Reports No. 39
100
80
60
40
20
public enterprises. Industrialization, export growth and
investment have been promoted. In the late 1980s, it
became clear that purely state-owned enterprises were not
efficient enough for further growth – they have been there-fore
privatized by "corporatization", listing on Kuala Lumpur
Stock Exchange and by selling large parts of government
shares. Foreign direct investment inflows have been suc-cessfully
encouraged – between 1989 and 1995 they aver-aged
around 7% of GDP.
Fast growth was possible thanks to the impressive rate
of investment, one of the highest in the world – in 1997
43% of GDP was invested [3]. The rate of savings was also
remarkable – in 1997 it amounted to 39% of GDP, private
and public saving had more or less equal shares in total sav-ings.
In 40 years time, Malaysia managed to transform itself
from an underdeveloped, third world commodity exporter
to a modern and industrialized country. The share of prima-ry
sector (agriculture, mining, fishing, etc.) in GDP was only
19% in 1997. Secondary sector (manufacturing, construc-tion)
had a share of 40%, while tertiary sector (services)
constituted 41% of GDP.
4.1.2. The Public Sector
In Malaysia, the (non-financial) public sector consists of
the federal government, 13 state governments (+2 federal
territories), 148 local governments as well as statutory bod-ies
and non-financial public enterprises. The federal govern-ment
is endowed with revenue collection power over its
most important sources (corporate, petroleum and person-al
income taxes and custom duties) and with expenditure
responsibilities over strategic domains (state expenditures,
infrastructure, etc.). As a result it acquires over 80% of all
public revenues and spends around 60% of all the expendi-tures.
Local (and state) governments have narrowly defined
responsibilities, i.e. provision of essential civic services to
local communities. It derives financing from agricultural and
urban taxes, non-tax revenues (like asset sales), and trans-fers
from federal government. The size and influence of
rather inelastic local government on fiscal situation is negli-gible
compared to the federal government, which is the only
body responsible of implementing fiscal policy objectives.
Figure 4-2. Malaysia: GDP by sector of origin
0
1960 1965 1970 1980 1990 1995
Agriculture etc. Industry Services
Source: IMF
[3] There are suggestions that official measures of investment rate are likely to be upward biased, because part of investment in Asian economies
is actually a disguised form of consumption.
Figure 4-3. Malaysia: Domestic demand componentes in 1996
Investment
public 12%
Consumption
private 43%
Investments
private 32%
Consumption
public 13%
Source: IMF](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-54-320.jpg)
![57
The Episodes of Currency Crises in Latin...
4
2
0
-2
-4
Generally speaking, the public sector is relatively small.
In 1997, the federal government raised 24% of GDP and
spent 21.4% of GDP. For the broad public sector (general
government), these figures are 28.7% of GDP and 25.3% of
GDP respectively. Direct taxes provide 40% of revenues,
the same amount comes from indirect taxes, remaining 1/5
are derived from other sources. On the expenditure side,
the budget is divided into current (operative) and develop-mental
parts. Total direct development expenditures in
1997 amounted to 25% of the total budget or 6% of GDP.
The importance of public sector in the economy is obscured
by widespread use of off-budget accounting, through which
significant part of government investment is done – espe-cially
in the case of big infrastructure projects. Data related
to such actions is limited. In 1998 off-budget spending
amounted to around 1.1% of GDP, for 1999 this figure was
1.8% of GDP. Government engagement in large firms (usu-ally
previously state-owned) effects in other quasi-bud-getary
activities like tax concessions and exemptions, lend-ing
on favorable terms, etc. with rather unclear conse-quences
for public finance.
Until the mid-1980s, the main purpose of fiscal policy
was the implementation of government development objec-tives.
The government became extensively involved in many
large-scale non-financial public enterprises – infrastructure
CASE Reports No. 39
projects, industrial investments, etc. This resulted in exces-sive
budgetary deficits that eventually became unsustainable
(reaching as much as 16.6% of GDP) and led to the eco-nomic
crisis and recession in 1985. Afterwards, the conduct
of fiscal policy was altered, with emphasis placed on macro-economic
stability and sustainable growth promotion, and
more attention paid to short-term aggregate demand man-agement.
Expected budgetary revenues started to be calcu-lated
realistically what resulted in somehow strange perma-nent
record of revenue and overall balance underestimation
[4]. As a result of the economy overheating from 1992–93,
the authorities adopted a rather conservative fiscal policy in
order to try to slow down and stabilize domestic demand
arising from large capital inflows. As a consequence, from
1993 till 1997, public sector consecutively recorded sizeable
budgetary surpluses.
Thanks to consecutive surpluses, the Malaysian authori-ties
succeeded in containing domestic debt – its volume was
steadily falling in years preceding the crises. In 1996, federal
government debt stood at decent level of 32% [5] and was
held primarily in government securities. The broad public
sector debt was somehow larger and totaled around 50%
of GDP. The external debt of federal government was rela-tively
small at end-1996, and amounted 4.1 billion USD, or
around 4% of GDP.
Figure 4-4. Malaysia: Federal budget fiscal position (% of GDP)
-6
1991 1992 1993 1994 1995 1996 1997 1998 1999
Source: IMF
Table 4-1. Federal Government Debt end-1996 (% of GDP)
Domestic Foreign Total
31.7 4.2 35.9
out of which Gov. securities 25.1 Market loans 2.1
Source: IMF, Malaysian authorities
[4] For 1990 the budget deficit has been overestimated by 1% of GDP. The same figures for 1991–1994 stand at 2%, 4%, 5% and 3% of GDP.
[5] As a memorandum: central government debt amounted to more than 100% of GDP in 1987.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-55-320.jpg)
![60
Marek D¹browski (ed.)
35
30
25
20
15
10
5
was relatively good. The proportion of non-performing
loans in total loans outstanding was 4.1% (end-1996). The
risk weighted capital ratio of 10.6% was also reasonable
comparing to 8% minimum requirement.
The side effect of liberalization was the emergence of a
large number of small and undercapitalized banks. To reme-dy
this problem, the BNM actively encouraged mergers in
the banking system. The banking sector took advantage of
the liberalization and engaged in rapid expansion of lending
activity, which in years preceding the 1997 crisis took a form
of a lending boom. The domestic credit growth averaged
about 25% per annum. As a percentage of GDP, lending
amounted to about 160% – one of the highest intermedia-tion
level in the world.
Lending to the broad property sector, against equity
and other assets, thrived – the share of loans to this sector
in total banking system portfolio (risk exposure) exceeded
30% and was the highest among merchant banks. The col-lateral
valuation was also very high – ranging from 80 to
100% – which in the presence of real estate price inflation
added much risk to this portfolio.
Together with financial sector deregulation, the authori-ties
pursued a policy of capital account liberalization. For the
highly open and emerging Malaysian market, this liberalization
was an important way of facilitating international trade and
providing investment capital. In the years before the crisis, the
capital control regime could be described as liberal. The ring-git
was externally convertible – was allowed as a currency of
trade settlements, which relieved resident importers and
exporters from the need to hedge against exchange risk.
There were relatively few restrictions on the ringgit transac-tions
with nonresidents.
As a consequence, the system of external accounts [6] and
offshore market for the ringgit developed, mainly in Singapore.
There were no restrictions on the source of funds placed in
the external account as well as on the transfer of funds into or
out of the external account. Malaysian banks were allowed to
provide forward cover against the ringgit to nonresidents.
Over-the-counter trading of the Malaysian stocks and bonds
took place in Singapore and Hong Kong. Borrowing abroad by
authorized dealers as well as their foreign exchange lending
activities were only subject to prudential regulations. Foreign
currency borrowing by residents was allowed, provided the
applicant could prove its earning in foreign exchange. Portfolio
capital inflow was unrestricted into all Malaysian financial
instruments. FDI inflows were actively encouraged (e.g.,
through tax exemptions); repatriation of nonresident invest-ment
and profit was completely free.
This policy, together with a stable, growth-oriented envi-ronment
and wide investment opportunities effected in con-stant
inflow of foreign direct investments, which in 1997
amounted to 6.7 billion of USD. On the other hand, a favor-able
MYR-USD interest rate differential and remarkable stabil-ity
of exchange rate could not go unnoticed by the short-term
investors. In response to a surge in capital inflow, that began to
get out of hand in 1994, the authorities decided to introduce
temporary inflow controls on portfolio transactions – these
restrictions were indeed soon lifted. Short-term inflows quick-
CASE Reports No. 39
Figure 4-8. Malaysia: Private sector credit growth in % p.a.
0
1992 1993 1994 1995 1996 1997 1998
total to property for consumption
Source: IMF
[6] External account is defined as a ringgit account maintained with a financial institution in Malaysia, where the funds belong to a nonresident indi-vidual
or corporation.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-58-320.jpg)
![65
The Episodes of Currency Crises in Latin...
of freedom to maneuver. Such situation falls under the
term "vulnerability". When this becomes common knowl-edge,
speculators join capital outflow and sell the domes-tic
currency short. At this point, it is virtually impossible to
end the crisis without paying a high price.
Malaysia was not an exception to the mechanism
described above. However, the extent of the crisis was
moderate compared to Indonesia, Thailand or Korea.
Malaysia managed to avoid the widespread bankruptcies,
bank runs, social unrest, and downward spiral of inflation
and recession. The key questions are: was Malaysia "vul-nerable",
if yes – why, and why matters did not go such a
disastrous way as in other Southeast Asian countries?
4.2.2. Malaysian Vulnerability Analysis
We start the analysis of Malaysian "vulnerability to cur-rency/
financial crises" in the end of 1996 with a review of
the academic research findings related to currency crisis
predictions. The most popular approach is called "early
warning system" and is based on the fact that usually some
macroeconomic indicators (called "leading indicators")
endowed with above-average predictive power exceed
their usual values and, therefore, issue a warning signal on
the possible crisis. The main drawback of this system is
that after each crisis researchers revise the set of leading
indicators and come up with the new ones that seem to
have a better predicting power. Afterwards, a new crisis
unfolds and a new revision takes place. However, there is
a consensus [see, for example, Edison, 2000] that the
most useful indicators are as follows: real exchange rate,
level of foreign reserves, current account balance, the
level of short term debt, domestic credit growth, fiscal and
monetary expansion, short-term capital flows as well as
other, hardly measurable features like the extent of moral
hazard ("the incentive structure of financial system"),
exposure to contagion, etc.
Essentially all the currency crises in the 1990s had a
short-term capital outflow and currency speculation as their
direct cause, and so were with Malaysia. From the point of
view of short-term international investor holding ringgit
assets, the most important parameters were the expected
MYR/USD exchange rate and domestic interest rate.
In order to defend a currency against speculative pres-sures
the central bank needs to have the sufficient inter-national
reserves. Malaysian foreign exchange reserves
stood at 27.7 billion USD at mid-1997.
The first question is whether the debtor had enough
foreign currency to pay the interest and amortization due.
Total external debt service in 1996 amounted to 6.4 billion
USD or 23% of the international reserves, so they were
sufficient to service the debt. In fact, Malaysia had the best
position in the region with respect to this issue. Sometimes
the debt service is presented as a ratio of export – again the
ratio of 8.2% was the best indicator in the region.
The second question is: once all short term creditors
would like to withdraw their funds at one moment, would
reserves be sufficient enough to meet their demand?
Again, the ratio of the total external short-term debt plus
external debt service to foreign reserves stayed at 70%,
the best among ASEAN-4 [7] countries plus Korea.
Figure 4-16. Malaysia: Foreign reserves (mln USD)
35000
30000
25000
20000
15000
10000
5000
0
1995M1
1995M4
1995M7
1995M10
1996M1
1996M4
1996M7
1996M10
1997M1
1997M4
1997M7
1997M10
1998M1
1998M4
1998M7
1998M10
1999M1
1999M4
1999M7
1999M10
Source: IFS
[7] Malaysia, Indonesia, Thailand, Philippines.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-63-320.jpg)
![66
Marek D¹browski (ed.)
However, there was some confusion about the defini-tion
of short-term debt. The above used data, coming
from the BNM, revealed (as mentioned above) that 27%
(around 10 billion out of 38.6 billion USD) of the total
external debt had a short-term character. When we take
into account the data of The Bank of International Settle-ments
whose members controlled 26 billion USD (2/3) of
the total Malaysian debt the picture looks somehow differ-ent
and less comfortable. In the BIS debt sub-sample the
proportion of short-term debt to total debt amounted to
about 50%. Hence, all the above indicators should be
adjusted (almost twice) accordingly.
International investors can also be afraid that in the
case of financial panic they will not get their hard currency
back. From the theoretical point of view, they can feel
secure when central bank liabilities (reserve money) are
covered by international reserves. So the ability of the cen-tral
bank to completely cover its liabilities with foreign
exchange reserves is a good sign for the solvency of the
system. The ratio of reserve money to official reserves at
end-1996 equaled exactly 134% [8]. However, BNM acted
as a lender of last resort to the banking system. So, in the
event of financial panic all liquid money assets could poten-tially
be converted into foreign currency. Hence, the ratio
of M2 to foreign reserves was another leading indicator of
possible distress. In Malaysia this ratio amounted to 480%,
which could be regarded as potentially dangerous but still
it was the best among ASEAN-4 [9]. On the other hand,
the informative content of this indicator is not very clear,
as M2/FX ratio is, to large extent, country specific and
reflects rather the development of domestic banking sys-tem.
Malaysian system in the 1990s was always character-ized
by a high degree of financial intermediation.
The importance of external short-term debt was
already discussed above – Malaysia had actually a very
decent record with respect to that. Almost equally impor-tant
was the share of short time debt in private sector
financing. The overdependence of the Malaysian corporate
sector on the equity market and short-term debt securities
pointed to the underdevelopment of long-term loan mar-ket.
The high share of short-term liabilities meant that the
authorities would restrain as much as possible from inter-est
rate hikes in the event of capital outflow and increased
pressure on the foreign exchange market. Instead, they
would seek to sterilize these outflows by direct interven-tions
on the interbank money market. The BNM policy of
stabilizing the interest rate meant that facing possible
devaluation expectation the central bank would hesitate to
compensate (with high interest rates) investors for the
devaluation risk.
Devaluation expectations could possibly come from
concern about overvaluation of the the real exchange rate.
The exchange rate was assessed to be overvalued about
5% based on CPI-measure [10]. Based on a real-export-unit-
value, the real exchange rate was overvalued by about
20–25%. Still this could reflect a temporary decline in
semiconductors and commodities prices. Although differ-ent
sources proposed different estimates there was a con-sensus
that the Malaysian currency was over its parity in
1996.
The prime source of concern in 1996 was a wide cur-rent
account deficit (around 10% in 1995 and around 5%
in 1996). All through the 1990s, the balance was perma-nently
negative giving rise to anxiety about its sustainability
[11]. The notion of "sustainability" is, however, hard to
define – what is sustainable today can become unsustain-able
tomorrow. The principal issue is CA deficit financing.
In the case of Malaysia, it was covered through inflows
of capital, notably FDI. The inflows of FDI in the 1990s was
just sufficient to cover the current account deficit. But in
the years preceding the crisis, FDI/CA deficit ratio was
rather on the long-term decline, and there were expecta-tions
that this trend might continue.
One way to make "sustainability" operational is to
introduce non-increasing foreign debt to GDP ratio. The
current account is sustainable if it doesn't cause an exces-sive
build-up of foreign debt. By taking an arbitrary 1%
difference between long-run interest rates and long-run
growth rate, Corsetti, et.al. (1998) show that a "sustain-able"
current account in case of Malaysia equals about
2.3% of GDP. In practice, the external debt/GDP ratio
stood almost unchanged since 1993 till the crisis.
Generally, there is nothing wrong with a current
account deficit as long as it reflects a consumption smooth-ing
process. Milesi-Ferretti and Razin (1996), argue that
Malaysian deficit development matched this pattern rather
closely. But the consumption smoothing theory predicts
that at some moment in future there will be a switch into
trade surplus. From a political and economic point of view,
the deficit will be "sustainable" if it can be reverted into
surplus without a crisis or drastic policy change. However,
in 1996–97 the current account imbalance seemed to have
[8] The money(M1)/foreign reserves indicator was 145%.
[9] It is worth to notice, that in November 1994, just before Mexican crisis M2/foreign reserves has been 9.1 in Mexico, and 3.6 in Brazil and
CASE Reports No. 39
Argentina.
[10] This method has, however, many drawbacks. First, inflation is usually closely monitored by authorities and they attempt to contain the increase
in the price of the most weighted components – so a change in inflation do not necessarily reflect the same change in competitivenes. Second, the so-called
Balassa-Samuelson effect should be taken into account – due to faster productivity growth in the tradable goods sector the exchange rate seems
overvalued (in CPI-terms) while it is actually not.
[11] The authorities were perfectly aware of the problem and tried to cool down the economy and reduce CA deficit in 1997 budget.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-64-320.jpg)
![[12] Some examples included: huge dam in remote Borneo, which rationale was questionable, new national airport, costing 9 billion US$ (almost
[13] The issue of capital productivity is closely linked to ongoing and yet unresolved debate about the causes of Asian miracle, namely whether the
fast growth of Asian countries resulted just from abundance of capital and labor force or from productivity growth. The first view was advanced by
Yong in the beginning of the 1990s and popularized by Krugman (1994, 1998). They argue that the total factor productivity in East Asia was significant-ly
lower (sometimes even close to zero) than the rate of GDP growth. There were also studies that contradicted this view. Sarel (1997) found that TFP
in Malaysia in 1978–1996 grew on impressive rate of 2% per year, while Claessens et.al. (1998) remarked that the return-on-assets indicator increased
in Malaysia from 5.5% between 1988–1994 to 6.3% in 1995–1996.
[14] It is worth mentioning that prices of some most weighted consumer goods in the CPI were effectively regulated and controlled in order to
[15] Sarno and Taylor (1999) conduct a formal test and cannot reject the hypothesis that asset prices took divergent (bubble) path in Malaysia before
67
The Episodes of Currency Crises in Latin...
Figure 4-17. Malaysia: FDI/current account deficit (%)
300
250
200
150
100
50
0
1991 1992 1993 1994 1995 1996 1997
Source: Corsetti et. al. (1998)
a structural character. The perspective of its financing was
closely connected to a "sustainability" of capital inflows
which, in turn, was connected to growth sustainability and
investment opportunities.
Malaysia's economy grew at the average rate of 8% per
annum in the 1990s. It was possible thanks to sizeable cap-ital
inflows, and vice versa the capital was attracted by the
expected high rates of growth and investment opportuni-ties.
However, the abundance of capital, high investment
rate and development promotion by the authorities wors-ened
the quality of investment projects. The government
engaged itself in "mega-projects" and massive infrastructure
constructions [12], many of them reflecting political or pro-pagandist
considerations, rather than efficiency justification.
Private sector, also facing increased supply of capital turned
to more risky projects. Indeed, the incremental capital out-put
ratio increased from 3.7 in 1987–92 to 4.8 in 1993–96,
indicating a sharp decline in investment efficiency and prof-itability
[13]. In such a situation it would be overoptimistic
in early 1997 to hold expectations that in short future
Malaysia would sustain 8–9% rates of growth without some
adjustment or structural reforms.
The signs of overheating were evident in end-1996 and,
eventually, optimistic growth expectations was revised in
early 1997 when the data revealed a sharp fall (60%) in
investment from both foreign and domestic sources. The
economic slowdown seemed inevitable, for what Malaysian
growth-oriented companies with ambitious fund rising aspi-rations
were not prepared. The authorities denied the
overheating problem pointing to low inflation. However,
rapidly expanding domestic credit fueled not only consumer
prices [14] but the asset market too.
The asset market development seems to be the main
factor of Malaysian vulnerability. Facing diminishing returns
in corporate sector investment, the banking system
switched from lending to manufacturing (growth in lending
fell from 30% in 1995 to 14% in 1996) to lending for equi-ty
purchases (and growth in loans granted for share pur-chases
rose to 20%, from 4% in 1995). Such loans were
granted mainly by finance companies and merchant banks.
As a result of such a policy and the wide availability of
property loans, asset prices (shares, real estate) increased
rapidly. A surge in the asset market was consistent with
(speculative) overinvestment story described above. The
KLSI gained about 25% in 1996 and a property and equity
related sector index rose over 50% [15].
The BNM made an effort to slow down the domestic
credit expansion by rising reserve requirements and intro-
10% of GDP), etc.
keep "inflation" down.
1997.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-65-320.jpg)
![68
Marek D¹browski (ed.)
ducing controls over consumer lending for cars and hous-es
in October 1995. It was also gradually raising interest
rates. In March 1997 the BNM tried to halt the asset mar-ket
bubble by placing restrictions and ceilings on property
and equity lending [16]. But this intervention probably
came too late.
Recognizing that these measures and BNM attitude
would eventually put an end to the property and stock
boom, investors started to withdraw their funds. This only
reinforced the downward trend on KLSE and on the prop-erty
market, which already started in the beginning of
1997. Few days after the restrictions were announced the
index was 17% lower than its heights month before. In the
first-half of 1997, the capital market witnessed some spec-tacular
failures in fund raising and initial public offerings.
Many companies were forced to suspend their invest-ments.
Still, there were no crisis expectations at the time.
Economists and analysts were only talking about a "slow-down".
The Malaysian economic fundamentals seemed
strong, and the BNM had a good reputation. In mid-May
the ringgit came under speculative pressure but a few days
of high interest rates (overnight rose as high as 18%) was
sufficient to counter this pressure and fend off the specu-lation
with virtually no impact on the exchange rate.
Malaysian securities had a high rating -rating agencies failed
altogether to anticipate the crisis.
4.2.3. Crisis Development
Malaysia encountered serious problems on July 2, 1997
when the Thai baht peg to the USD collapsed. Immediate-ly,
market confidence to Southeast Asian economies was
reassessed – unexpected devaluation of the baht meant
that any country with similar economic structure, compa-rable
state of fundamentals and export structure was like-ly
to give up its exchange rate policy under similar circum-stances.
A prolonged period of fierce speculative pressure
has started. The yield curve inverted dramatically. On July
8, 1997, the BNM was forced to heavily defend the ringgit.
Short-term rates reached the level of 50%. On July 11,
1997, the Philippines gave up supporting its peso. The ring-git
collapse was a matter of days. After the weekend, on
July 14, the BNM abandoned its ringgit peg to the USD.
Ten days after, Prime Minister Mahathir publicly blamed
"rogue speculators" as responsible for the crisis. Later on,
on several occasions, he suggested that currency specula-tion
should be banned. Such statements further under-mined
investors' confidence in Malaysia. On September 4,
the ringgit broke the psychological barrier of 3 MYR/USD
and continued to depreciate. In January 1998 it hit his bot-tom
ever of 4.5 MYR/USD.
The actual direct cause of the crisis was a rapid rever-sal
of short-term capital flows that finally turned into finan-cial
panic. Malaysia, which in 1996 experienced inflows of
over 4 billion USD and expected billions more in 1997, lost
Figure 4-18. Malaysia: Capital flows (bln USD)
10
5
0
-5
-10
1994 1995 1996 1997 1998
total long term short term
Source: IMF
[16] The restrictions introduced limited further loans to 15% (30%) for commercial (merchant) banks and ceilings on existing outstanding stock
of loans, which actually was lower than the proportion of property loans in banks portfolio. That implicitly meant that banks have to contract their lend-ing,
CASE Reports No. 39
cease to roll-over pending credit, so new loans wouldn't actually be granted soon.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-66-320.jpg)
![69
The Episodes of Currency Crises in Latin...
4 billion USD in 1997 and about 5 billion in 1998. The equi-ty
and property asset bubbles burst. KLSI in 1997 lost 50%
of its value, while its property sector subindex lost almost
80%. Vacancy rate in central business district rose from
10% in June 1997 to 17% in June 1998. Prices and rents of
office and residential property fell.
The crisis brought a 7% decline in Malaysian GDP in
1998 (instead of 8% growth as expected). It affected the
Malaysian economy via the following transmission channels.
The stock exchange crash was a shock in a country with
such a heavy reliance of the corporate sector on the capital
market. Many projects had to be halted, companies had to
postpone their investments. There was also a widespread
uncertainty about the direction of the economy that con-tributed
to the investment slowdown. This sharp decline in
investment activity was the major source of a drastic fall in
aggregate demand, which in turn became responsible for a
recession. Private investment demand fell by 58% in 1998.
Rapid depreciation of asset prices brought massive nega-tive
wealth shock. Individuals and the economy as a whole
Table 4-5. Domestic demand collapse components
private
public
Source: BNM
Figure 4-19. Malaysia: Percentage of nonperforming loans in portfolio
40
35
30
25
20
15
10
5
0
1994Q1
1995Q1
1995Q2
1995Q3
1995Q4
1996Q1
1996Q2
1996Q3
1996Q4
1997Q1
1997Q2
1997Q3
1997Q4
1998Q1
1998Q2
1998Q3
1998Q4
1999Q1
Source: IMF
CASE Reports No. 39
postponed their consumption. Private consumption, a second
major factor in aggregate demand decline, contracted by
12%. As explained below, the overall contribution of public
sector to aggregate demand change was also negative.
A sharp ringgit depreciation brought a rapid improve-ment
in the current account balance. From a deficit of 5%
of GDP in 1997, it turned to a 13% surplus in 1998. As
export and service balances held steady, all the current
account adjustment happened by a squeeze in imports
which declined by 18%. The positive net external demand
partially (in half) offset domestic demand contraction, so
the overall aggregate demand fall by 25%.
Exchange rate depreciation triggered inflationary pres-sures
in the economy. From the end of 1997, inflation
started to rise and at its peaks reached about 6% in mid of
1998. Afterwards, however, thanks to the stabilization of
the exchange rate, inflation began to fall and reached pre-crisis
level in the beginning of 1999. Inflation was substan-tially
lower than expected [17]. The consumer price index
revealed only a part of a price increase process – a large
Change from 1997 to 1998 in %
consumption -10.3
investment -57.8
consumption -3.5
investment -10
commercial ban ks finance co mpanies
merchant banks banking system
[17] Malaysian authorities estimated that 1% of exchange rate depreciation would have the impact of 0.176% on inflation. Given around 40%
depreciation of the ringgit, inflation should have risen at least the additional 6–7 percentage points.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-67-320.jpg)
![70
Marek D¹browski (ed.)
part of the impact of ringgit depreciation was not passed
on to consumers but instead absorbed by firms.
As opposed to other countries in Southeast Asia,
Malaysia did not have a large burden of private sector
external debt. Its vulnerability to exchange rate deprecia-tion
was therefore of less concern and did not pose a seri-ous
threat to the economy. Movement in exchange rate
also had little impact on banks portfolio quality because of
BNM strict prudential supervision over foreign currency
borrowing.
The collapse of asset prices caused heavy losses of
usually highly leveraged financial institutions engaged in
stock market gambling, real estate lending, or other col-lateral-
based credit activities. The banking system
became insolvent, the ratio of non-performing loans
soared, and the economy entered a credit crunch. The
main source of concern became merchant banks and
financial companies – exposed the most to asset market
risk.
The banking system portfolio quality deteriorated
slowly during 1997. After a crash of the stock and prop-erty
markets this process accelerated rapidly. The pro-portion
of non-performing loans in the total credit vol-ume
increased from 6% in December 1997 to 24% in
March 1999. The worst was the situation of merchant
banks, which had 37% of their total assets in non-per-forming
status. In 1998, the banking system reported 2.2
billion MYR pre-tax losses, in sharp contrast with 7.6 bil-lion
MYR profit in 1997. Losses belonged only to finan-cial
companies and merchant banks, as commercial
banks managed to maintain positive return on assets
[18]. Despite losses the capital base of the banking sys-tem
increased by 1.2 bln MYR in 1998 but only thanks to
4.6 bln MYR injection of fresh capital provided by the
authorities. As a result of an asset meltdown through
1998, at the end of this year 9 out of total 78 financial
institutions ceased to meet minimal capital criteria and
other basic regulatory requirements. In addition to bank-ing
sector problems, there were also widespread prob-lems
in the shortly indebted corporate sector. Most of
the companies' short-term debt was to domestic banks,
establishing the link between financial sector problems
and corporate sector distress. Some companies went
insolvent or illiquid because their banks were unable to
roll over their short-term debt. Similarly, the lack of cap-ital
and falling asset prices caused many corporate
defaults, increasing the proportion of bad loans in the
financial sector.
4.3. Response to the Crisis
4.3.1. Introduction
Hostile towards international institutions, as well as
those who tolerated currency speculation which suppos-edly
caused the financial crises in sound and solvent
emerging markets, the Malaysian authorities rejected the
help of the International Monetary Fund (together with its
restructuring and reform directives), and decided to cope
with the crisis themselves.
4.3.2. Fiscal Policy Response
The 1998 budget presented in October 1997 was
designed to deal with large current account imbalance and
overheating through further fiscal contraction. The fiscal
surplus of 3% of GDP was planned, up from over 2% of
GDP in 1997. In addition to these measures, the Prime
Minister finally postponed several multibillion construction
and infrastructure projects.
In November-December 1997, it became obvious that
Malaysia was going to experience a significant slowdown
and the authorities announced dramatic policy tightening
(fiscal and monetary as well) to counteract the crisis. Bud-get
spending were cut by 18% (partly to avoid a possible
deficit), regulation of big import was introduced. Projec-tion
of budgetary surplus was revised down to 1.5% of
GDP in anticipation of lower tax revenues. These steps
came far too late as the crisis spread to other parts of the
economy. The authorities realized this in the beginning of
1998 and decided to stimulate the economy. From March
1998, the policy stance was gradually changed to the
expansionary one. In March 1998, the fiscal package of 3
billion MYR (1% of GDP) was announced. At the same
time a drastic revision of federal budged projected the fis-cal
deficit of 2.6% of GDP. In July 1998, the additional fis-cal
package of 7 billion MYR (2.5% of GDP) was declared.
New funds were to be spent on development projects.
Some of previously suspended undertakings were
resumed. The fiscal expansion continued in 1999 with an
expected deficit of about 5% of GDP.
Despite this effort, and despite the fact that due to con-secutive
budget surpluses in the 1990s Malaysia was pre-pared
for a period of fiscal expansion, the federal govern-ment
response failed to provide the expected stimulus.
Actually, the public sector contribution to domestic
demand was negative – public consumption and investment
[18] Although commercial banks didn't engage themselves excessively in equity and broad property credit they also incurred heavy losses because
CASE Reports No. 39
of their capital interdependence with finance companies and merchant banks.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-68-320.jpg)
![Figure 4-20. Malaysia: monetary base and currency in circulation
100
90
80
70
60
50
40
30
20
10
0
1994M1
1994M4
1994M7
1994M10
1995M1
1995M4
1995M7
1995M10
1996M1
1996M4
1996M7
1996M10
1997M1
1997M4
1997M7
1997M10
1998M1
1998M4
1998M7
1998M10
1999M1
1999M4
1999M7
1999M10
[19] Income tax is based on the income received during the preceding year. The direct tax revenues were 8% larger than planned.
[20] Spread emerged because the domestic interest rates were kept artificially low relative to market sentiments. At the same time, ringgit funds
were needed for currency speculation, driving offshore interest rate high. The important source of speculative ringgit funds for nonresidents were offer
side swaps.
71
The Episodes of Currency Crises in Latin...
fell, compared to 1997, by about 3% and 10%, respec-tively.
Actual deficit of the federal government of 1.5% of
GDP occurred to be smaller than planned. Tax revenues
were underestimated [19]. Also expenditures fell 0.5%
GDP short of what was expected – a result of "institution-al
and operational obstacles". Because the authorities relied
on inert infrastructure projects spending as an instrument
for reviving the economy, the fiscal stimulus got delayed
and spread to 1999–2000.
4.3.3. Monetary Policy Response and Capital
Control
There were many phases and policy stance changes in
the central bank's response to the crisis. In July 1997, BNM
engaged in the heavy support of the ringgit, As a conse-quence
of its interventions, short-term interest rates
remained very high for some time and foreign exchange
reserves were sharply falling. After realizing that continuous
pressure on the ringgit and capital outflow was not going to
ease soon and that high interest rate was likely to damage
fragile financial system and after depleting 4.9 billion USD or
20% of its reserves in fruitless market intervention the
BNM gave up its direct support to the currency.
It started to focus on domestic money aggregates and
financial market stabilization, thus opting to sterilize the
capital outflow. As a result, domestic interest rates
returned to their usual levels, the monetary base
remained virtually unchanged (it actually rose slightly), as
well as M1 aggregate – but a spread emerged between
offshore and domestic interest rate encouraging further
capital outflows [20]. In an attempt to interrupt this
process, controls were imposed on banks to limit non-commercial-
related offer-side swap transactions to 2 mil-lion
MYR per foreign customer. Short selling of stocks on
KLSE was also prohibited.
The BNM turned to the domestic market overlending
problem. It established guidelines aimed at reducing annu-al
credit growth to 25% by end-1997, to 20% by end-
March 1998, and to 15% by end-1998, put restrictions on
financial companies consumption lending and on all bank-ing
sector property projects.
Relatively loose monetary policy continued for some
time. The measures taken did not meet their aim – capi-tal
outflows were not stopped, reserves continued to fall,
and the ringgit was permanently depreciating at a very
stable pace. These developments increased the burden of
public and private external debt, increased the price of
essential import items, and most notably, set off inflation-ary
pressures in the economy. The inflation-depreciation
spiral became a real threat.
From November-December 1997 through February
1998, the BNM changed its policy stance to a contrac-
CASE Reports No. 39
Monetary base Currency in circulation (bln MYR)
Source: IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-69-320.jpg)
![72
Marek D¹browski (ed.)
90
80
70
60
50
40
30
20
10
tionary one, increasing several times its intervention inter-est
rate to as high as 11% but at the same time lifting
some lending restriction as domestic banking sector con-dition
deteriorated. The monetary policy remained tight
until August 1998 leading to a sharp contraction in mone-tary
base followed by decrease in M1 aggregate money. In
the meantime, the current account sharply improved and
inflation, which prospects were in the beginning overesti-mated,
was successfully subdued [21]. But the tight mon-etary
policy contributed to further weakening of the
economy while, after a short break, the ringgit was again
started its downward descent. BNM analysts could hardly
see any stable correlation between interest rate and
exchange rate [22]. The offshore-onshore interest rate
spread reached 20%, constraining the effectiveness of
domestic monetary policy and causing capital outflow to
continue. In this environment, the Malaysian authorities
decided to impose severe capital account controls on Sep-tember
1, 1998.
The main objective of capital controls was to regain
monetary independence [23] and to terminate capital out-flows
which was to be achieved by the effective elimina-tion
of offshore market and insulating domestic interest
rates from external developments. Restrictions eliminated
practically all legal channels for the transfer of the ringgit
assets abroad, required the repatriation of the ringgit held
offshore to Malaysia within a month, prohibited the repa-triation
of portfolio capital held by nonresidents for 12
months, and imposed tight limits on the transfer of capital
abroad by residents. Residents were prohibited to grant
ringgit credit to nonresidents. However, payment and
transfers for international trade and FDI was not subject-ed
to restrictions. All import-export transactions had to
be settled in foreign currency. All transactions with
Malaysians assets (stock, securities, etc.) could only be
concluded and registered through authorized institutions
like KLSE. Securities had to be repatriated to Malaysia –
some 170.000 investors had their portfolios totaling 10
billion MYR frozen.
In such an environment the BNM could easily exercise
monetary policy and regain control over foreign exchange
market. The ringgit has been pegged to the dollar at 3.8
MYR/USD and remains stable at this level until now
(November 2000). Monetary policy swung from contrac-tionary
to expansionary and central bank became engaged
in providing liquidity for the domestic market. Rapid cuts
of interest rates from 10 to 6% followed what was wel-comed
by the domestic corporate and financial sector.
CASE Reports No. 39
Figure 4-21. Malaysia: money M1 (bln MYR)
0
1994M1
1994M4
1994M7
1994M10
1995M1
1995M4
1995M7
1995M10
1996M1
1996M4
1996M7
1996M10
1997M1
1997M4
1997M7
1997M10
1998M1
1998M4
1998M7
1998M10
1999M1
1999M4
1999M7
1999M10
Source: IFS
[21] Not only was inflation smaller than expected, unemployment did not become a serious problem, increasing from 2.6% in 1997 to 3.9% in
1998. The main burden of labor market adjustment was absorbed by migrant workers. Right after the emergence of crisis the government canceled
the law giving migrant workers an automatic prolongation of their working-contract after expiration.
[22] Gould and Kamin (2000) remark that monetary tightening may lead to counter-intuitive result of further currency depreciation, because it
threatens to further weaken the banking system and the economy as a whole. The unstable relationship between interest rate and exchange rate can
reflect market confusion about that issue.
[23] As it is well known, the authorities can maintain only two out of three following things: control over exchange rate, freedom of capital move-ment
and monetary policy independence at the same time.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-70-320.jpg)
![[24] In some countries, e.g. USA, large pension funds are prohibited by law to invest in securities below "investment standard" according to rating
73
The Episodes of Currency Crises in Latin...
Figure 4-22. Malaysia: Yearly inflation (%)
7
6
5
4
3
2
1
0
1994M1
1994M4
1994M7
1994M10
1995M1
1995M4
1995M7
1995M10
1996M1
1996M4
1996M7
1996M10
1997M1
1997M4
1997M7
1997M10
1998M1
1998M4
1998M7
1998M10
1999M1
1999M4
1999M7
1999M10
Source: IFS
Malaysian authorities argued that the capital control was
only a temporary measure in the crisis recovery plan but
the international reaction was negative. Malaysian securi-ties
were downgraded by rating agencies and the country
lost a lot confidence among investors.
Malaysia modified and eased capital account restrictions
in February 1999 – partly because of the fear of massive
capital outflow after the 12-month moratorium and because
the domestic and external market situation stabilized, but
also to prove that controls were indeed transitory. 12-
month moratorium period was lifted and replaced with a
system of declining exit tax. The highest levy amounted to
30% and was to be gradually decreased. This step was con-sidered
as an improvement, so in April 1999 Malaysia was
upgraded by rating agencies. Only a small proportion of
invested funds were withdrawn in the period following the
relaxation of capital controls [24]. Authorities continued to
relax monetary policy – interest rate further decreased and
fell below 3% in mid-1999.
There is no consensus whether Malaysian capital restric-tions
have been necessary, neither have they been effective.
The main problem is the difficulty to separate the impact of
these restrictions from the usual market developments.
There are arguments that the vast majority of investors who
wanted to withdraw from Malaysia had already withdrawn
before September 1998, so the restrictions might not have
actually been necessary. On the other hand, as Edison and
Reinhart (2000) point out, the absence of speculative pres-sures
on the ringgit, the exchange rate stability, sharp
decrease in interest rates, reviving economy, steady inflow
of new FDI's, and absence of any parallel or black exchange
market provide the arguments that restrictions has been
accepted and probably effective.
4.3.4. Financial and Corporate Sector
Restructuring
The consolidation of the financial system started in
March 1998 and is continuing. The main components of
Figure 4-23. Malaysia: Unemployment rate (%)
5
4
3
2
1
0
1994 1995 1996 1997 1998
Source: IMF
agencies.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-71-320.jpg)
![74
Marek D¹browski (ed.)
this process are: setting up the institutions in charge of
cleaning up bank's non-performing loans (Danaharta),
injection of capital to undercapitalized banking institutions
(Danamodal), corporate debt restructuring, company
merger program, and tightening of prudential regulations.
Danaharta is a financial institution set up by the
authorities with the mandate to deal with the problem of
non-performing loans, and was expected to push borrow-ers
and banks for fast restructuring. Its staff came mainly
from the private sector. At the first stage, Danaharta
acquired from banks at market value bad loans [25] that
could not be restructured by banks themselves, with a
price averaging about 40% of the book value [26]. The
capital essential for the bailout operation was provided by
the Ministry of Finance (1.5 billion MYR), and came from
the issuance of zero-coupon bonds that had explicit gov-ernment
guarantee. Through March 1999 Danaharta pur-chased
23 billion MYR of non-performing loans, or a quar-ter
of such loans outstanding.
In the second stage, Danaharta started to manage,
restructure or/and execute the loans. In order to shorten
maximally the period of financial clean-up and force banks
to give up bad loans at prevailing low market value, the
company built up a very strong system of incentives.
Banks selling loans to Danaharta have a right to 80% of
any profit that this company makes out of the loan, they
can also amortize for five years the loss resulting from
selling assets under its book value instead of recognizing it
immediately. (In case of undercapitalization caused by
such a transaction banks are eligible for recapitalization
carried out by the Danamodal). Non-interest earning bad
loans are exchanged with government guarantees, zero
risk interest paying bonds. Danaharta has also special and
unusual powers over the borrowers, whose debt has
been acquired. Another body set up especially for restruc-turing
large corporate loans is called the Corporate Debt
Restructuring Agency. Through May 1999, the CDRA was
reviewing 57 applications encompassing 31 billion MYR,
out of which 2.5 billion was restructured.
Danamodal is an institution set up by BNM in order to
provide the additional equity to undercapitalized financial
companies. Its capital is funded by BNM (3 billion MYR)
and by the government zero coupon bonds (8 million
MYR). It intervenes to restore adequate capital ratio,
however, the effectiveness of each investment has to be
assessed by the independent international specialists.
After intervention, Danamodal takes a stake in the institu-tion
proportional to its involvement, and engages in the
management of the company (it appoints at least two,
high rank board members). By mid-1999, Danamodal
injected 6.2 billion MYR to 11 banking institutions.
Another way of restoring adequate capital ratio in the
system is provided by a merger program, which is con-ducted
under the BNM encouragement and supervision.
There is some concern, however, that the activities of
Danamodal mean a hidden form of banking system nation-alization.
On the other hand, the authorities take into
consideration (re-)selling distressed financial institutions
to private sector, mainly to foreign banks.
What concerns prudential regulations, from 1998 the
BNM requires from individual institutions to conduct
monthly tests based on parameters given by the BNM. It
has also issued "Minimal Standards on Risk Management
Practices of Derivatives", "Minimum Audit Standards for
Internal Auditors", and other guidelines for bank man-agers.
In addition, it requires incorporating off-balance
sheet items into loan classifications, and market risks into
the capital adequacy norms. The BNM increased the fre-quency
of on-site inspections, etc. The regulatory frame-works
of capital market have been strengthened as well as
disclosure and transparency rules.
4.4. Conclusions
Generally speaking, Malaysia has avoided all the eco-nomic
and social havoc characteristic of the crises in
neighboring Indonesia, Thailand, and Korea. GDP fell tem-porarily
(-7.5% in 1998) and then rebounded. The
exchange rate depreciated around 50% and remained sta-ble
at the level of 3.8 MYR/USD. There was no social
unrest or widespread poverty problem. Malaysian eco-nomic
fundamentals occurred to be much sounder than
those of its neighbors [27]. The external debt remained
manageable, exposure to exchange rate risk limited,
domestic debt low enough to absorb the costs of econo-my
restructuring, and level of moral hazard in financial
institutions not so high. Financial institutions had a lower
ratio of non-performing loans and higher capital, and
there was a stronger banking culture, with better supervi-sion
and prudential regulation than in other Asian
economies.
In 1999, the Malaysian economy grew by 5.8%, while
in the first quarter of 2000 real GDP increased by 11.9%,
[25] Minimal eligible loan amounted to 5 million of USD.
[26] But excluding one large and extremely faulty loan, the average price jumps to 63% of the face value.
[27] There are some opinions that Malaysia has been hit by the crisis unjustified, or by accident or by contagion only. Other opinions suggest that
even without Asian crisis in 1997 Malaysia would have its own financial crisis in 1998–1999 due to overheating and speculative overinvestment in equi-ty
and asset markets and excessive domestic credit expansion.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-72-320.jpg)
![75
The Episodes of Currency Crises in Latin...
and in the second quarter by 8.8%. FDI continued to flow
in, inflation moderated to 1.5%, trade surplus remained
on the high level of 7.7 billion USD, and BNM's interna-tional
reserves – at the level of 30 billion USD. The over-all
risk-weighted capital ratio of the banking system
amounted to 12.6% with non-performing loan propor-tion
equal to 6.4%. The cost of fiscal stimulus amounted
to 5% of GDP, also less than projected. The exchange
rate remains stable, but on somehow undervalued level
[28], which gives good prospects for current account but
might put some inflationary pressure on the economy.
Asset prices stand at around 60% of their pre-crisis level.
The cost of restructuring of the banking sector totaled to
about 10% of GDP, which was less than expected
Contrary to the macroeconomic "soundness", micro-economic
reforms have been very slow. The consolida-tion
of the financial system has not been completed. Until
August 2000, Danaharta acquired, restructured and dis-posed
a total 31.5 billion MYR of non-performing assets
and CDRA completed 27 restructuring cases, involving
23.6 billion MYR. However, there are allegations, that
some of the "restructuring" that take place is just a reshuf-fling
of debt among subsidiaries in order to avoid insol-vency.
In other cases a debt solution is obstructed by the
political connections of debtors. Although Malaysia is a
leader in restructuring among post-crisis Asian countries,
the failure to clean up the debt overhang can have serious
negative consequences for the country.
The stock market has not yet resumed the role of a
fund-rising institution.
Malaysia is still maintaining a rather restrictive capital
account regime and because of that is probably not going
to fully regain investors' confidence any time soon.
Appendix: Chronology of the Malaysian
1997–1998 Crisis
1997
March 28, Malaysia central bank restricts loans to
property and stocks to head off a crisis.
early-May, Japanese officials, concerned about the
decline of the yen, hinted that they might raise interest
rate. Investors start gradual withdrawal from South East
Asian markets.
mid-May, The BNM defends ringgit with few days of
very high interest rates.
July 2, Thai baht collapses and turmoil begins.
July 8, Malaysian central bank, Bank Negara, has to
intervene aggressively to defend the ringgit. The inter-vention
works for a while (the currency slightly appreci-ates).
July 14, Malaysia central bank abandons the defense of
the ringgit and engages in stabilizing domestic money
market with relatively loose monetary policy.
July 24, Ringgit hits 38-month low of 2.65 MYR/USD.
July 26, Prime Minister Mahathir blames George Soros
and other "rogue speculators" for the attack on the ring-git.
September 4, Ringgit breaks through 3 MYR/USD.
September 20, Mahathir tells the public, that specula-tion
is immoral and should be stopped.
October 1, Mahathir repeats his call for tighter regu-lation,
or a total ban on forex trading. The ringgit falls 4%
in less than 2 hours to a low of 3.4MYR/USD.
October 17, Malaysia tightens budget in effort to stop
the economy from sliding into recession.
December 5, Malaysia finally and radically changes its
policy and imposes tough reforms in order to deal with a
crisis. These include an 18% cut in government spending,
restriction on large-volume import, on bank credit and in
stock market regulations. There were to be "no question
of bailout" for financially ailing companies.
1998
January-February, Several increases in BNM's inter-vention
interest rate were planned to stop the currency
from depreciating and restrict inflationary pressures
March, The severity of the crisis is gradually recog-nized
and the fiscal policy changes to more expansionary.
The fiscal package of 3 billion MYR (1% of GDP) is
announced and a drastic revision of federal budged aims
the deficit of 2.6% of GDP.
July, Another additional fiscal package (7 billion MYR
(2.5% of GDP)) announced in order to stimulate the
economy.
January-August, Despite the austerity fiscal measures
and firm monetary policy the crisis and the capital outflow
continue.
September 1, Malaysia introduces capital controls;
financial investment can be repatriated only after a 1-year
period. Rental and profits from sales can be repatriated.
1999
February 5, Malaysia replaces one year holding period
with exit tax. Repatriation of principal and profits will be
subjected to a maximum levy of 30%.
[28] The estimates of current undervaluation (exchange rate of 3.8 MYR/USD) reach even 19%.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-73-320.jpg)
![78
Marek D¹browski (ed.)
Figure 5-3. Indonesia: Employment by sector
CASE Reports No. 39
Figure 5-2. Indonesia: GDP by sector
Agriculture,
mining, etc.
25%
Manufact.
and costruction
34%
Services
(and other)
41%
Source: IMF
relatives of senior government officials) in which the latter
took a share in profits in exchange for providing protection
and preferential access to contracts, concessions, licenses,
tax exemptions etc. The same system of concessions and
restrictions prevented others from participating in the most
profitable economic activities. The children of the President
were owners and chiefs of banks, oil firms and other strate-gic
companies – corruption and nepotism were widespread.
Indonesia was notoriously ranked in the last position in the
Transparency International rankings [1]. The economy has
been, to large extent, monopolized – from oil extraction to
clove cigarettes – there was no Western standard competi-tion
law. Imports and distribution of essential food items
were monopolized by state agency Bulog; food, electricity
and other everyday products were subsidized.
Nevertheless this system, which derived its legitimacy
from sustained improvement in the standard of living of
the mass Indonesians, has been so far successful in achiev-ing
its goal of fast development. As a result of political sta-bility
and the impressive mobilization of country resources
(saving rates around 30%, investment rates around
30–35% of GDP- in the 1990s), Indonesia experienced
three decades of permanent growth and transformed itself
from an underdeveloped, impoverished, agricultural and
mineral-exporting country to a rapidly urbanizing industri-al
economy. In the years preceding the crisis, growth aver-aged
around 7% per year. Recent reforms reoriented the
economy so as to reduce its dependence on the oil sector,
encourage the creation of competitive non-oil export-ori-ented
industrial infrastructure and expand the role of the
private sector. In early 1997, GDP per capita stood at
around 1140 USD, while total GDP was about 225 bln
USD. The primary sector (agriculture, mining, etc.) had
25% share in total GDP and about 40% of total employ-ment.
Manufacturing and construction constituted 35%
and employed about 20% of the labor force. Tertiary sec-tor
(services etc.) had around a 40% share in both GDP
and employment.
5.1.2. Monetary Policy and the Financial Sector
The central bank of Indonesia is Bank Indonesia (BI). Its
task is to conduct monetary policy to stabilize macroeco-nomic
environment, issue the national currency, the rupiah
(IDR), handle foreign exchange assets and debt servicing, as
well as exercise control over national banking and financial
system. To this end, BI monitors broad money, credit aggre-gates
and reserve money. In addition, the authorities moni-tor
the real value of the rupiah against an undisclosed basket
of currencies. The main instrument in BI actions are open
market operations involving Bank Indonesia papers (intro-duced
in 1984) and commercial bank papers (from 1985), as
well as reserve requirements and foreign exchange inter-ventions.
In the 1990s, BI has succeeded in limiting inflation
only to rather moderate level of 7–11%, however, in a peri-od
leading to the crisis inflation was steadily falling and
reached about 5% in mid-1997.
Foreign reserves, together with reserve money, have
been rising largely due to capital inflows, while domestic
credit accelerated, triggered by financial liberalization. In the
1990s, BI had permanent problems with rapidly growing
[1] In 1995: 41/41 (41 place out of 41 reviewed countries), 1996: 47/54, 1997: 46/51, 1998 80/95.
Agriculture,
mining, etc.
42%
Manufact.
and costruction
20%
Services
(and other)
38%
Source: IMF](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-76-320.jpg)
![79
The Episodes of Currency Crises in Latin...
credit (over 20% a year in the period preceding crisis).
Because of that, BI has conducted a relatively restrictive pol-icy
of high interest rates. Reserve requirements have been
raised in accordance with BI plan to reduce credit growth to
17% in 1997 [2]. Bank lending rates in the 1990s have been
usually 10% or more above inflation. The exchange rate
policy can be described as real exchange rate targeting –
with the rupiah gradually and predictably depreciating with
respect to the USD about 4% year, i.e. from around 1900
10
8
6
4
2
1997M1 1997M2 1997M3 1997M4 1998M1 1998M2 1998M3
CASE Reports No. 39
IDR/USD in 1990 to around 2400 IDR/USD in mid-1997.
Such a policy combined with capital account liberalization
created incentives to borrow abroad to take advantage of
interest rate differentials and prepare the ground for large
capital inflows. BI tried to counteract excessive inflows by
widening the rupiah's trading bands from 2% to 3% around
daily mid-rate in 1995, and further to 5% in June 1996, and
to 8% in September 1996, adding therefore some risk to
foreign exchange market.
Before the crisis, the size of the Indonesian financial sec-tor
amounted to some 60% of GDP [3]. The leading finan-cial
institutions in Indonesia were commercial banks,
accounting for 87% of total assets, out of which the biggest
seven state-owned banks accounted for around half of that
figure, the other half being distributed among around 170
private banks (1995). Other financial institutions like insur-ance
companies, pension funds, stock brokerages or other
financial intermediaries played only a minor part in the sys-tem
and had no impact on the overall picture.
The process of liberalization of the banking system
began in 1983 with interest rates liberalization and the elim-ination
of credit ceilings. But ever since the government
opened up the system to new entrants in 1988–89, the sec-tor
started to thrive. In 1988, reserve requirements were
reduced from 15% to 2%, licenses for new private and
joint-venture banks issued, and state-owned firms were
allowed to put 50% of their funds with private banks. The
following year, the requirement of BI license in long-term
loans granting and offshore loans ceiling was removed. The
number of banks soared from 112 to around 240 as anyone
Figure 5-4. Indonesia: Size of Indonesian banking system, assets of
banks in 1996
Joint 6%
Foreign 4%
Local state 3%
State
commercial 40%
Private
national
47%
Source: IMF
Figure 5-5. Indonesia: Trading volume on forex market (bln USD/day)
0
forward&swap spot
Source: IMF
[2] The plan failed to large extent and eventually credit growth reached 23%.
[3] Data for 1994. Compare Malaysia 100% of GDP and Thailand 110% of GDP.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-77-320.jpg)
![80
Marek D¹browski (ed.)
CASE Reports No. 39
800
700
600
500
400
300
200
100
with access to around 3 mln USD minimum capital could set
up shop. So did bank credit that rose by 350% from 1988
till 1995. In 1994, new private banks overtook state banks
with lending activity. Without a proper supervision frame-work,
and in combination with a severe shortage of trained
and experienced bankers, this quickly led to a problem with
prudent asset management and bad debts. To counteract
this problem, BI introduced a minimum capital adequacy
ratio of 8% and gradually increased minimum capital need-ed
to open a bank to around 30 mln USD. In 1992, state
owned banks were converted into limited liabilities compa-nies.
However, the Ministry of Finance announced in 1994
that it would not permit a state bank to default on its oblig-ations.
This was not the end of banking fragility and problems.
Financial scandals and bank failures were quite frequent,
with government intervening to bail out bankrupt banks
and cover deposits [4]. Noncompliance in lending limits and
off-balance sheet operations were widespread. Strong polit-ical
dependence of state banks, which were used as a source
of cheap capital for government affiliates' enterprises,
explains their permanent inferior performance, in terms of
both return on assets and bad loans.
After liberalization, the 1990s witnessed a rapid devel-opment
of the securities market. In 1988, the market was
opened to foreign investors, in 1994 Bapepam (the state
supervisory agency) implemented new accounting stan-dards,
and in 1995 a new computerized automatic trading
system was introduced allowing for much higher trading vol-ume.
Despite these changes, the securities market never
became a major source of commercial finance. In 1994,
equity market capitalization was only of 30% of GDP [5].
Moreover, its importance is still overestimated because cap-italization
includes shares that have never been sold – 70%
of the shares are held by companies' founders [6], while the
Figure 5-7. Indonesia: Equity market size in % of GDP
35
30
25
20
15
10
5
0
1989 1990 1991 1992 1993 1994
Source: IMF
Figure 5-6. Jakarta Composite Index
0
1992M1
1992M5
1992M9
1993M1
1993M5
1993M9
1994M1
1994M5
1994M9
1995M1
1995M5
1995M9
1996M1
1996M5
1996M9
1997M1
1997M5
1997M9
1998M1
1998M5
1998M9
1999M1
1999M5
1999M9
Source: Bloomberg
[4] Examples include looting of Bank Bapindo, or Bank Lippo scandal.
[5] Compared to 280% in Malaysia and 95% in Thailand.
[6] Indonesia has the highest ownership concentration in corporate sector among Southeast Asian countries. As it has been stated above, main
Indonesian companies are characterized by the close connections with the authorities, strong family ties, and are monopolized by the small elite of Chi-nese
businessmen, and senior government officials.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-78-320.jpg)
![81
The Episodes of Currency Crises in Latin...
government holds large stakes in privatized and listed for-mer-
state-owned companies. When adjusted for that the
stock market has provided about 15% of total business
finance. The small size of the market and open access to it
are the reasons why the Indonesian stock exchange was
70% dependent on foreign investors and relatively volatile.
International investors complained about small liquidity,
poor audit standards, doubtful fairness and quality of com-panies'
disclosures as well as notorious insider trading.
Capital account liberalization started with a gradual pro-motion
of foreign direct investment in designated sectors
and with efforts to restructure and modernize the economy
from oil-export dependence [7]. Investment procedures
were simplified, certain restricted fields gradually opened,
equal treatment with domestic investment sequentially
granted, etc. In 1989 so called priority list ("in what fields to
invest") was exchanged by so called negative list consisting
of domains in which foreign investment was restricted. The
number of items on the negative list was decreasing. As a
result, foreign direct investment has been steadily pouring
into Indonesia, especially from 1995 (in 1996/97 accounting
year [8] FDI inflow amounted to 6.5 bln USD). In 1988,
Indonesia accepted the obligation of Article VIII [9] – the
foreign exchange market was developed and the swap
transactions liberalized. The forex market remained rela-tively
shallow with a 10 bln daily turnover in mid-1997. In
1989 and 1996 the authorities liberalized portfolio capital
inflows by eliminating quantitative limits on banks' borrow-ing
from nonresidents, foreigners were permitted to freely
invest in stocks up to 49% of share capital, and the central
bank withdrew from its obligatory intermediation in foreign
exchange transactions.
High interest rates, a stable rupiah exchange rate, a
growth-oriented environment and an expanding stock mar-ket
resulted in constant inflows of foreign portfolio and
short term capital (7 bln USD in acc. year 1995/96 and 6 bln
USD in acc. year 1996/97). As a result, foreign reserves in
Bank Indonesia have been steadily increasing. BI attempted
to sterilize the resulting increase in base money by sales of
central bank papers and through interventions on foreign
exchange market. Capital inflows mainly took the form of
borrowing of commercial banks, which in turn were con-verted
it into local currency and lent to domestic corporate
sector.
5.1.3. The External Sector
As a result of years of trade protection, Indonesia is not a
particularly open economy in Southeast Asian terms – the
average trade [10] is about 25–30% of GDP. The most
important export items are oil and natural gas – their share in
total exports is one-fourth. Indonesia has made an effort to
reduce its dependence on oil, and the export composition
weights towards manufactured goods, such as textiles, wood
products and electrical appliances, and now the share of man-
Figure 5-8. Indonesia: Capital account (bln USD)
15
10
5
0
-5
-10
-15
92/93 93/94 94/95 95/96 96/97 97/98 98/99
[7] In the beginning (mid-1980s) firms were encouraged to export all their production and to invest in remote areas.
[8] In Indonesia the reporting and accounting year starts in April.
[9] Article VIII of the IMF's Articles of Agreement, i.e. general obligations of members.
[10] (import+export)/2/GDP.
CASE Reports No. 39
total net official (=new loans-repayment) FDI short term
Source: IMF](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-79-320.jpg)
![82
Marek D¹browski (ed.)
Figure 5-9. Indonesia: rupiah exchange rate (IDR/USD)
IDR/USD
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
95-11-16
96-01-16
96-03-16
96-05-16
96-07-16
96-09-16
96-11-16
97-01-16
97-03-16
97-05-16
97-07-16
97-09-16
97-11-16
98-01-16
98-03-16
98-05-16
98-07-16
98-09-16
98-11-16
99-01-16
99-03-16
99-05-16
99-07-16
99-09-16
99-11-16
00-01-16
00-03-16
00-05-16
00-07-16
00-09-16
Source: Bloomberg
ufactured goods in export is about 30%. Indonesia is also
dependent on oil/gas sector-related imports (that accounts
for 10% of total import) as well as on intermediate and capi-tal
goods imports. There is one major item to be mentioned
with respect to imports – Indonesia was importing food (rice,
wheat, etc.) for about 2–3 bln USD in years before crisis.
Major trade partners are Singapore, US, Japan and the Euro-pean
Union. To meet the WTO criteria Indonesia was gradu-ally
reducing its import tariffs and they averaged 12% in 1997.
Because the authorities were monitoring and targeting
the real exchange rate, the rupiah overvaluation was not a
major issue in Indonesian external position – in 1990s before
crisis it has been fluctuating around 100% (+/- 3.5%). At
the end of 1996, due to an extremely weak Japanese yen,
the CPI-based real effective exchange rate was somehow
overvalued and stood at 105% [11]. But with a tradition of
sluggish and steady rupiah adjustment, there was no threat
of any drastic exchange rate adjustment.
Figure 5-10. Indonesia: Current account (% of GDP)
20
15
10
5
0
-5
-10
-15
92/93 93/94 94/95 95/96 96/97 97/98 98/99
total trade services+net factor income
Source: IMF
[11] It must be noted that there are well known problems in assessing real exchange rate, so estimates differ depending on the source. There is a
CASE Reports No. 39
consensus, however, that the real exchange rate was in the neighborhood of its parity equilibrium.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-80-320.jpg)
![84
Marek D¹browski (ed.)
record of budget surpluses in the 1990s. Rather conserva-tive
fiscal policy was aimed at reducing domestic demand
stemming from rapid credit expansion, as well as containing
the current account deficit. In the accounting year 1996/97,
the conventionally measured government budget surplus
stood at 1.4% of GDP.
Prudent fiscal policy, combined with high rates of eco-nomic
growth, led to a declining public debt ratio, which in
1997 equaled some 25% of GDP [12].
The government external debt in 1996/97 amounted to
56 bln USD or 23% of GDP – 44% denominated in USD
and 37% in Japanese yen. But the main problem was a pri-vate
sector external debt officially estimated at 57 bln USD.
Given attractive interest rate differentials and a pretty free
capital flow regime, the private sector got heavily indebted
( mostly with short term debts). There is no reliable data on
the extent of the debt or the precise maturity structure [13]
– authorities were obviously interested in underestimating
its share. According to the World Bank, 25% of the debt
was short term, among Bank of International Settlements
members, controlling most of Indonesian debt, this ratio
was 61%.
Total external debt was officially 113 bln (around 50% of
GDP). Other sources estimate this number to be around
130 bln USD, or even more. The debt required 15 bln USD
debt-service payments in 1996/97.
5.2. The Crisis
5.2.1. Introduction
The direct cause of the Indonesian financial crisis was
contagion from its Southeast Asian neighbors. After the Thai
baht was floated, the general Asian market risk was
reassessed sharply and caused adjustments which Indonesia
couldn't endure or counteract. The mechanism of crisis
development was more or less similar to other countries
[14]. A gradual (yet fairly visible) deteriorating macroeco-nomic
situation in the country was making international cap-ital
market anxious about the further profitability of capital
investments in a given country, and generally about the
future prospect of Asian-style capitalism. A long period of
constant growth gave domestic firms a false impression of
stability. A highly competitive and poorly regulated banking
sector engaged itself in more risky projects to take advan-tage
of an investment boom. After the economy overheat-ed,
a slowdown was expected, some short-term capital was
withdrawn, and a number of domestic firms hedged against
possible currency depreciation. The downward pressure on
the currency forced the authorities to choose one of the
alternative solutions.
First, if they had sufficient foreign reserves, they could
defend the currency by hiking interest rates, but it would
lead to a contraction of the economy and could be accept-able
only as a temporary measure (longer debt maturities
are usually not affected). This problem becomes more seri-ous
if a large portion of domestic debt is short-term. A pro-longed
period of high short-term interest rates can force
some cash-short companies to postpone investments or
even default on their obligation. One default can cause
another default, and suddenly the economy falls into illiquid-ity
with serious contractionary consequences. Things get
even worse if banks and firms invest in assets (like stocks,
real estate) as asset prices usually decrease substantially in a
crisis.
Secondly, the authorities could let the exchange rate
depreciate. But if the economy has a large stock of external
debt, especially in the private sector, or if companies bor-row
in foreign currency, having receipts in domestic curren-cy
(and unhedged) can become overwhelmed by the
increase in their debt and get cut from any financing and
eventually fall. There is usually no need for creditors to liq-uidate
pending deposits – when the debt is mainly short
term it is sufficient only to postpone or refuse to roll over.
Illiquidity turns into insolvency – the ratio of bad loans
soars and if the banking system is weak, poorly managed and
undercapitalized, some banks may go bankrupt. If there is
little confidence in the financial system, there could be a run
on other banks and a collapse as well. Authorities then have
to get engaged in an emergency bail-out and massive capital
injection, the monetary base explodes, inflation rises, cur-rency
plummets, depreciation fuels inflation (and vice
verse), and external debt problems increase. Confidence
among foreign investors evaporates. The lack of political will
to stick to harsh anti-crisis measures and social unrest only
makes the problem more serious.
So if there is both a large stock of short-term debt and a
large stock of unhedged external debt, combined with other
ingredients like a shortage of foreign reserves, a weak finan-cial
system or political instability, the authorities might have
no degree of freedom to maneuver and no power to reas-sure
foreign and domestic investors about the safeness of
[12] Compared to 55% in 1987.
[13] Bank Indonesia probably lost track of all private sector foreign borrowings. Other estimates suggest 65 bln USD or even 80 bln. There are
also well understood problems with maturity classification of different loan arrangements – the average maturity of foreign debt was in the neighbor-hood
of 18 months.
[14] Japan was the first example, that the times of highly regulated economies whose growth is based on extensive mobilization of national (or for-eign)
resources might come to an end, and that their growth prospects has to be revised.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-82-320.jpg)
![85
The Episodes of Currency Crises in Latin...
their investment. Such a situation falls under the term "vul-nerability".
When this becomes common knowledge, spec-ulators
join capital outflows and sell the domestic currency
short. At this point, it is virtually impossible to reverse the
crisis without a cost. In this respect, Indonesia typifies the
worst-case scenario.
5.2.2. Indonesia's Vulnerability Analysis
For decades, economists were making an effort to
improve crisis predictability and their work is far from being
accomplished. For some time, the occurrence of a curren-cy
and financial crises is explained in terms of "vulnerability".
When certain conditions are satisfied there is an increased
probability of a crisis, but its timing and actual occurrence
are determined by market sentiment and shifts in expecta-tions.
Usually some macroeconomic indicators (called "leading
crisis indicators") endowed with above average predictive
power exceed their standard values before the crisis and
this is the most frequent way of describing vulnerability. It is
possible to construct a so-called "early warning system"
(see, for example, Edison (2000)) based on the combination
of these indicators that would issue a signal warning against
a possible crisis. However, the performance of such systems
leaves much to be desired [15]. Many researchers come up
with many proposals for leading indicators but the most
popular include: real exchange rate, foreign reserves, cur-rent
account, the level of short term debt (external and
domestic), rapid credit growth, fiscal and monetary expan-sion,
short-term capital flows as well as other, hardly mea-surable
features like the extent of moral hazard ("the incen-tive
structure of financial system"), exposure to contagion
etc.
From the point of view of a domestic investor borrow-ing
externally, or a short-term international investor holding
the rupiah, the most important point is IDR/USD exchange
rate, at which they exchange their rupiah receipts into hard
currency. They thus count on the authorities that the policy
of assuring exchange rate stability won't be altered. For the
defense of the currency against pressures, there is an essen-tial
need of foreign reserves. Indonesian foreign exchange
reserves stood at about 20 billion USD in mid-1997. Total
external debt service to foreign reserves ratio gives an
answer to the question of whether the country can serve its
current obligations. This debt service (interest and amorti-zation)
was about 15 bln USD in 1997, producing a danger-ously
high ratio of around 75%. This means that without
further foreign loans, Indonesia would have had problems
meeting its obligations. Debt can be serviced also by the
export proceeds. But the ratio of external debt service to
export stood at around 30% – the highest among Southeast
Asian countries. Once all short-term creditors would like to
withdraw their funds at one moment, would the reserves
be sufficient enough to meet their demand? Again, the ratio
of total external short-term debt plus external debt service
to foreign reserves, depending on the estimate, ranged
from 210% to 320%. It has been therefore absolutely
essential for Indonesia to have its short-term debt rolled
over, otherwise it would have to default on its obligations.
The country's external finances depended wholly on the
sentiments of short-term foreign creditors and domestic
investors' willingness to hedge foreign debt, as well as their
Figure 5-13. Indonesia: Total reserves minus gold (bln USD)
30
25
20
15
10
5
0
1996M1
1996M3
1996M5
1996M7
1996M9
1996M11
1997M1
1997M3
1997M5
1997M7
1997M9
1997M11
1998M1
1998M3
1998M5
1998M7
1998M9
1998M11
1999M1
1999M3
1999M5
1999M7
1999M9
1999M11
Source: IFS
[15] Actually, Indonesia did fairly well in the rankings of crisis probability.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-83-320.jpg)
![86
Marek D¹browski (ed.)
beliefs about the future exchange rates and the behavior of
other market participants.
A second indicator of financial solvency is the money to
foreign reserves ratio. In the event of financial panic and
irrational domestic asset selling, the central bank should be
able to cover its liabilities (reserve money) – theoretically all
the rupiah (cash or power-money) can be exchanged into
hard currency if the central bank sticks to its exchange rate.
A ratio below 100% is a good sign for the soundness of the
system. In Indonesia it was 100%. In practice, BI acted as a
lender of last resort, so if it was ultimately determined to
support the banking system in the event of financial panic, all
liquid money assets (M1 or even M2 [16]) could potentially
be converted into foreign currency. Therefore the ratio of
monetary aggregates to foreign reserves is another leading
indicator of possible distress [17]. The ratio of money (M1)
to foreign reserves was 135%, but more commonly used
M2/foreign reserves ratio, as a result of thriving banking sec-tor
activities, was about 700%, which can be regarded as
very dangerous [18]. On the other hand, the informative
content of this indicator is not very clear, as M2/FX ratio is
to large extent country specific and reflects rather the
development of domestic banking system.
Although the indicator of the real exchange rate and fis-cal
stance were well below the warning level, the current
account deficit needs to be addressed. Throughout the
1990s, the balance was negative, while standing at 3% of
GDP in 1996 it was not a big problem in the period leading
to the crisis. Nevertheless, when adjustment had been
made for usual oil/gas surplus the deficit would have risen to
more than 5% of GDP. The variability of oil prices and ever
increasing foreign debt service payments might have cast
some doubts on its sustainability.
The notion of sustainability can be implemented by
introducing non-increasing foreign debt to GDP ratio. The
current account is sustainable if it doesn't cause an excessive
build-up of foreign debt. By taking arbitrary 1% difference
between long-run interest rate and long-run growth rate
Corsetti, et.al. (1999) show that a sustainable current
account in case of Indonesia equals about 3.3% of GDP,
which was more or less equal to its actual record. In accor-dance
to that finding, Milesi-Ferretti and Razin (1996) argue
that the Indonesian deficit pattern in the 1990s matched the
consumption smoothing theory rather closely, so it should
be nothing wrong with such a deficit. But the consumption
smoothing theory predicts that at some moment in future
there would be a switch into a surplus. So, from a political-economic
point of view, the deficit is sustainable if it can be
reverted into a surplus according to a optimal development
path, without a crisis or drastic policy change. But in fact, in
1996–97, the current account imbalance seemed rather
structural. FDI inflows, the main source of current account
financing, were never sufficient to cover the deficit, and
Indonesia had to rely on new foreign loans. This led us to the
issue of sustainability of capital inflows, and thus, sustainabil-ity
of economic growth.
Indonesia grew at average annual rate of about 7% in
the first-half of the 1990s. That was possible thanks to
sizeable capital inflow, and, vice versa, the capital was
attracted by anticipated high rates of growth and invest-ment
opportunities. However, the abundance of capital,
high investment rates and rapid credit expansion led to
deterioration of the quality of investment projects. Politi-cally
connected monopolies paid no attention to cost
reduction, a bulk of government investments was designed
under political or propagandist considerations rather than
out of efficiency reasons. The private corporate and bank-ing
sector, facing increased supplies of capital and a tradi-tion
of poor regulation, turned to more risky projects.
Claessens et.al. (1998) remark that the return-on-assets
ratio has decreased in Indonesia from 8% between
1988–1994 to 5.5% in 1995–1996 [19].
However, the main vulnerabilities of Indonesia originat-ed
from a building up of short-term external debt, a shaky
domestic banking system and rising political instability.
Aging President Soeharto showed not only no signs of
retiring but had also been very reluctant to even discuss the
issue of his successor and his idea of how the transition of
power might be accomplished without political and social
tension what in presence of Indonesian political strife and
social/ethnical unrest has been a matter of importance.
There was also a lack of confidence that in an event of a cri-sis
the corruption-ridden government could deal with it
effectively.
[16] If banks allow the depositors to break time deposits.
[17] Compare Obstfeld and Rogoff (1995).
[18] It is worth to notice that in November 1994, just before Mexican crisis M2/foreign reserves were 9.1 in Mexico, and 3.6 in Brazil and Argenti-na.
In Malaysia it was 4.8.
[19] The issue of capital productivity is closely linked to the ongoing and yet unresolved debate about the causes of the Asian miracle, namely
whether the fast growth of Asian countries results just from abundance of capital and labor force or from productivity growth. The first view has been
advanced by Young in the beginning of the 1990s and popularized by Krugman (1994, 1998). They argue that the total factor productivity (TFP) in East
Asia has been significantly lower (sometimes even close to zero) than the rate of GDP growth. There are also studies that contradict this view. Sarel
(1997) finds that TFP growth in Indonesia in 1978–1996 has been quite remarkable and amounted to 1.2% per year. Also the estimates of incremen-tal
capital output ratio (ICOR – a measure of capital productivity) show the decrease form 4.0 in 1987–92 to 3.8 in 1993–96 indicating a some improve-ment
in investment efficiency. But on the other hand (according to World Bank data), ICOR has been rising (weakening efficiency) till 1987 and again
CASE Reports No. 39
from 1994.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-84-320.jpg)
![87
The Episodes of Currency Crises in Latin...
Table 5-1. Noncompliance to prudential rules (in number of institutions). Capital adequacy ratio (Car), Legal lending limits (Lll), Loan-deposit
ratio (Ldr)
What concerns the banking sector, its weakness was
structural. The audit standards, transparency and compli-ance
to prudential principles record was astonishingly poor.
In 1995, half of private banks and 40% of state banks failed
with respect to either capital adequacy ratio, legal lending
limits or loan deposit ratio.
Another issue was that state banks did not necessarily
make loans on a commercial basis and were subject to polit-ical
pressures. This problem was also present among private
banks often maintaining close connections with their bor-rowing
customers [20] which created incentives for risky or
even fraudulent lending to these customers and did not
encourage accurate loan monitoring. Poorly capitalized and
monitored banks competing with other similar small banks
on a segmented market had even more incentives to make
riskier loans, especially if the management expected to
become bailed-out if things go wrong (moral hazard).
This quickly led to the problem of bad loans – about
10% of total loans were classified as non-performing [21].
State banks had an especially bad record with this share at
17%, while private national banks had, on average, 5% of
their loans non-performing. Indonesia also had previous
experience with banking scandals, financial sector bankrupt-cies
and even bank runs. Instead of closing insolvent and
bankrupt banks, the authorities arranged bailouts, encour-aged
mergers and provided other forms of support. They
also announced that no state bank would be left alone to
default on its obligations. Such actions were the reason why
bank managers could have an impression of having (implic-it)
government guarantees. The moral hazard problem
became even more serious in the presence of a poor bank-ruptcy
law and inconsistencies in law enforcement.
The combination of a relatively open financial market,
growing eagerness of global investors to put money into
Asia, high Indonesian interest rates and expanding Indone-sian
corporate sector resulted in large capital inflows –
indeed larger than banks and companies could wisely invest.
The dangerous side effect of this inflow and market circum-stance
was a mismatch in the balance sheets of banks and
corporation. Maturity mismatch resulted from the use of
short-term debt to finance long-term projects, while cur-rency
mismatch emerged from the use of foreign-currency
denominated loans to credit local currency earning projects.
In the pre-crisis period this mismatch has not created
many problems because of exchange rate stability and a tra-dition
of a smooth rollover of short-term debt.
The above-mentioned developments (large interest rate
differentials, open capital market, (false) impression of
exchange rate stability, growth environment, smooth debt
rollover) were responsible for the key component of
Indonesian vulnerability, i.e. the existence of immense
unhedged foreign currency liabilities. In July/August 1997,
one of the major global financial consulting companies sur-veyed
34 Indonesian chief financial officers. 2/3 of them had
more than 40% of their debt in foreign currencies. Half of
this amount was completely unhedged, and most of the rest
CASE Reports No. 39
Total in category Car Lll Ldr
State 7 0 2 1
Private national 166 18 56 11
Local development 27 2 3 0
Foreign and joint 40 1 9 6
Total 240 21 70 18
Source: Montgomery (1997), BI, data for 1995
Table 5-2. Banks' performance: Return on assets (Roa) and Car
96/97 97/98 98/99
ROA CAR ROA CAR ROA CAR
Comm. banks 1.17 12.2 0.38 4.3 -22.6 -24.6
State 0.82 13.9 0.34 2.4 -24.9 -28.4
Private forex 1.13 10.3 -0.47 5.3 -29.2 -18.8
Private nonforex 0.31 9.7 0.97 15.9 -0.35 10.4
Joint 2.49 18 1.54 4.8 -9.88 -7.7
Foreign 4.48 13.8 5.18 12.8 -0.77 12.9
Source: BI
[20] Like the ownership of poorly regulated banks by non-financial companies or political connections between bank managers and borrowing firm
management.
[21] Data for 1995. For end-1996 estimates were about 13%.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-85-320.jpg)
![88
Marek D¹browski (ed.)
Figure 5-14. Indonesia: maturity structure of banks
120
100
60
40
20
*<3m
3m<*<6m
6m<*<1y
had well under half of their debt hedged. Borrowing US dol-lars
was part of life in Indonesia – "it was like going to
McDonald's" [22]. Everyone assumed that the money would
always be available, and took advantage of this situation.
After 30 years of steady economic growth, the corporate
sector didn't seem to fear economic downturn.
Some of the loans were used to finance speculative
investments in such areas as equity purchases and real
estate. Property loans grew at an annual rate of more than
60% during 1992–1995 (compared to 20–25% percent rate
of growth for total credit) and in April 1997 accounted for
19.6% of outstanding bank credits. To restrain the growing
1y<*<3y
3y<*
deposits
exposure of the banking system to this sector BI restricted
in July 1997 commercial banks from extending new loans for
land purchases and property development (except for low-cost
housing). Total credit growth also continued, in spite of
consecutive statutory reserve requirements increases (from
2% to 3% in January 1996 and an announcement of a rise
to 5% in April 1997) and other attempts by BI to contain it.
The growing uncertainty about the future development
on the exchange rate market was reflected by an increase in
the volume of forward and swap rupiah transactions. Their
average daily turnover rose from around 4.5 bln USD in
early 1997 to about 6.2 bln USD in June/ July 1997. The
increase in trading illustrates increased hedging activities
among externally indebted domestic companies. However,
nobody expected that a sharp downturn and such a severe
crisis would erupt.
5.2.3. Crisis Development
Indonesia's troubles began on July 2, 1997 when the Thai
baht peg to the USD collapsed. Immediately, market confi-dence
in Southeast Asian economies was reassessed. Unex-pected
devaluation of the baht meant that any country with
similar economic and export structure and comparable fun-damentals
is likely to give up its exchange rate policy under
similar circumstances. Such an event is called a "wake up
call", or a focal point for coordinating market expectations -
a "sunspot".
A prolonged period of downward pressure started. The
yield curve inverted dramatically. On July 11, the Philippines
gave up supporting its peso, while the Indonesian authorities
CASE Reports No. 39
100
80
60
40
20
0
[22] After The Wall Street Journal 31 XII 1997 quoting one Singapore based Indonesia-investing fund manager.
credit
0
80
Source: BI, IMF
Figure 5-15. Indonesia: bank credit growth (% p.a.)
-20
1993 1994 1995 1996 1997
total property sector
Source: IMF](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-86-320.jpg)
![Figure 5-16. Indonesia: rupiah exchange rate (IDR/USD)
18000
16000
14000
12000
10000
8000
6000
4000
2000
0
97-03-17
97-04-17
97-05-17
97-06-17
97-07-17
97-08-17
97-09-17
97-10-17
97-11-17
97-12-17
98-01-17
98-02-17
98-03-17
98-04-17
98-05-17
98-06-17
98-07-17
98-08-17
98-09-17
98-10-17
98-11-17
98-12-17
99-01-17
99-02-17
99-03-17 Source: Bloomberg
Figure 5-17. Indonesia: Money market rate
90
80
70
60
50
40
30
20
10
0
1996M1
1996M3
1996M5
1996M7
1996M9
1996M11
1997M1
1997M3
1997M5
1997M7
1997M9
1997M11
1998M1
1998M3
1998M5
1998M7
1998M9
1998M11
1999M1
1999M3
1999M5
1999M7
1999M9
1999M11
Source: IFS
[23] To what extent the rupiah collapse has been caused by speculators or by domestic investors suddenly starting to hedge against exchange risk
is a subject of great debate. Indonesian authorities (Minister of Justice in August) claimed that the speculators were guilty and their activities could be
interpreted as subversive criminal actions (there is a death penalty for subversion), while hedge-fund managers and other participants suggest that this
was not the case. Major funds were fully invested in the rupiah and they even supposedly bet on the rupiah rebound at some moment. But it was the
case that many unhedged domestic companies decided to insure in forward market against the rupiah decline. Of course, speculators joined the mar-ket
89
The Episodes of Currency Crises in Latin...
widened the trading bands from 8% to 12% (and inter-vened
when the rate moved outside the band). On July 14,
Malaysia gave up its currency peg. It took one more month
and a 15% depreciation until Bank Indonesia floated the
rupiah and doubled short-term interest rates to over 25%
to support its value [23]. On August 29, BI introduced
restrictions (up to 5 mln USD per customer) on nonresi-dents'
trading in forward currency contract (i.e. supposed
speculation). Despite this, capital outflows continued and by
October 8, the rupiah/dollar exchange rate had already
depreciated cumulatively by 46%. By that day, matters
went so badly with domestic financial and corporate sector
when depreciation seemed to be inevitable.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-87-320.jpg)
![90
Marek D¹browski (ed.)
Figure 5-18. Indonesia: Inflation annualized
90
80
70
60
50
40
30
20
10
0
1996M1
1996M5
1996M9
1997M1
1997M5
1997M9
that Indonesia's government decided to request IMF assis-tance.
The financial and the corporate sector was confronted
with an increase in the rupiah value of their foreign indebt-edness.
Most private companies were able for some time to
cover their foreign exchange losses but they were dramati-cally
running out of cash, unable to refinance their short
term debt and watching it rapidly expanding (banks and for-eign
creditors refused to rollover the existing short term
debt). On the other hand, banks were unable to crack down
on their debtors because of a weak and inefficient bank-ruptcy
law. This, together with tight liquidity and high inter-est
rates, gradually pushed many banks and companies into
technical bankruptcy. At end-October, rating agencies
downgraded the ratings of 10 big Indonesian banks from
neutral to negative, further limiting their borrowing abilities.
The already poor confidence in the national banking system
brought about a gradual build-up of runs on some of the pri-vate
banks, reflecting a "flight to quality", as depositor per-ceived
state banks to be safe and were moving deposits
from presumably troubled private banks. As a result of a cri-sis,
the stock market index immediately lost 30%, and then
a further 10%, before November 1997 because investors
lost confidence. The real estate market collapsed as well.
Office, residential and retail property rent and prices (usual-ly
quoted in USD) fell from 30% to 80% between June
1997 and June 1998.
On October 31, 1997, the IMF unveiled the 23 bln USD
aid package for Indonesia and on November 5 approved a
10 bln USD standby loan facility. Apart from a request for
1998M1
structural reforms and tight monetary and prudent fiscal
policy, the package included the requirement for the closure
of the 16 most insolvent (bankrupt) banks. Authorities,
however, failed to extend appropriate deposit guarantees
and a panic erupted among depositors running the whole
system [24]. A massive and sudden withdrawal of deposits
started, a large number of banks failed to meet their obliga-tions
and had to resort to central bank liquidity support.
One reason for a sudden drop in overall confidence was
that people saw the end of the regime approaching quickly.
They doubted the political capacity of the government to
fulfil its commitments to the IMF. In the beginning of Decem-ber
1997, Soeharto was ordered to retire to bed and disap-peared
from public life for about a month. The implementa-tion
of IMF packages was already delayed or off-track. The
rupiah collapsed badly to almost 6000 IDR/USD. Soeharto
reemerged in public on January 6, 1998, only to unveil the
1998/1999-draft budget that had virtually nothing to do with
the reforms agreed upon with the IMF. At that moment,
confidence in the Indonesian government was lost com-pletely
and on black Thursday, January 8th , the rupiah plum-meted
to 10000 IDR/USD – and later even to 14000. The
market panic across the country in anticipation of food
shortages and overall social unrest and violence started.
Food prices skyrocketed indeed and through 1998
increased by 100%, compared to 70% of total CPI increase.
Financial panic continued, the surge in liquidity provided by
BI to tumbling banks (about 7% of GDP before end-Janu-ary)
far exceeded the real liquidity needs of the economy
and contributed to sharp rise in inflation and put further
[24] Authorities guaranteed only deposits up to about 5000 USD. The guarantees covered 90% of depositors but not even 20% of total deposits.
However, the lack of confidence in banking system was so great that there were hardly any awareness of any deposit guarantees (or belief in such guar-antees)
CASE Reports No. 39
– depositors with deposits less than 5000USD were also running on banks.
1998M5
1998M9
1999M1
1999M5
1999M9
Source: IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-88-320.jpg)
![92
Marek D¹browski (ed.)
inflow of official aid, while the outflow of private capital was
not reversed. Inflation escalated to the level of 80% in 1998
in response to panic food buying, interrupted production,
social violence and increased prices of import commodities.
Unemployment rose from 5% to 28% at end-1998.
On March 10 1998, Soeharto was reelected to a 7th term
in office. On May 4, the government announced sharp price
increase of gasoline and other utilities. Widespread protests
erupted, among them most importantly student-led anti-regime
demonstrations calling for the President's resignation.
The army cracked down on protesters. Embassies and for-eign
companies evacuated non-essential staff. On May 19, stu-dents
started parliamentary compound occupation. The stu-dent
demonstrations seeking political reforms were accom-panied
by rioting, widespread looting, destruction, crime, as
well as religious and ethnic conflicts. Anti-Chinese rioting
directed mainly at shopkeepers in small town resulted in
complete disruption of supply distribution channels and short-age
of basic products. The general erosion of social order
went out of control. Over 1000 dead were reported in the
May riots. The hard-won relative stability of the rupiah was
immediately lost, runs on banks and massive deposit with-drawals
started again, the currency crisis renewed and the
rupiah plunged to over 16000 IDR/USD. It took five months
to bring it back under 10000 IDR/USD. On May 21, urged by
his affiliates, Soeharto resigned.
Changes in key positions in the administration con-tributed
to delays in the implementation of economic
reforms. With reference to that and to harsh economic cir-cumstances,
the IMF rearranged its agreements with
Indonesia towards easier conditionality. The situation start-ed
to stabilize. Food security has been gradually restored
through emergency import and increased food subsidies.
Monetary stability gradually returned around October 1998,
inflationary pressure eased and the rupiah stabilized around
9000 IDR/USD. Price levels also finally stabilized and the
beginning of 1999 saw some deflation. The sluggish process
of financial system and corporate debt restructuring started.
5.3. Response to the Crisis
5.3.1. Introduction
When the crisis unfolded, market confidence in Indonesia
disappeared and domestic conditions started to deteriorate
quickly as economists and market participants came to the
conclusion that the rescheduling of Indonesia's debt and the
rescue of the financial system would require the backing of
western governments and international institutions: "No one
would want to buy Indonesian debt if it was just Indonesian
debt" [25]. The Indonesian authorities also noticed that with-out
additional backing and emergency loans the country would
soon default on its debt. So, the IMF assistance was requested
in October 1997 and the first Letter of Intent, aid package of
23 bln USD and a 10 bln Standby arrangement was announced
in early November. The IMF, not expecting the seriousness of
the crisis, insisted on tough monetary and fiscal policies as well
as economic reforms, including the closure of some bankrupt
banks. These measures, however, only aggravated the bank-run
problem. On the other hand, there was no political will to
implement the reforms. As result, confidence further col-lapsed.
Indonesia and the IMF signed the second agreement in
mid-January 1998. The hardships with implementation of
agreed reforms and civil unrest again threw the program off-track.
The IMF has been several times threatening to postpone
or suspend the aid. From the third letter of intent in April, and
an agreement with the new government after the fall of Soe-harto,
the IMF acknowledged the severe economic conditions
and approved the implementation of less stringent measures.
In July 1998, the Standby agreement was replaced with an
Extended Fund Facility program.
As of June 2000, the international aid commitment to
Indonesia amounted to 50 bln USD: 12 bln from the IMF, 10 bln
from multilateral financial institutions (like World Bank or Asian
Development Bank) and 15 bln from bilateral agreements and
programs (like Japanese Miyazawa plan). About half of this
package, i.e. 22 bln USD, has already been disbursed.
5.3.2. Monetary Policy Response
The government responded to the first run on the rupi-ah
in July/August 1997 with a drastic credit contraction. BI
stopped repurchasing central banks certificates, decreasing
the supply of local currency in an effort to discourage mar-ket
participants from exchanging the rupiah into hard cur-rency,
interest rates rose from around 15% to 30%.
Although official statistics do not show a sharp decline in for-eign
reserves, part of the stock was probably tied down in
forward contracts, so the usable reserves were actually
lower. From September 1997, the reserves started to shrink
much faster. Facing a deteriorating domestic situation, from
September BI gradually cut yields on its commercial papers.
Monetary policy was being carried out in an environ-ment
of high debt-equity ratios and overall financial system
distress, which made a prudent policy of high interest rates
almost impossible to implement. Eventually, the Indone-sian
authorities had to resort to IMF help. One of the main
goals that the IMF program pursued was to restore market
confidence. Based on the presumption that the Indonesian
economy was suffering from structural weaknesses than
CASE Reports No. 39
[25] After The Financial Times 30 I 1998 quoting a regional economist of one major western banks branch in Jakarta.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-90-320.jpg)
![Figure 5-23. Indonesia: rice prices (IDR/kg)
3500
3000
2500
2000
1500
1000
500
0
1995Q1
1995Q2
[26] For example: if there is no budget deficit there is also no temptation to monetize it in a crisis period; on the other hand, increase in public sav-ing
contributes to current account improvement (in 1997 current account deficit has been regarded as a problem).
95
The Episodes of Currency Crises in Latin...
started to move within designated bands, monetary condi-tions
stabilized and BI regained much control over the finan-cial
market. The interest rate decline started since Septem-ber/
October 1998 together with monetary stabilization and
inflation decrease (CPI rise halted and then turned into
slight deflation in 1999).
5.3.3. Fiscal Policy Response
The initial November 1997 IMF plan of fiscal tightening
was expected to restore confidence [26], demonstrate the
authorities' eagerness for reforms and make room for pos-sible
bank restructuring costs. Already in September 1997,
about 80 infrastructure projects (including 13 power plants
and 36 toll roads) were suspended. According to the plan,
wide ranging cuts in public spending and the postponement
of about 35 bln USD in infrastructure projects were to be
implemented in order to reduce the current account deficit
and generally improve the soundness of the economy. The
government budget surplus was planned to amount to 1%
of GDP. In the sphere of structural reforms, the dismantling
of state monopolies, trade liberalization and other similar
measures were envisaged.
However, the authorities ignored their most important
commitments which was revealed in the 1998/1999-budget
proposal in January 1998. The constitutional validity of the
IMF-supported stabilization program was also questioned
on the grounds that it goes against "family values". The dete-riorating
situation forced the government to seek another
agreement with the IMF on January 15, 1998. The macro-economic
assumption of the second program was revised
downward: 0% GDP growth in 1998, 20% inflation and
5000 IDR/USD rate. the fiscal stance was eased to meet a
1% deficit. Calls for structural reforms were reiterated.
Budgetary support, tax and credit privileges to the new air-plane
and national car projects (owed by Soeharto family)
were to be canceled, cartels in cement, paper and plywood
dissolved, domestic agriculture deregulated, import and dis-tribution
restrictions lifted, fuel subsidies gradually
removed, fiscal transparency improved, autonomy for mon-etary
policy granted. Nevertheless, the authorities were still
very reluctant to fulfill these demands, as many of them
were directly targeted at businesses from which govern-ment
officials' relatives and friends profited.
Later in 1998, in accordance with the changing political
and (worsening) economic situation, the program was
revised and included not only accommodation of the shock,
but also some additional fiscal stimulus. Current account
(that actually quickly turned into surplus by itself) and confi-dence
issues ceased to be main problem. The severity of
the recession was not taken adequately into consideration
while designing previous programs. In accordance with the
agreement with the IMF the government raised expendi-tures
that were associated with the social safety net as well
as subsidies of oil-based fuel, electricity, medicine and food-stuff.
Subsidies increased dramatically from 0.3% of GDP in
1996/97 to 3.1% of GDP in 1997/98 and 4.4% of GDP in
1998/99 budget (however less than planned 6.6%). The
state budget was planned and estimated to run into deficit
CASE Reports No. 39
1995Q3
1995Q4
1996Q1
1996Q2
1996Q3
1996Q4
1997Q1
1997Q2
1997Q3
1997Q4
1998Q1
1998Q2
1998Q3
1998Q4
Source: IMF](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-93-320.jpg)
![96
Marek D¹browski (ed.)
Figure 5-24. Indonesia: government subsidies (%GDP)
5
4
3
2
1
0
92/93
93/94
94/95
95/96
96/97
97/98
98/99
Source: IMF
of about 8.5% of GDP in 1998/99. In the end, however, the
government failed to provide a sufficient boost to the econ-omy,
the realized deficit reached only 2.2% due to lower
government expenditures attributed to exchange rate
appreciation and technical constraints in general [27]. As for
deficit financing, foreign borrowing financed 99% of it, while
remaining 1% was financed domestically.
5.3.4. Banking System and Debt Restructuring
The authorities have begun to restructure the banking
system through a mixture of bank closures, mergers and
takeovers. After the closure of 16 banks in November 1997
and the following bank runs, BI guaranteed eventually in Jan-uary
1998 all deposits at domestic banks. At the same time
the establishment of Indonesia Bank Restructuring Agency
(IBRA) was announced. The task of IBRA was to assume
control over troubled private banks, review them for liqui-dation
or recapitalization and manage non-performing loans.
There were initial problems with the operation of that body,
as bank managers failed to change their behavior in accor-dance
with IBRA's recommendation. By April 1998, it
became apparent that forceful ("hard") intervention [28] was
necessary. Efforts were made to recover the liquidity credit
extended to banks by the central bank, so IBRA focused and
finally took over (effectively nationalized) seven private
banks responsible for 75% of all liquidity support and
accounting for 16% of total banking system liabilities.
Another seven very small and completely insolvent banks
were closed. The operations of IBRA were subject to rising
uncertainty among a public unaccustomed to the implica-tions
of bank takeovers. Several "IBRA banks" were being
run on for some time. In September and December 1998,
the authorities announced a comprehensive plan of restruc-turing
of the banking system. Banks were categorized
depending on their capital adequacy ratios (CAR). Banks
with CAR above 4% would be allowed to continue opera-tion.
Banks with CAR below –25% were given one month
to recapitalize, failing which they would be merged or
closed. The rest of the banks were to submit reliable busi-ness
plan, which would be assessed by independent experts.
Banks were required to meet capital adequacy ratio of 4%,
8% and 10% by the end of 1998, 1999 and 2000 respec-tively.
These requirements were strictly executed and dur-ing
the financial year 1998/99 the government closed 48
banks. State banks were jointly recapitalized and four of
them merged into new state bank (Bank Mandiri) which
became the largest bank in the system with about 30% of all
deposits. After mergers and closings, the number of banks
dropped from 238 to 157. As a result there was a dramatic
change in the ownership structure of the banking system –
the government's stake rose from 40% to 70%. Despite
these efforts, the state of the banking system, as of March
2000, still leaves much to be desired with non-performing
loans ratio of 32%.
The recapitalization program has been financed by the
issuance of bonds worth over 50% of GDP [29]. The cost
of the bailout will be a substantial burden for public finances
– the public debt amounts already to over 90% of GDP.
IBRA liquidity credits were to be converted into equity or
subordinate debt. Some of the capital is planned to be
regained by consequent privatization. Restrictions on for-eign
investors to own banks in Indonesia have been accord-ingly
removed. However, the prospects of asset recovery
from bankrupt and restructured banks are very dim – the
market estimate of the IBRA portfolio is 20% of its book
value.
The second urgent problem was corporate and inter-bank
external debt restructuring. Foreign banks were very
reluctant to rollover the debt of falling companies. On Janu-ary
27 1998, the government had to announce a corporate
debt payment moratorium. Talks with a steering committee
of private bank creditors concerning the restructuring of
interbank and corporate debt began in February 1998 and
were concluded on June 4, 1998, in Frankfurt. Agreement
on interbank debt involved an offer to exchange the debt
maturing by end-March 1999 with the new loans. They
[27] Similarly the 1999/2000 fiscal deficit was only 1.5% of GDP and fail to reach 5% planned in the budget.
[28] Suspension of shareholders rights, assumption of ownership by IBRA and management replacement.
[29] This cost is significantly higher than the costs other crisis countries had to incur. The cost of the banking sector restructuring in percent of GDP
CASE Reports No. 39
were: 17% in Korea (1997- ), 29% in Thailand (1997-.), 29% in Chile (1981-87), 19% in Mexico (1994–99).](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-94-320.jpg)
![[30] This was similar to Mexican corporate debt restructuring framework called "Ficorca".
[31] The London approach to debt resolution is a voluntary, non-binding framework in which creditors agree to keep credit facilities in place, seek
[32] In USD terms Indonesian shares were 12 times cheaper (!) in September 1998 than in June 1997 – this is what people call "the fire sale FDI".
97
The Episodes of Currency Crises in Latin...
were backed by a full dollar guarantee of Bank Indonesia
and of maturities from one year (not more than 15% of the
new loans) to four years (at least 10% of the new loans) at
an interest around 300 basis points above Libor. Foreign
banks committed to maintain trade financing as far as possi-ble.
To eliminate a crunch in international trade payments
and kick-start import/export activities, Bank Indonesia set-tled
the trade arrears of commercial banks amounting to
more than 1 bln USD. Unlike with the banking system, the
Indonesian government was reluctant to extend direct sup-port
to the private sector, but preferred instead to provide
a government-supported umbrella for restructuring private
sector debt with some tax concessions and preferential
financing rates, but without any formal guarantees [30].
This was reflected in the scheme of corporate debt
restructuring agreed in Frankfurt. It provided a framework
for voluntary restructuring of external debt through direct
negotiations between debtors and creditors with a support
and mediation of a new governmental body called the
Indonesian Debt Restructuring Agency (INDRA) established
in August 1998. Its task was to provide exchange rate guar-antees
under condition that the agreement met certain con-ditions
(a minimum eight years maturity and three years
grace period). INDRA would not guarantee payment, but
only the exchange rate and would supply foreign exchange
using the best 20-day average rate before June 1999, with a
reset option if the rupiah appreciate more. The INDRA
scheme was complemented by so called "Jakarta Initiative",
i.e. the set of guidelines for debt restructuring workout
based on a London approach [31]. In the meantime the
new, tough bankruptcy law took effect in August 1998, and
the reluctant companies had an additional incentive to join
INDRA-scheme. The market reactions and the experience
with implementation are not too satisfactory. By February
1999, some 120 companies with total debt of 18 bln USD
were registered to Jakarta Initiative. By July 2000, only 5 bln
USD has been rescheduled, i.e. not much more than 1% of
total eligible debt. Numerous institutional and political
obstacles and the failure of the legal system to pose a cred-ible
threat to the debtors obstruct the process. The corpo-rate
debt resolving continues, but its slow pace undermines
the economic recovery and market confidence in Indonesia.
In December 2000, the IMF warned Indonesia of the possi-ble
consequences and urged the authorities to deal with the
problem.
The stock market index hit an all-time low in September
1998 but as soon as monetary and political conditions stabi-lized
the market rebounded quickly as foreign investors
took advantage of unbelievably cheap equity prices [32].
The bourse reached pre-crisis levels in early 2000 but soon
after that, in the first month of 2000, lost 40% due to a pro-longed
crisis and higher US interest rates. The property
market remains weak with a 35% vacancy rate in office real
estate.
5.3.5. Prospects for the Future
In early 1999, new electoral laws were adopted and in
June the country's first free and honest elections were
held. In contrast with the past, there was more than one
candidate for presidential post, which finally was won by
Abdurrahman Wahid, an open-minded and relatively liber-al
leader of the major Muslim organization. GDP grew a
slight 0.1% in 1999 and is estimated to grow between 3%
and 4% in 2000 and 4–5% in 2001. The rupiah gradually
depreciated from 6900 IDR in October 1999, right after
the new elections, to around 9500–10000 IDR/USD in
December 2000. Indonesia is still heavily dependent on
international support. Failure to meet the IMF's and for-eign
creditors' expectations can still have serious conse-quences
but pressure from outside strengthens the pro-reform
faction in the government. So, ironically, the
authorities (can) take advantage of the crisis to push
through some important reforms.
5.4. Conclusions
Indonesia is the most hard hit country among the East
Asian crisis' victims. GDP fell 14% in 1998 and only after
two years does it show any sign of recovery – it is going to
take a long time until GDP growth returns to pre-crisis lev-els
of 7%. As a result of the crisis, the political regime col-lapsed,
social and ethnic tensions erupted, the country's
integrity has been threatened and a poverty problem
emerged.
This paper tried to answer why this was the case. The
answer is that Indonesia had probably the worst economic
fundamentals of all the East Asian countries. The economy
was ridden by corruption and monopolized. The weak
financial system engaged itself in reckless credit expansion,
dangerously risky investments or even quasi-criminal activi-ties.
Overoptimistic corporate sector accustomed to the
abundance of capital ceased completely to insure against
economic risk. Short-term external debt mounted, with a
presumption that it would never have to be paid back and
out-of-court solution and work together in good will.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-95-320.jpg)
![99
The Episodes of Currency Crises in Latin...
Appendix 1: The chronology of the
Indonesian crisis
1997
May: Thai currency comes under speculative pressure.
July: Thai, Malaysian, Philippine, and Indonesian curren-cies
all depreciate.
August 14: Indonesia abolishes its system of a man-aged
exchange rate. The rupiah starts to depreciate
September 16: 15 government "mega-projects" are
postponed.
October 8: Indonesia says it will ask the IMF for
financial assistance.
October 31: Indonesia's IMF package is unveiled. It pro-vides
for more than 23 USD bln in aid.
November 1: Sixteen banks are closed as first step in
IMF package, what causes panic and bank runs among
depositors.
November 5: IMF approves a US$10 billion loan for
Indonesia as part of the massive international package.
December 5: Soeharto takes 10 days rest after a 12-day
world tour and misses ASEAN summit.
December 9-12: Finance Minister fails to negotiate the
debt rollover in Washington.
1998
January 6: Indonesia unveils an expansionary 1998/99
budget, contrary to IMF demands of a budget surplus. The
rupiah loses half its value over a five-day period.
January 9: Ratings agency Standard & Poor downgrades
Indonesia's currency to "junk bond" status.
mid-January: More or less direct calls start to be made
for a change of the regime.
January 15: Soeharto signs a new IMF agreements.
January till mid-February: Anti-Chinese food riots take
place in at least a dozen places throughout Indonesia.
mid-January: All but 22 of the 286 companies listed on
the Jakarta stock exchange are technically bankrupt. Prop-erty
companies are the worst.
January 27: Government announces a moratorium on
repaying debts and interest, and promises to guarantee all
deposits of commercial banks.
February: Soeharto proposal of a currency board is
announced, criticized and finally turned down.
February : Talks with a steering committee of private
bank creditors concerning the restructuring of interbank
and corporate debt begin
March 10: Soeharto is re-elected to a seventh five-year
term with Habibie as vice president.
May 4: Fuel prices are increased by up to 71 percent.
Three days of riots follow.
May 9: Soeharto leaves for a week-long visit to Egypt.
May 12: The army troops shoot four students at Jakarta
protest.
May 13-14: Rioting spreads throughout Jakarta. Estimat-ed
1,200 people die in two days. When Soeharto returns
from Egypt, he faces a flood of calls to resign.
May 21: Soeharto resigns and hands power to Habibie.
June 4: Agreement concerning debt restructuring is
reached in Frankfurt.
June 17: The rupiah again hits 17,000 against the dollar.
July 29: The central bank certificates auctions system
was improved and changed in attempt to regain control
over monetary aggregates.
September 24: Paris Club reschedules $4.2 billion of
sovereign debt. Annual inflation rises to 82.4 percent in Sep-tember.
September 29: Indonesia strengthens bank recapitaliza-tion
scheme.
October: Monetary stability gradually returns, inflation-ary
pressure eases and the rupiah stabilizes around 9000
IDR/USD.
November 10: Special session of the Parliament begins
to discuss election and political reforms.
1999
March 13: Government closes 38 insolvent banks.
June 7: Indonesia holds first democratic election since
1955.
August 6: Finance Minister admits there were "irregular-ities"
in loan-recovery process. The scandal prompts IMF
and World Bank to threaten loan suspension.
October 1: Indonesia announces seventh month of
deflation, with annual inflation of 1.25 percent.
October 20: Wahid has been elected President.
[33] This chronology wes much to "CNN Asia Now", October 1999 and "Inside Indonesia", report No. 54, April-June.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-97-320.jpg)
![101
The Episodes of Currency Crises in Latin...
Part VI.
The South Korean Currency Crisis, 1997–1998
by Monika B³aszkiewicz
6.1. Was Korea Different?
6.1.1. Introduction
Financial crises' episodes of the 1990's differ from those
of the 1980's in that recently they have occurred in coun-tries
on their way to social and economic development. The
initial success in implementing reforms, and the evident
prosperity for the future, encouraged foreign investors to
diversify their portfolios towards emerging markets. Addi-tional
capital inflows allows financing economic growth and
speeds up the process of integration with the global market.
On the other hand, the same capital can cause troubles
when it becomes an abrupt and sharp outflow. Why and
when does it happen?
The group of countries that came under speculative
attack in 1997 in Southeast Asia can be roughly summa-rized
within the above scenario. Yet the roots underlying
the crisis in each country within the same group should be
treated as country-specific. Of course, there were some
common features among them, but in order to identify
main characteristics for an individual economy, a separate
analysis is required. This is even truer for South Korea
(hereafter referred to as Korea) where strong macroeco-nomic
performance until late 1997 did not let most ana-lysts
foresee the crisis. Even the eruption of crises in coun-tries
from the neighboring region, marked by the July Thai
bath devaluation, did not downgrade the assessment given
by the international rating agencies to Korea.
This paper aims to explore the major factors lying
behind the Korean financial crisis. It further looks at the
sources of the crisis that took its roots in highly leveraged
companies with a weak balance sheet, and a poor func-tioning
banking system. It shows that while macroeco-nomic
imbalances played a minor role, the close relation-ship
among banks, corporations and the government cre-ated
CASE Reports No. 39
problems, which resulted in numbers of bankruptcies
and in the end led to the sharp and unexpected economic
downturn.
6.1.2. Background to the Crisis
The currency crisis, which erupted in Korea in the end
of 1997 hardly fits the first or second-generation conceptu-al
frameworks, where irresponsible government policies
and investors' panic play a crucial role. The sudden collapse
also cannot be solely attributed to the possible contagion
effect across the Asian countries facing similar problems at
that time. The 1997 Korean experience is an example,
which confirms that financial crises occur not only when
macroeconomic but also microeconomic indicators identify
vulnerabilities. Although indicators like GDP growth, infla-tion
or fiscal balances are important measures of economic
soundness, healthy financial and corporate sectors are an
essential prerequisite to a successful financial system dereg-ulation
as well as liberalization of capital account.
In an oversimplification of the classification, the eco-nomic
fundamentals can be divided into two broad cate-gories:
macro and micro-economic. Looking solely at the
former, many failed to predict the 1997 Korean crisis. This
is because at the onset of the crisis, macroeconomic funda-mentals
in Korea remained relatively sound and did not
show many signs of vulnerability. Real GDP growth rate
oscillated around 8 percent between 1994 and 1996. At the
same time, inflation measured by CPI was under control
and averaged at 5.1 percent per annum. The price stabi-lization
led to a gradual decline in nominal interest rates.
The three-year corporate bond yields, declined from 16.2
percent in 1992 to 11.9 percent in 1996. The consolidated
central government position was balanced or even in sur-plus
and public debt was less than 10 percent of GDP in the
end of 1996 [1].
[1] This number, however, can be misleading due to substantial quasi-fiscal burden in Korea arising from the governmental support to privately own
banks. In this case large public expenditures did not appear on the general government balance sheet. Additionally, such a practice posed a moral haz-ard
problem.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-99-320.jpg)
![102
Marek D¹browski (ed.)
1994 1995 1996 1997 1998
Real GDP (percent change) 8.3 8.9 6.8 5.0 -5.8
Inflation (CPI) 6.2 4.5 4.9 4.5 7.5
Fiscal balance* 0.1 0.3 0.0 -1.7 -4.2
Gross national savings* 35.5 35.4 33.5 32.5 33.4
Gross domestic investments* 36.5 37.2 37.9 34.2 20.9
Yield on 3-year corporate bonds 12.9 13.8 11.9 13.4 15.1
Current Account balance* -1.0 -1.7 -4.5 -1.7 10.9
*Percent of GDP
Source: IMF, IFS; own calculation
CASE Reports No. 39
Before financial liberalization, which started in the early
1990's, Korea was targeting monetary aggregates like M2 or
MCT [2]. This helped the authorities to maintain financial
stability since all other instruments were set to achieve the
growth rates of money, inflation and interest as well as
exchange rate. Despite the progress made towards the
financial sector opening in the 1990's, this policy was con-tinued
as the government was convinced about its advan-tages.
Two years before the crisis, the annual growth rate of
M2 oscillated around 15 and 16 percent. Although this was
3 percent lower than the early 1990's average, domestic
credit rose at the very rapid pace during 1994–97 achieving
18.5, 14.1 and 20.1 percent, respectively. Additionally, the
broad money aggregate expressed in ratio to foreign
reserves was around 6, which was high even in comparison
to other crisis economies. In Malaysia this liquidity indicator
was equal 4 and in Thailand 4.9 at that time.
Nevertheless, other macro-indicators were accept-able.
Unemployment rate did not exceed 2.3 percent
throughout three pre-crisis years, 1994–1996. Until mid-
1995 investment and saving rates were soaring, averaging
approximately at 37 and 35 percent of GDP, respectively.
The only exception was the current account deficit, which
deteriorated to 4.5 percent of GDP at the end of 1996
and was mostly covered by short-term portfolio invest-ments.
There were two factors responsible for the perfor-mance
of this indicator. The first related to strong capital
inflows during the whole year (especially in the second
quarter); the second was a terms of trade shock, repre-senting
a 12 percent drop from the previous year (accord-ing
to many empirical research shocks to this variable
increase the probability of financial crisis). In particular,
the unit export price of semi-conductors during 1996 fell
by more than 70 percent in the semi-conductor manufac-turing
industry. The magnitude of the current account
deficit even though large was not tremendous. Many
countries suffering from the episodes of financial crises
Table 6-1. Macroeconomics fundamentals
Figure 6-1. Real effective exchange rate
95
90
85
80
75
70
65
60
55
50
Jan-93
Jul-93
Jan-94
Jul-94
Jan-95
Jul-95
Jan-96
Jul-96
Jan-97
Jul-97
Jan-98
Jul-98
Jan-99
Jul-99
Jan-00
Index Value, 1990=100
Source: Moodys database
[2] MCT is composed of M2, certificates of deposits and trust accounts.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-100-320.jpg)
![103
The Episodes of Currency Crises in Latin...
experience more severe imbalances. The same is true for
the behavior of the won/ dollar exchange rate [3].
Considering tight Korean linkages with Japan (widely
fluctuating yen/dollar exchange rate was a key determi-nant
of the Korean competitiveness) and appreciation of
US dollar vis-a-vis Japanese yen at the beginning of 1995,
the won/ dollar exchange rate would be a poor approxi-mation
of the total overvaluation of won (annual percent-age
changes in 1994 and 1995 were 7.5 and 6.1, respec-tively;
in 1996 the real won/ dollar exchange rate was
actually depreciating, comparing with its 1995 value).
Thus, the real effective exchange rate would be a more
adequate way of measurement. But even then, the over-all
magnitude of appreciation was not excessively unsus-tainable.
From mid-1995, the real effective exchange rate
was appreciating comparing with its 1990 value; from May
1996 on it was steadily depreciating.
6.1.3. Signs of Vulnerability
Korea's rapid growth during the past decades and its
ability to maintain prudent macroeconomic fundamentals
masked important structural weaknesses. The successful
industrialization process, which transformed the country
from one of the poorest nations of the world into one of
the most promising, was achieved mainly at the expense
Figure 6-2. Net capital inflow
15
10
5
0
-5
-10
-15
-20
-25
-30
1990Q1
1990Q3
1991Q1
1991Q3
1992Q1
1992Q3
1993Q1
1993Q3
1994Q1
1994Q3
1995Q1
1995Q3
1996Q1
1996Q3
1997Q1
1997Q3
1998Q1
1998Q3
1999Q1
bill. USD
Source: IMF IFS, own calculation
CASE Reports No. 39
of an excessively indebted corporate sector. The 1970's
and 1980's strategy, when Korean companies were small
and were taking advantage of economies of scale, proved
to be wrong in the 1990's. This was mainly due to the
overall international environmental change towards
tighter linkages among countries that increased an internal
and external competition. The external pressure to
deregulate and open the financial system in Korea was
considerable.
The fact that borrowing from abroad was half the
price of borrowing domestically (the three-month interest
rate on corporate bonds in Korea was as high as 11–12
percent, whereas in the United States it was averaging
around 5.5 percent in the second half of the 1990's) com-pounded
the foreign exchange exposure from the domes-tic
side. Between 1990 and 1996, net capital inflows were
equivalent to $69 billion of which $51.8 billion took the
form of portfolio investments. Net foreign direct invest-ments
were actually negative and equal to $7 billion. Yet,
a continuing inflow of short-term foreign capital kept the
overall balance of payment in surplus helping to fuel
investments and growth.
The official foreign reserves accumulated. Neverthe-less,
in terms of monthly import (in 1995 and 1996
reserves were enough to cover just three-month imports'
obligations) and considering the growing stock of short-term
external debt, they were not sufficient to protect
[3] Between 1980 and 1989 Korea followed a policy of the 'Managed Basket Peg' to adjust current account. Due to the large investment boom in
1990–91which shifted current account from surplus into deficit, the government decided to adopt the policy called "Market Average Rate System".
Within this policy the exchange rate was allowed to fluctuate within a band up to 2.25 percent a day. However, because of still existing capital account
restrictions, the system was more similar to fixed but adjustable peg (see OECD, 1998; Black, 1996).](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-101-320.jpg)
![104
Marek D¹browski (ed.)
CASE Reports No. 39
60
50
40
30
20
10
against the liquidity problems. These numbers were even
smaller for gross usable foreign reserves, which are calcu-lated
as the difference between official stocks of foreign
reserves (minus gold) less overseas branches' deposits of
domestic financial institutions. In practice they are hardly
available, when the need arises. In Korea the ratio of
usable reserves to short-term debt in 1996 and 1997 was
equivalent to 0.3 and 0.1, respectively. Due to the above
there is no doubt, the Korean crisis had its roots in near
depletion of foreign currency reserves.
Rushing for additional funds from Korean firms and finan-cial
institutions failed to put priority on cash management
and other forms of provision against the risk. At the same
time, banks and other financial institutions failed to detect
the full picture of individual enterprises before offering loans
to these companies. Implementation of market principles in
an socio-economic environment characterized by over-reg-ulation,
concentration of resources around business groups
and implicit government intervention in the banking system
resulted in the high exposure of Korea to systemic risk. In
the end, it eroded the asset side of financial intermediaries
when the highly leveraged firms became unable to meet
their obligations.
A key problem associated with the accelerating stock
of external liabilities in Korea was the high proportion of
short-term debt in total borrowing. Together with low
productivity of investments (the discussion on the invest-ments'
efficiency is carried out in the next section) and the
low stock of foreign reserves it affected the sustainability
of current account deficit, since it mainly depends
whether the level of external liabilities is consistent with
the country's debt servicing capacity.
6.2. The Role of Cheabols in the Future
Development of the Crisis
The role of the state in Asia's development in the
1970's and 1980's was substantial. It is often believed that
it was exactly the government's intervention and planifica-tion
that made the "miracle" possible. However, as the
1997 meltdown showed, this common view is question-able.
For many years, investments in Korea were concen-trated
around cheabols, the multi-company business
groups operating in a range of markets under common
supervision and financial control. Although, each compa-ny
within a group was legally independent, in reality
cheabols were fostered by government policies. It was
the Presidential Declaration on Heavy and Chemical
Industrialization Policy of January 1973, which encouraged
large companies to invest in strategic industries such as
semiconductors, shipbuilding, steel etc. [IMF, 1999;
OECD, 1998]. The government support, apart from the
implicit risk share for preferential industries, was massive.
For example, the state-owned banks were pressured to
allocate more than half of their loan portfolios to particu-lar
sectors. By the same token, strategic industries were
provided loans that carried out low interest rates [Nam et
al., 1999]. Even after the financial sector's liberalization
and privatization in the 1980's, the governmental support
for the large firms, affiliates of cheabos, did not vanish.
The large economic concentration around business
groups that were subject to special regulations in Korea is
clearly evident in terms of capital stock invested and the
Figure 6-3. Reserves
0
1993 1994 1995 1996 1997 1998
bill. USD
Usable Gross Reserves (left scale) Usable Gross Reserves/Import (right scale)
Total Reserves - minus Gold (left scale) Total Reserves - minus Gold/Import (right scale)
5
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
months of import
Source: Bank of Korea](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-102-320.jpg)
![105
The Episodes of Currency Crises in Latin...
Table 6-2. Share of cheabols in mining and manufacturing
number of people employed. According to the OECD 1998
Survey on Korea, the thirty largest business groups subject
to special regulations accounted for about two-fifths of the
capital stock in mining and manufacturing sectors and
almost a fifth of employment in 1996. In terms of shipments,
their 1994 market share was 39.6 percent (data for 1995
and 1996 was not available). In 1995 the top thirty cheabols'
value added accounted for 16 percent of GNP and 41 per-cent
of value added in the manufacturing sector [Borenstein
and Lee, 1999].
Capital concentration around business groups was
definitely important, but not the only problem of weak
corporate governance in Korea. The other relates to the
corporate concentration of ownership around founding
family, relatives and affiliate firms that created corruption.
Most of the time, cross-guaranteed debt financing led to
the chain reaction, resulting in a collapse of the whole
business group. As the OECD report states, in 1996 there
were three cheabols with a share of founding family high-er
than 20 percent. In 1997 all three were insolvent. Even
considering the declining trend in internal ownership
(between 1983 and 1997 it fell from 57 to 43 percent)
due to capital market development, only about a quarter
of the 669 firms affiliated with the top thirty cheabols
were listed on the stock market in 1995.
Another feature practiced by cheabols was so called
"empire-building", the term that relates to diversification
of business groups into the broad range of industries.
From 1970 to 1996 the number of companies affiliated
with thirty largest cheaboles' increased by 18 from aver-age
4 to 22 companies investing in almost 19 industries
[OECD, 1998]. Diversification of business into a wide
range of different economic activities itself is not a nega-tive
practice since it protects against the possible loss at
one market by gaining profits at another. Nevertheless,
the fact still remains that in Korea, companies felt pro-tected
not only by the diversification of their business
portfolio, but also because of the governmental interven-tion
ensuring takeover rather than bankruptcy. Feeling
free of risk, cheabols were engaging themselves in invest-ments
based heavily on debt financing.
6.2.1. Debt Financing
The total corporate debt measured as the ratio
between company's liabilities and its capital employed
900
800
700
600
500
400
300
200
CASE Reports No. 39
1984-89 1991 1992 1993 1994
Top five
Shipments 22.5 23,4 23,8 23,0 24,6
Employment 10.02 10,8 10,8 10,4 11,1
Top thirty
Shipments 38 38,8 39,7 38,1 39,6
Employment 17.95 17,7 17,5 16,6 17,7
Source: Yoo and Lim (1997) and Fair Trade Commission cited in 1998 OECD Survey on Korea
Figure 6-4. Total corporate debt
100
1990 1991 1992 1993 1994 1995 199 6 199 7 1 998
%%, Trillion of Won
Source: Bank of Korea](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-103-320.jpg)
![106
Marek D¹browski (ed.)
was increasing steadily throughout the 1990's, achieving
its peak in 1997.
There were several remarkable features of this debt,
which explicitly contributed to the collapse of many cor-porations
and implicitly, through the growing number of
non-performing loans, to the bankruptcy of banks and
non-banking financial institutions (NBFIs).
First, the increasing trend towards indirect financing of
the corporate sector in Korea mirrored the relatively
undeveloped bond and equity markets (between the first
half of 1996 and 1997 the exposure of banks and NBFIs to
cheabols almost doubled while direct financing declined
by 20 percent (OECD, 1998)). In 1996, in terms of capi-talization,
the equity market in Korea was equal to 25.4
percent of GDP. This fell far below that of the developed
world (108.7, 67.6, 47.8 for the United States, Japan and
G-10 Europe, respectively) and represented a sharp fall
from 1990, when the equity market capitalization was
equal to 43.6 percent (BIS, 1997 Annual Report). Highly
leveraged cheabols became prone to shocks that cause a
fall in cash flow (i.e. a terms of trade drop) or an increase
in payment obligations (due to an interest rate increase).
The situation became worse at the beginning of 1997
when an almost 50 percent slide of market equity value
was observed compared to its 1995 high. This affected
not only cheabols, but also banks as cheabols were pur-chasing
equity for loans granted. The similar trend was
observed in terms of the market capitalization of shares of
domestic companies (main and parallel markets, excluding
investment funds). In 1994, the capitalization was equiva-lent
to around 118 percent of GDP. However, by the end
of 1997 it dropped together with the stock market
decline to only 23 percent of GDP.
Secondly, the corporate sector debt in Korea was, to a
high degree, concentrated in the thirty largest cheabols.
What is more, as the Table 6-3 shows, the exposure of the
thirty largest chaebols to non-bank financial institutions in
Korea was increasing.
Between 1988 and 1992 the share of banks in corpo-rate
debt financing dropped by 6.3 percent, but that of
NBFIs increased by 8.1 percent. Taking into consideration
the minimum supervision imposed on non-banking finan-cial
intermediaries, there was no doubt they were eager
to make loans to the business sector and individuals.
Overall, the high dependence of the Korean corporate
sector on debt as opposed to equity finance was clearly
evident and extremely high even by international stan-dards.
Throughout all the 1980's and 1990's the debt/
equity ratio averaged from around 400 percent to 500
percent plus [4]. On the other hand, the debt/ equity ratio
for the United States oscillated around 50–100 percent,
for the United Kingdom slightly less at that time. Even
Japan, which expanded beyond 350 percent in 1980, in
1994 was down to around 150 percent. Enormous debt/
equity imbalances in Korea had their roots in the system
of debt guarantees within cheabols, lax capitalization rules
and low effective tax rates on interest income implement-ed
to pursue the rapid growth of the economy [IMF,
1999].
6.2.2. Investments
Although it is true that investment rates in Korea were
high, the central question remains if they were profitable
CASE Reports No. 39
Table 6-3. Share of loans to the 30 largest chaebols in total loans by financial institutions, percent
1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
Banks* 28.6 26.3 24.2 20.7 19.0 18.9 17.9 15.6 15.0 13.9
NBFIs - - 32.4 36.6 37.8 38.5 40.5 - - -
*Deposit money banks only.
Source: The Bank Supervisory Board, The Korea Investors Service, Inc. Quoted from Nam, et al. (1999)
Table 6-4. Top five largest cheabols (unit: trillion won, %)
Debt/ Equity ratio
1996* 1997 1998
1.Hyundai 440 572.3 449.3
2.Samsung 279 365.5 275.7
3.LG 345 507.8 341.0
4.Daewoo 391 473.6 526.5
5.SK 352 466.2 354.9
*Data for 1996 is from April, otherwise end of the year.
Source: Dongchul Cho and Kiseok Hong (1999), Ministry of Finance and Economy
[4] The debt/ equity ratios for the top thirty cheabols can be found in Appendix 1 of this paper.](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-104-320.jpg)
![Figure 6-6. Profitability of the corporate sector
4. 0
3.5
3 . 0
2.5
2 . 0
1.5
1.0
0.5
0
98
96
94
92
90
88
[5] The incremental capital-output ratio (ICOR) is subject to possible faults as its value can be influenced by factors not solely dependent on the
[6] Manufacturing ordinary income is calculated as total sales less cost of sales plus selling and general administrative expenses, less net non-oper-ating
cost.
107
The Episodes of Currency Crises in Latin...
Figure 6-5. ICOR*
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
1988 1989 1990 1991 1992 1993 1994 199 5 1996
* The calculation was done as a five-year moving average to avoid cyclical effects.
Source: IMF IFS
enough to contract such a high debt? By comparing the
cumulative investments with changes in GDP, it is almost
certain that some of them were misdirected [5].
Since the late eighties, the value of the capital-output ratio
(ICOR) was increasing, indicating the falling quality of fixed
capital formation. A slight drop was observed between 1991
and 1993. But after then, the ICOR index was systematically
growing. Considering the slow down in GDP growth rates
since 1995, there is no doubt there was not enough capacity
in the corporate sector in Korea to cope with such high rates
of credit. The overheating pressures amplified.
Apart from ICOR, other indices like the rate of return
on assets, the growth rate of ordinary income as well as
the break-even point, showed the low productivity of
investment and the growing vulnerability of the Korean
corporate sector [6]. The last variable indicates the level
efficiency of invested capital. Other factors include structural weaknesses and capital deepening.
CASE Reports No. 39
1991 1992 1 993 1 994 1995 1996
% %
86
%%
Ordinary income (left scale) Break-even-point (right scale)
Return on Assets (left scale)
Source: Bank of Korea](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-105-320.jpg)
![108
Marek D¹browski (ed.)
of cost of production just covered by income (the higher
the value of the break-even points, the higher the cost).
As the chart shows, from 1995 onwards, all three indi-cators
– the break-even point, ordinary income and
return on assets – demonstrated the falling profitability of
the corporate sector.
The high leverage of the corporate sector and its inad-equate
governance was an important, if not the major, fac-tor
of Korea's ensuing collapse. The question to be
answered is why it was sustainable in the 1980's but
turned into a collapse in the second half of the 1990's?
The simplest answer lies probably in the unfavorable
behavior of the terms of trade, which significantly con-strained
the cash flow of chaebols, the major exporters in
Korea. But there were other factors like the appreciation
of the US dollar vis-a-vis the Japanese yen at the beginning
of 1995 that weakened the competitive position of firms.
Furthermore, the collapse of Hanbo Steel Co. in January
1997, the first big bankruptcy in decades, undermined
investors' confidence that Korean firms were 'too big to
fail'. This was followed by Moody's decision to lower the
long-term rating on three Korean banks, which had a sig-nificant
exposure to Hanbo [Park, et. al, 1998]. The com-mon
belief in the government willingness to bail out falling
companies disappeared together with the declining stock
of foreign reserves.
Low-productivity of investments was mirrored in the
burden of non-performing loans (NPLs) in the banking
system. In 1995 NPLs (included loans classified as sub-standard
and doubtful) for commercial banks were equal
to 5.1 percent of total loans raised. In 1997 it was already
6 percent, according to the Financial Supervisory Service
(cited by J. Fleming). Soon after the eruption of the crisis
the problem of NPLs in total loans magnified. In 1998 the
percentage number increased to 7.4. Although, these fig-ures
are high, they maybe even higher since regular
reports on NPLs have been available only recently.
6.3. Korean Financial System and its
Liberalization
Before the liberalization of the Korean financial sec-tor
the early 1980's, the government had intervened
heavily to pursue its industrial objectives. As the reforms
progressed, several commercial banks and non-bank
financial institutions were added to the system. Never-theless,
the attempts to liberalize were only partially
successful, still leaving many regulations in force (i.e. low
interest rate ceiling to increase profits and retain earn-ings
for selected firms, commercial banks' lending to
preferential sectors) [7].
The 1988 plan to deregulate the majority of bank and
non-bank's lending rates as well as interest rates on
money market instruments was mostly reversed, because
of pressures arising from earlier beneficiaries of the pref-erential
access to low interest rates credit. On the other
hand, the second attempt to implement the plan in 1992
was suppressed by the stock market slump [IMF, 1999].
The next phase of financial system reform took place
in 1993 when the first democratically elected civilian gov-ernment
came to power. The new government under the
President Kim Young Sam was highly committed to finan-cial
liberalization. There were two reasons for speeding
up the process. One of them was the perspective of join-ing
the OECD, the other was the growing ability of pri-vate
and already credible firms to borrow funds from
abroad. However, most capital flows attracted by firms
through the stock market were not free from explicit or
implicit quantitative controls, with the exception of trade
related short-term financing [Dooley et al. 1999]. During
that time, large structural changes led to further rapid
growth of non-banking financial institutions. As the Bank
of Korea states, the market share of non-banking financial
institutions in terms of Korean won deposits between
1980 and 1998 increased from about 29 to 72 percent.
The numbers for banking institutions (commercial and
specialized banks) were 71 and 28 percent, respectively
[OECD, 1998]. The trend in loans and discounts was sim-ilar,
increasing for NBFIs and falling for banks. Between
1996 and 1997, the share of funds raised by the business
sector from non-banks to the total funds raised increased
by 10 percentage points, from 13.9 to 23.9 percent. At
the same time borrowing from banks dropped from 14 to
12,9 percent (Bank of Korea).
6.3.1. Non-banking Financial Institutions
In Korea, non-banking financial institutions can be
roughly classified into five categories according to their
business activities. These are: development, savings,
investment, insurance, and other institutions (Bank of
Korea). The role they played in causing future deteriora-tion
in the financial sector balance sheet was significant
since they were allowed greater freedom in their man-agement
of assets and liabilities. What is more, they were
able to charge higher interest rates on their loans as well
as apply higher interest rates on their deposits. Regarding
[7] Commercial banks in Korea were nationalized in the 1960's and since then the government was influencing the sectoral allocation of credit with
CASE Reports No. 39
smaller or greater intensity [IMF; 1999].](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-106-320.jpg)
![109
The Episodes of Currency Crises in Latin...
Table 6-5. Fund Raising by the Corporate Sector
the troubles Korea faced in 1997, the number of new
licenses issued to merchant banking corporations was an
important factor. In 1993 there were just 6 merchant
banks. By 1996 this number increased to 30 as a result of
deregulation on financial transactions. In principle, mer-chant
banks were supervised by the Ministry of Finance
and Economy, but this was minimal as there was no asset
classification, capital, or provisioning rules [IMF, 1999].
Besides, most of them were owned by cheabols and were
used to finance activities within a group. To attract funds,
merchant banks, for example, were offering cash man-agement
accounts to their customers (within these
accounts, apart from getting checkbooks and credit cards,
banks' clients were able to raise loans). Banks were main-ly
investing in short-term commercial papers and notes.
The contribution of non-banking financial institutions,
and merchant banks in particular, in financing investments
of corporations was significant. Lax regulations let banks
provide loans and guarantees of up to 50 percent of their
capital. Additionally, the practice of cross-guarantees was
common, where affiliates merchant banks were financing
activities of other firms from the same business group.
Conflict of interest between these two resulted in banks'
failure to monitor the performance of their debtors
[OECD, 1998]. This problem in economic literature is
known as an adverse selection, the situation where
lenders have an incomplete knowledge of the creditwor-thiness/
quality of borrowers.
6.3.2. Capital Account Liberalization
Alongside financial liberalization, the capital account
was also progressively liberalized. Deregulation of foreign
exchange capital account transactions led to the expan-sion
of foreign branches of Korean banks. Between 1994
and 1996, Korean banks opened 28 foreign branches in
addition to 24 financial companies which were allowed to
engage in foreign exchange business upon the conversion
into merchant banks described above. In accordance with
the design program, the short-term capital movements
were liberalized in advance of long-term ones. Regard-less
of the increasing list of industries open to foreign
direct investments (FDIs), they were still subject to tight
restrictions of unclear form. Very often they were denied
because they could "disrupt the market" in the situation of
surge short-term flows. In 1993, limits on the long-term
foreign-currency denominated loans were relaxed, but
the long-term borrowing remained restricted. This deci-sion
led Korean banks to borrow funds from abroad for
the short-term and lend these funds to domestic compa-nies
for the long-term. It created a serious maturity mis-match
where banks became prone to shocks such as an
increase in interest rates. In this case, the burden
imposed in the end of 1997 was a natural consequence of
their net worth reduction. This is because, by definition,
higher interest rates increase value of banks' long-term
assets more than lowering short-term liabilities.
According to the BIS-IMF-World Bank's statistics, lia-bilities
to banks – due within the year – oscillated
between 63 and 70 percent of total external liabilities in
1994–96. The situation looked even more precarious in
terms of international reserves accumulated. By the end
of 1997, total reserve assets only covered 38% of short-term
external liabilities.
The external financial liberalization, which led to the
accumulation of short-term liabilities, exposed the Kore-an
banking sector to problems like liquidity tightening,
when some adverse news about the market caused sharp
investor reactions. Firstly, when in 1997 investors
CASE Reports No. 39
1996 % 1997 % 1998 %
Fund Raising 118,769 100.0 118,022 100.0 28,360 100.0
Indirect finance
Borrowings from
DMBs
Borrowings from non-banks
33,231
16,676
16,555
28.0
14.0
13.9
43,375
15,184
28,191
36.8
12.9
23.9
-15,003
54
-15,487
-52.9
0.2
-54.6
Direct finance
(Commercial paper)
(Stocks)
(Corporate bonds)
56,097
20,737
12,981
21,213
47.2
17.5
10.9
17.9
44,087
4,421
8,974
27,460
37.4
3.7
7.6
23.3
49,749
-11,678
13,515
45,907
175.4
-41.2
47.7
161.9
Borrowings from
abroad
Others (trade credits,
borrowing from
governments, etc.)
12,383
17,058
10.4
14.4
6,563
23,997
5.6
20.3
-10,196
3,810
-36.0
13.4
Source: Bank of Korea, Flows of Funds, 1999](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-107-320.jpg)
![110
Marek D¹browski (ed.)
200
180
160
140
120
100
80
60
40
20
stopped believing in the capacity of the government to
bail out falling companies, banks and NBFIs, they rushed
to pull their money out of the country. Secondly, the exist-ing
currency mismatch (Korean banks' foreign liabilities
were excessive to domestic assets) limited the ability to
convert domestic currency into foreign currency. Thirdly,
when the won/ dollar exchange rate started to depreciate,
short-term foreign currency obligations of banks and non-banking
financial institutions as a share of domestic assets
increased significantly. The lack of liquidity of Korean
banks was also clear by international standards. On aver-age,
between 1995 and 1997, the ratio of liquid assets to
liquid liabilities (a three months period is considered to be
"liquid") was 60 percent in comparison to 100 – an inter-national
standard [Nam et al.1998].
It is also important to note that even though the total
external debt as a ratio of GNP increased from 13 to 22
percent between 1990 and 1996, it was not as large as in
CASE Reports No. 39
Figure 6-7. External debt
0
1994 1995 1996 1997 1998
%%, USD
Liabilities to banks - d ue within a year/
/Total liabilities
International reserve assets
(excluding gold)/Short-term liabilities
Source: BIS-IMF-World Bank joint statistics, BIS web side
Figure 6-8. Net foreign assets
20000
15000
10000
5000
0
-5000
-10000
-15000
Jan-94
May-
Sep-94
Jan-95
May-
Sep-95
Jan-96
May-
Sep-96
Jan-97
May-
Sep-97
Jan-98
May-
Sep-98
Billion of Won
Source: IMF IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-108-320.jpg)
![111
The Episodes of Currency Crises in Latin...
Table 6-5. Liquidity ratio of the 10 largest korean banks
18
16
14
12
10
8
other Asia and Latin America countries [8]. In 1996 the
ratios for the Philippines and Thailand were 54 and 46
percent, respectively. For Mexico, prior to the 1994 cri-sis,
it was 35 percent [Park et al, 1998]. The same con-clusion
is drawn from other indicators. The percentage
share of foreign liabilities in total liabilities of the banking
system was about 13 percent in 1996–97 for Korea,
whereas in Indonesia averaged around 25 percent at that
time. In Argentina on the other hand, it surged up to 20
percent in 1997 (IMF, IFS). There were two facts that
seemed to be more important than the overall magnitude
of external debt. One was its increasing trend since the
early 1990's; another was associated with the high expo-sure
of the banking system to short-term foreign bor-rowing
discussed above.
Progressive capital account liberalization also covered a
higher ceiling on stock investments for non-residents, with
the aggregate ceiling of 26 percent and individual ceiling of
7 percent by November 1997. Others included borrowing
from international bond markets by Korean companies with
prior notification, foreign purchasing of certain types of
bonds or non-guaranteed corporate and SME bonds [IMF,
1998]. But despite of liberalization of capital account trans-actions
some restrictions remained (see Appendix 2).
6.3.3. Credit Expansion
The industrialization strategy implemented in Korea
fuelled by the financial system liberalization resulted in
domestic credit expansion. Furthermore, the surge in
banking borrowing was typified by five important charac-teristics
(all key for the future of the banking system):
– Credit, in the most part, was extended to the pri-vate
sector to finance new investments; public sector
borrowing played a minor role,
CASE Reports No. 39
1995 1996 March 1997 September
1997
80-90% 1 3 2 2
70-80% 2 2 1 1
60-70% 4 2 4 5
Below 60% 3 3 3 2
Source: Shin and Hahm (1998) cited in Nam et al. (1998)
Figure 6-9. Currency mismatch*
6
1994Q1
1994Q3
1995Q1
1995Q3
1996Q1
1996Q3
1997Q1
1997Q3
1998Q1
1998Q3
1999Q1
1999Q3
Source: *Foreign liabilities over domestic assets, deposit money banks
Source: IMF, IFS; own calculation
[8] This data does not include offshore borrowing of domestic financial institutions, overseas borrowing of foreign branches of domestic financial
institutions and borrowing of overseas branches of domestic enterprises. When these are incorporated, the total external debt jumps from 121 to 170
billion of dollars at the end of 1997 [Park et al, 1998].](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-109-320.jpg)
![112
Marek D¹browski (ed.)
40
30
20
10
0
-10
-20
-30
-40
-50
-60
– Explicitly, it indicated a falling quality of loans and ampli-fied
the probability of accelerating non-performing loans,
– Borrowing, for the most part, was short-term and
foreign currency dominated,
– Non-banking financial institutions were the major
intermediaries of funds,
– Expansion of overseas branches of domestic financial
institution resulted in the situation where 70 percent of
total debt accounted for the banking sector, direct finance
played a minor role [Dooley et al. 1999].
Bank credit grew more than 20 percent per annum in
1996 and 1997. Moreover, the ratio of bank credit to GDP
was also increasing at a very high pace. The fact that loans
were invested in the risky business of low profitability and
declining rates of return (the discussion on efficiency of
investments and profitability of the corporate sector was
carried out in the previous section) led many of them to
become non-performing, putting an extraordinary burden
on the banking sector.
Alongside the domestic credit growth, total claims on
the private sector were also increasing. Between 1994
and the first quarter of 1996, total claims of deposits in
banks as a percentage of GDP was averaging around 55
percent of GDP. From the second quarter on, it was sys-
CASE Reports No. 39
Figure 6-10. Net domestic credit (annual percentage changes)
-70
1995M1
1995M4
1995M7
1995M10
1996M1
1996M4
1996M7
1996M10
1997M1
1997M4
1997M7
1997M10
1998M1
1998M4
1998M7
1998M10
%%
Source: IMF, IFS
Figure 6-11. Claims on the private sector
80
75
70
65
60
55
50
45
40
1994Q1
1994Q3
1995Q1
1995Q3
1996Q1
1996Q3
1997Q1
1997Q3
1998Q1
1998Q3
Source: IMF, IFS](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-110-320.jpg)
![113
The Episodes of Currency Crises in Latin...
tematically growing, reaching 65 percent at the onset of
the crisis.
6.3.4. Risk Assessment in the Banking System
The sharp increase in bank lending – fuelled by very lax
provisioning rules and insufficient risk assessment – was
mirrored by the growing number of non-performing loans
(NPLs) in Korea. The fact that there were no regular
reports did not permit an adequate assessment of the health
of the banking sector. The depth of the problem is clear
when the data on non-performing loans as a percentage of
total loans from 1996 is compared with the revised data for
the same time period. The 1996 number for NPLs as a per-centage
of total loans states for 0.8 percent, whereas the
revised one for 4.1 percent. Partially, this discrepancy is
connected with the classification of substandard loans in
Korea, partially with the lack of regular reports already
pointed. Usually, loans being three months plus in arrears
are considered as substandard. But in the Korean Republic,
this rule was extended to six months plus. Compulsory pro-visioning
imposed on these loans varied from 20 to 75 per-cent,
although this depended on the types of collateral and
guarantees. In many cases, only bad loans (NPLs not cov-ered
by collateral) were reported as non-performing. The
transmission mechanism between the banks, non-banking
financial institutions and Korean corporations led to the pre-sumption
that the real number of compulsory provisioning
was closer to the lower bounder – or was even below it.
Adding to this story the number of corporate and banking
bankruptcies in the end of 1997 and the beginning of 1998,
it is obvious that Korean banks failed to adequately assess
credit risk.
The economy experienced troubles also in terms of sol-vency
indicators of the banking sector. The capital adequa-cy
ratio based on the Basle Core Principle requires a mini-mum
ratio of eight. Yet, the domestic regulation was looser
and required only four percent. In 1996 the actual capital
adequacy ratio in Korea was just 9.1 percent representing a
2-percentage point drop from the 1993 value [9]. In Thai-land
and Hong Kong, meanwhile, it was equal to 11.3 and
17.5 percent respectively. There were other factors like
soft accounting rules, which gave the authorities the room
to manipulate the capital adequacy ratio. The growing mer-chant
banks' off-balance sheet credit guarantees were
alarming prior to the crisis. In 1996, off-balance sheet cred-it
guarantees were 49.3 percent expressed in ratio to total
assets. It was higher by 12.5 percentage points than the
1993 average. For commercial banks the number was small-er
and equal to 8 percent (the 1.2 percentage points drop
from 1995). This piece of evidence points on the immense
role the merchant banks played in the debt-financed growth
strategy in Korea.
Of course Koran banks had some prudential regula-tions
to limit the probability of financial difficulties, but
they were not successful in preventing excessive risk tak-ing.
For example, in 1996, in order to prevent excessive
risk-taking, Korean banks had constraints on foreign cur-rency
exposure. The sum of long positions as a percent-age
of total capital was limited to 15; the sum of short
positions to 10 percent. The spot short positions were
limited to 3 percent of bank capital or to 5 million of USD,
whichever was greater. Maximum lending to a single bor-rower
was 15 percent on the total bank capital, the same
number as in United States. Nevertheless, the growing
currency mismatch in Korea suggests that the imposed
limits were relatively flexible and that foreign investors
were mostly responsible for the growth of foreign assets
of Korean banks.
6.4. The Onset of the Crisis
The overview of the financial and corporate sectors'
situation preceding the crisis showed that microeconom-ic
distortions in Korean economy were accumulating for
quite some time. Even if the crisis was initiated by the
subsequent massive and abrupt capital outflow, the fact
remains that many other factors contributed to the 1997
crash more than a simple herding. Definitely financial lib-eralization
without adequate regulations imposed, long-lasting
governmental control of the financial sector as well
as the weak corporate governance were fundamental to
the crisis. But apart from structural weaknesses sketched
in the previous sections, there were other signs of vul-nerabilities
building up in the economy and eventually
causing the external liquidity crisis:
– the slow down in the GDP growth rate,
– the steady decline in stock price levels, lowering the
value of banks and corporate equities and further reduc-ing
the value of their net worth,
– the sharp drop in the terms of trade affecting not
only Korean export, but also disturbing banks and
cheabols' balance sheets. According to Cho (1999) cited
in Mishkin, et al (2000), the 20 percent drop in terms of
trade accounted for more than 70 percent loss in aggre-gate
corporate profits,
– depreciation of the Japanese yen against the US dol-lar
additionally weakening the country external position,
– financial crashes in neighboring countries.
[9] The capital adequacy ratio for Korea includes commercial banks only and is calculated under the Korean provisioning standards. Soft rules and
many exceptions given to non-banking financial institutions would probably further decrease this number.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-111-320.jpg)
![115
The Episodes of Currency Crises in Latin...
Sammi Steel on March. The sovereign credit rating was still
set at AA- level and did not decline until October.
As in many other events of financial (banking) crises, it is
extremely difficult to assess which model would be the most
suitable one for Korea. The scenario of speculative attacks
due to expected currency depreciation couldn't facilitate the
run on Korea simply because the existence of tight regula-tions
on currency forwards backed by corresponding cur-rent
account transactions and the absence of currency
futures market made this framework impossible [Mishkin, et
al, 2000]. It was rather a consequence of creditors' herding
behavior to withdraw their loans. The fact that Korea had a
high level of short-term debt and only moderate interna-tional
reserves favors this scenario. In retrospect however,
and regarding the structural weakness of financial and cor-porate
sectors this panic could be judged as rational.
The set of macroeconomic indicators developed to mea-sure
the probability of financial crises would blur the true pic-ture
about the health of Korean economy. Here the problem
is deeper and goes back to the early 1990's when the financial
liberalization eased restrictions on short-term capital flows
but left long-term ones in force. Together with the govern-ment's
reluctance to give up its influence in most sectors of
the economy, a series of bankruptcies of corporations under-mined
investors' confidence in the quality of their assets.
Thus, the Korean crisis was a result of interactions between
financial and corporate sectors which failed to set prudential
governance rules. The role the government played in calm-ing
down the market was also of considerable importance
since the policy was misdirected and instead of ceasing, it
fuelled the crisis. The harsh tone and desperate announce-ments
concerning good prospects of the economy gave a
warning sign to investors. The fact that the authorities were
bailing out collapsing corporations and banks raised doubts
about the solvency of the whole economy. When the govern-ment
announced the stock of official reserves at the end of
November, the market confidence about the magnitude of
usable foreign reserves was undermined and caused a further
decline in the stock market index (regarding the course of
events in Korea at the end of the 1997, see the overview of
the chronology of selected news from the financial markets
presented in Appendix 3).
6.5. The 1998 Recession and 1999
Recovery
6.5.1. The IMF Intervention in Asia
On November 20, 1997 the exchange rate band was
widened from 2.5 to 10 percent. The day after, the gov-ernment
asked the International Monetary Fund to lend
CASE Reports No. 39
support. On December 4, the IMF's Executive Board
approved a Stand-By Arrangement with Korea of $21 bil-lion
over the next three years based on the assumption of
the GDP growth in 1998 of 2.5 percent. The World Bank
and the Asian Development Bank provided an additional
$14 billion as well as technical assistance. Individual coun-tries
agreed on another $22 billion as a second line of
defense credit. The total sum granted to Korea was equal
to $58.4 billion, but the disbursement of money was
based on certain conditions:
1. comprehensive financial sector restructuring with a
clear exit policy, strong market and supervisory discipline,
and independence of the central bank. Nine merchant
banks were suspended; two commercial banks were cap-italised
by the government; all commercial banks were
required to submit plans for recapitalization,
2. the bases for corporate income and VAT widened in
order to bear the costs of financial restructuring (fiscal
measures were equivalent to about 2 percent of GDP and
were consisted with the balance budget target),
3. an end to close the relationships between govern-ment,
banks and corporations as well as an improvement
in accounting (corporate financial statements had to be
prepared on the consolidated basis), auditing (certification
of external auditor) and disclosure standards,
4. the implementation of trade and capital account lib-eralization
measures (liberalization of FDIs flows and
opening money, bond and equity markets to capital
inflows),
5. labor market reforms,
6. the publication and dissemination of key economic
and financial data (IMF on line service).
Nevertheless, the announcement of the program did
not help to stabilize the exchange rate and the usable
gross reserves fell to 8.9 billion US dollars at the end of
December 1997. In the last quarter of the 1997, net cap-ital
outflows were observed ($24.9 billion compared to $7
billion of inflows in the corresponding period of 1996).
This was greater than the IMF initially expected (the pro-gram
assumed net outflows of $11.1 billion in December
1997 and net inflows of $3.3 billion in 1998). Revision of
the program was necessary, but even then it failed to pre-dict
the total net capital outflows of $14.8 billion in 1998.
The IMF prescriptions for crisis management may be a
source of concern given the long-term adjustment condi-tions
imposed, but the role of the Korean government is
also questionable. As the government was injecting capital
to falling companies, investors did not really believe in its
commitment to market restructuring. As a result, the
exchange rate dropped by 40 percent between Decem-ber
7 and 14, the stock market declined by 17 percent
during December 3–10. It was the sharpest reduction
since the beginning of October. As the crisis accelerated,
the defense of the won became impossible and the](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-113-320.jpg)
![117
The Episodes of Currency Crises in Latin...
support of 41 billion won (10 percent of GDP) for their
disposal as well as for banks' restructuring. In result most
of Korean banks achieved capital adequacy ratio between
10 and 13 percent.
An acceleration of the recession was so abrupt that
the GDP growth rate was once again corrected to -5 per-cent
in August 1998. A supplementary budget was intro-duced
to increase expenditure needed for restructuring
distressed sectors. However, the exchange market stabil-ity
allowed further reductions in the overnight call rate,
from 25 percent at the beginning of the year to 8.5 per-cent
at the end of September. This help disturbed com-panies
and reduced the number of bankruptcies from
3000 to 1400 during the same period [Nam et al., 1999].
6.5.2. Macroeconomic Environment after the
Crisis
After the eruption of the crisis, Korea moved into a
severe recession which was not expected even by the
IMF staff. As indicated above, the growth forecasts were
revised downward, along with almost all the other
restructuring program's elements. The fiscal deficit was
also adjusted in order to accommodate weaker econom-ic
activity and costs of financial restructuring. High inter-est
rates (reaching 25 percent in January 1998) necessary
to achieve exchange rate stabilization imposed a high bur-den
on the real sector. In the first half of 1998, the gross
domestic product plunged to -5.3 percent comparing to
the year before and declined to 6.7 percent during the
whole year. The reduction in output was mostly the
result of a sharp reduction in domestic demand. The
annual percentage changes in private and public con-sumption
were equal to -9.6 and -0.1, respectively. The
gross fixed capital formation dropped by 21 percent, and
was probably the most responsible factor of the total out-put
squeeze. Exports had declined by 13 percent, yet the
slide in import was sharper and equal 22 percent. This
resulted in the current account surplus of 10.9 percent of
GDP in 1998. The unemployment rate jumped to 6.3 per-cent
during the 1st half of the year and reached 6.8 per-cent
at the end of the year (during the pre-crisis period it
did not exceed 3 percent). The consolidated central gov-ernment
budget was in deficit of 4 percent of GDP com-paring
to the pre-crisis surplus (the initial IMF program
assumed 0.2 percent surplus). The annual CPI inflation in
the first quarter of the year reached 8.9 percent, but in
the last quarter was down to 3.9 percent, which was
lower than before the crisis.
The 1998 recession seemed to be over in 1999 with
some signs of stability visible already in the second-half of
1998 which helped to fuel investments. In the third quar-ter
of 1999, gross fixed capital formation increased by 4.8
percent compared to its 1998 level. GDP growth boomed
and exceeded 10 percent at the end of 1999 – supported
by a monetary and fiscal policy expansion. The fact that
short-term interest rates were cut from 23–25 percent to
5 percent significantly diminished the costs of borrowing.
Low interest rates also helped banks and cheabols to bear
the costs of debt restructuring. The budget deficit was
reduced to 3 percent of GDP. The growth in imports
reduced the current account surplus to around 6 percent
of GDP. The exchange rate was stable oscillating around
W1200 per US dollar. Nevertheless, there was a couple of
exceptions to this impressive recovery. One of them was
an unemployment rate of 6.3 percent, which despite of
deceleration was still in a sharp contrast to the pre-crisis
average; the other was the collapse of the Daewoo group
in the middle of the year.
6.6. Conclusions
The financial turmoil which erupted in Korea in 1997
did not really fit into any group of theoretical models of
financial crisis existing in the economic literature at that
time. It is just recently when researches tried do develop
so-called third generation models with new sets of funda-mentals.
Broadly speaking, they aggregate macro and
microeconomic indicators of economic soundness.
The Korean case is a good example of the great
importance of complex reforms once the economy is
switching from the centralized to market governance.
The healthy macroeconomic background is a prerequi-site,
but the high rates of growth, on their own, are not a
safeguard of the long-lasting economic success.
In the Korean Republic, highly leveraged cheabols
were vulnerable to shocks like a collapse in investor con-fidence.
Inadequate liberalization instead of hosting long-term
capital welcomed short-term inflows. Additionally,
and even more importantly, cheabols were not efficient
and facilitated family connections. Tight relationships
among cheabols, banks and the government made the
transparency of financial and corporate sectors virtually
impossible. Even if, on average. the prudential rules on
banks in Korea were similar to those in western coun-tries,
existing exceptions created a scope for choosing the
"regain strategy". The number of financial instruments
available increased once the financial sector was liberal-ized,
and made the government attempt to control the
stock of foreign capital ineffective.
In the last Letter of Intent dated August 13, 2000
which finished the IMF intervention in Korea, the mem-bers
of the IMF Executive Board stressed the impressive
Korean rebound from the 1997 crisis. They said it was
possible due to "supportive macroeconomic policies and
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-115-320.jpg)
![118
Marek D¹browski (ed.)
the competitive exchange rate; a wide range of structural
reforms that addressed the weaknesses that contributed
to the 1997 crisis and increased market orientation; a
favorable external environment; and an improvement in
confidence resulting from both the implementation of
strong economic policies and the recovery and build-up in
foreign exchange reserves".
Despite the changes however, there is still a lot to be
done especially in the area of structural reforms. In the
special joint report the IMF and Korean Government
agreed on major policies regarding further financial and
corporate sectors restructuring in Korea. They identified
at least a few problems, which despite the progress made
in both sectors need to be solved. Bearing in mind all the
reasons for the financial crisis in Korea, a number of them
are key. Ongoing policy plans of the financial sector
include the following measures:
– while still being the owner of some commercial
banks the government will let banks to operate on a fully
commercial basis and will not be involved in the day-to-day
management,
– public funds should be only used to the extent nec-essary
to facilitate the liquidation of failed institutions and
restructuring of weak but viable banks,
– financial transactions between affiliated companies
or between institutions and their shareholders are con-fined
to transactions in assets for which market price
exist,
– all related party transactions will be disclosed in
audited accounts and reported to FSC,
– exposures of commercial banks, merchant banks,
specialized and development banks in excess of the new
20 percent or 25 percent limits will be subject to pro-gressive
reduction.
Further corporate sector restructuring consists of
preparation of the consolidated statements assessing the
overall financial structure as well as Daewoo resolution.
Others encompass further improvements in financial
transparency and accountability (Financial Supervisory
Service) [10].
In spite of all this, the latest news from Seoul is not
optimistic. The unsuccessful auction of Daewoo group
undermined investor confidence and was immediately
mirrored in the stock market decline. Additionally, the
scale of corporate indebtedness before the crisis was so
great that two years time after the crisis South Korea's
conglomerates are still uanble to overcome it. The
progress towards reducing bad loans has not been
observed yet; the target of cutting the debt-to-equity
ratio in the end of 1999 to 200 percent was not achieved.
This begs the question about the efficiency and the
speed of reforms carried out in the country. Indeed, one
can question the IMF restructuring program applied in
Korea, which failed to foresee the deepness of the crisis
and caused the sharp output downturn, but in the light of
current episodes one can also ask if the imposed condi-tionality
was severe enough to bring about 'the second
miracle.
[10] Information about elements of banking system and corporate restructuring during 1998–99 can be found in Appendix 4.
CASE Reports No. 39](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-116-320.jpg)
![122
Marek D¹browski (ed.)
Banking System Restructuring
In the banking sector, a credit crunch resulted from high-er
interest rates that increased the stock of non-performing
loans (net domestic credit of the banking system dropped
by 50 percent on the annual basis). At the end of March
1998, the ratio of loans being three months plus in arrears
to total loans was 16.9 percent for banks and 14.5 percent
for all financial institutions (Sang- Loh Kim; 1998). In order
to deal with this problem, in April 1998 an independent
supervisory authority was established to apply international
prudential standards. Additionally, to help banks resolving
bad loans government committed W32.5 trillion to the
NPLs' Resolution Fund and planned to resolve W100 trillion
of bad loans via auctions, capital increase or subordinated
bond issues. Five banks were shut down as well as 16 mer-chant
banks in 1998. The Korea First Bank and the Seoul
Bank had been nationalized after they repealed 87.5 percent
of their stock.
The Financial Supervisory Commission (FSC) set a mini-mum
target for capital adequacy ratios according with the
BIS standard for banks and merchant banking corporations
to be achieved by the end of 2000. Together with the Min-istry
of Finance measures on improving the disclosure,
accounting and accounting standards were introduced
(accounting of securities was going to be based on the mar-ket
instead of the book value). Starting from July 1, loans of
at least three months in arrears but less than 6 months were
categorized as non-performing. Prompt corrective action
system was introduced imposing recommendations, mea-sures
and orders on the unsound financial institutions (i.e.
banks with the capital adequacy ratio lower than 8 percent
minimum).
Corporate Restructuring
Corporate restructuring in Korea proceded on two sep-arate
tracks. One was a debt workout for the smaller
cheabols and other large corporations; the other included a
package for the top five cheabols. Reform bills were passed
by the National Assembly in February 1998 and among oth-ers
included:
– Tax Exemption and Reduction Control Act – tax
breaks for company restructuring were provided,
– Bank Act – the limit on bank ownership of a corpora-tion's
equity increased from 10 to 15 percent, or higher with
Financial Supervisory Commission (FSC) approval,
– Corporation Tax Act – non-deductibility of interest on
"excessive" debt was moved from 2002 to 2000,
– Foreign Direct Investment and Foreign Capital Induce-ment
Act – takeovers of non-strategic companies by foreign
investors without government approval were allowed. For-eign
investors could acquire 33 percent of shares without
board approval (23 percent increase compared to the pre-crisis
limit),
– Antitrust and Fair Trade Act – new cross guarantees
were prohibited. Elimination of existed cross guarantees
was planed by March 2000,
– Financial Supervisory Committee maintained relatively
lax rules on accounting for restructured debt in order to
help Korean banks to negotiate substantial rate reductions
and conversions of debt into equity or low-yield convertible
bonds [Mako, 1999].
CASE Reports No. 39
Appendix 4: Banking System and Corporate Restructuring](https://image.slidesharecdn.com/rc39-141023050203-conversion-gate02/85/CASE-Network-Reports-39-The-Episodes-of-Currency-Crisis-in-Latin-American-and-Asian-Economies-120-320.jpg)
CASE Network Reports 39 - The Episodes of Currency Crisis in Latin American and Asian Economies
- 1. M a r e k D ¹ b r o w s k i ( e d . ) The Episodes of Currency Crisis in Latin American and Asian Economies W a r s a w , 22 0 0 1 No. 39
- 2. The views and opinions expressed in this publication reflect Authors’ point of view and not necessarily those of CASE. This paper was prepared for the research project No. 0144/H02/99/17 entitled "Analiza przyczyn i przebiegu kryzysów walutowych w krajach Azji, Ameryki £aciñskiej i Europy OErodkowo-Wschodniej: wnioski dla Polski i innych krajów transformuj¹cych siê" (Analysis of the Causes and Progress of Currency Crises in Asian, Latin American and CEE Countries: Conclusions for Poland and Other Transi-tion Countries) financed by the State Committee for Scien-tific Research (KBN) in the years 1999–2001. The publication was financed by Rabobank SA Key words: currency crisis, financial crisis, Asian economies Argentina, Mexico,Thailand, Indonesia, South Korea, Malaysia DTP: CeDeWu Sp. z o.o. Graphic Design – Agnieszka Natalia Bury © CASE – Center for Social and Economic Research, Warsaw 2001 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, without prior permission in writing from the author and the CASE Foundation. ISSN 1506-1647 ISBN 83-7178-257-8 Publisher: CASE – Center for Social and Economic Research ul. Sienkiewicza 12, 00-944 Warsaw, Poland e-mail: case@case.com.pl http://www.case.com.pl
- 3. 3 The Episodes of Currency Crisis in Latin... Contents Introduction by Marek D¹browski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7 Part I. The Mexican Peso Crisis 1994–1995 by Wojciech Paczyñski . . . . . . . . . . . . . . . . . . . . . . . . . . .9 1.1. History of the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .9 1.2. In Search of the Causes of the Crisis: Macroeconomic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 1.2.1. Fiscal Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 1.2.2. Savings and Investment Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 1.2.3. Private and Public Debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 1.2.4. Exchange Rate Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14 1.2.5. Monetary Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 1.2.6. Foreign Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 1.2.7. Balance of Payments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 1.3. In Search of the Causes of the Crisis: Microeconomic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 1.4. Political Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 1.5. Crisis Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 1.6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19 Appendix: Chronology of the Mexican Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .20 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 Part II. The 1995 Currency Crisis in Argentina by Ma³gorzata Jakubiak . . . . . . . . . . . . . . . . . . . . . .23 2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 2.2. Overview of Economic Situation Before and During the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 2.2.1. Reforms of the Early 1990s and Crisis Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 2.2.2. Monetary and Exchange Rate Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24 2.2.3. Fiscal Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 2.2.4. Private and Public Debt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 2.2.5. Savings and Investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .26 2.2.6. Foreign Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27 2.2.7. Balance of Payments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27 2.2.8. Real and Nominal Rigidities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 2.2.9. Banking System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 2.2.10. Domestic Financial Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .31 2.1.11. Private and Public Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 2.3. Political Situation and Management of the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 2.4. Post-Crisis Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 2.5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 Appendix: Chronology of the Argentinian Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37 Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37 CASE Reports No. 39
- 4. 4 Marek D¹browski (ed.) Part III. The 1997 Currency Crisis in Thailand by Ma³gorzata Antczak . . . . . . . . . . . . . . . . . . . . . . .39 3..1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39 3.2. The Way to the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39 3.2.1. Macroeconoomic Signs of Vulnerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .41 3.2.2. Microeconoomic Signs of Vulnerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42 3.3. The Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44 3.3.1. Managing the Crisis. The IMF Intervention in Thailand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .44 3.3.2. Macroeconomic Environment after the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 3.4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47 Appendix: Chronology of the Thailand's Currency Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54 Part IV. The Malaysian Currency Crisis, 1997–1998 by Marcin Sasin . . . . . . . . . . . . . . . . . . . . . . . . .55 4.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 4.1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .55 4.1.2. The Public Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56 4.1.3. Monetary Policy and the Financial Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 4.1.4. The Corporate Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61 4.1.5. The External Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62 4.2. The Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 4.2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 4.2.2. Malaysian Vulnerability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .65 4.2.3. Crisis Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 4.3. Response to the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 4.3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 4.3.2. Fiscal Policy Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 4.3.3. Monetary Policy Response and Capital Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 4.3.4. Financial and Corporate Sector Restructuring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .73 4.4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .74 Appendix: Chronology of the Malaysian 1997–1998 Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .76 Part V. The Indonesian Currency Crisis, 1997–1998 by Marcin Sasin . . . . . . . . . . . . . . . . . . . . . . . . .77 5.1. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77 5.1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77 5.1.2. Monetary Policy and the Financial Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .78 5.1.3. The External Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .81 5.1.4. The Public Sector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 5.2. The Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 5.2.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84 5.2.2. Indonesia's Vulnerability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 5.2.3. Crisis Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 5.3. Response to the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 5.3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 5.3.2. Monetary Policy Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 CASE Reports No. 39
- 5. 5 The Episodes of Currency Crisis in Latin... 5.3.3. Fiscal Policy Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 5.3.4. Banking System and Debt Restructuring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 5.3.5. Prospects for the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 5.4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Appendix: The Chronology of the Indonesian Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .100 Part VI. The South Korean Currency Crisis, 1997–1998 by Monika B³aszkiewicz . . . . . . . . . . . . . .101 6.1. Was Korea Different? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 6.1.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 6.1.2. Background to the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 6.1.3. Signs of Vulnerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 6.2. The Role of Chaebols in the Future Development of Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .104 6.2.1. Debt Financing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105 6.2.2. Investments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .106 6.3. Korean Financial System and its Liberalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 6.3.1. Non-banking Financial Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 6.3.2. Capital Account Liberalization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109 6.3.3. Credit Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111 6.3.4. Risk Assessment in the Banking System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 6.4. The Onset of the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 6.5. The 1998 Recession and 1999 Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 6.5.1. The IMF Intervention in Asia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 6.5.2. Macroeconomic Environment after the Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 6.6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 Appendixes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Appendix 1: 30 Largest Cheabols: April 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Appendix 2: Foreign Capital Controls in Korea, June 1996 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 Appendix 3: Chronology of the Korean Crisis 1997 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 Appendix 4: Banking System and Corporate Restucturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .122 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 Comments to Papers on Asian Crises by Jerzy Pruski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 CASE Reports No. 39
- 6. 6 Marek D¹browski (ed.) CASE Reports No. 39 Ma³gorzata Antczak Ma³gorzata Antczak is an economist at the Centre for Social and Economic Research. After she graduated from the Department of Economics at the Warsaw University in 1994, she joined the CASE Foundation. She works in the fields of macroeconomics in transition economies and social policy issues. The research activity included also education issue and it's relationship to labour market in Poland Monika B³aszkiewicz Economist, Ministry of Finance The author received MA in International Economics from the University of Sussex in January 2000. Presently, Monika B³aszkiewicz works for the Ministry of Finance at the Department of Financial Policy, Analysis and Statistics. Her main inter-est lies in short-term capital flows to developing and emerging market economies and the role this kind of capital plays in the process of development and integration with the global economy. In every day work she deals with the problem relat-ed to Polish integration with EU, in particular in the area of Economic and Monetary Union. Marek D¹browski Marek D¹browski, Professor of Economics, V-Chairman and one of the founders of the CASE – Center for Social and Economic Research in Warsaw; Director of the USAID Ukraine Macroeconomic Policy Program in Kiev carried out by CASE; from 1991 involved in policy advising for governments and central banks of Russia, Ukraine, Kyrgyzstan, Kazakhstan, Georgia, Uzbekistan, Mongolia, and Romania; 1989–1990 First Deputy Minister of Finance of Poland; 1991–1993 Member of the Sejm (lower house of the Polish Parliament); 1991–1996 Chairman of the Council of Ownership Changes, the advi-sory body to the Prime Minister of Poland; 1994–1995 visiting consultant of the World Bank, Policy Research Department; from 1998 Member of the Monetary Policy Council of the National Bank of Poland. Recently his area of research interest is concentrated on macroeconomic policy problems and political economy of transition. Ma³gorzata Jakubiak Ma³gorzata Jakubiak has collaborated with the CASE Foundation since 1997. She graduated from the University of Sus-sex (UK; 1997) and the Department of Economics at the University of Warsaw (1998). Her main areas of interest include foreign trade and macroeconomics of open economy. She has published articles on trade flows, exchange rates, savings and investments in Poland and other CEE countries. During 2000-2001 she was work-ing at the CASE mission in Ukraine as resident consultant. Wojciech Paczyñski Economist at the Centre for Eastern Studies, Ministry of Economy, Warsaw. Graduated from the University of Sussex (1998, MA in International Economics) and University of Warsaw (1999, MA in Economics; 2000, MSc in Mathematics). Since 1998 he has been working as an economist at the CES. In 2000 started co-operation with CASE. His research inter-ests, include economies in transition, political economy and game theory. Marcin Sasin Marcin Sasin has joined CASE Foundation in 2000. He is an economist specializing in international financial economics and monetary policy issues. He obtained Master of Science at the Catholic University of Leuven, Belgium in 2000. He also holds MA. in Oriental Studies at the Warsaw University.
- 7. 7 The Episodes of Currency Crisis in Latin... The decade of the 1990s brought a new experience with financial instability. While earlier currency crises were caused mainly by the evident macroeconomic mismanage-ment (what gave theoreticians an empirical ground for a construction of the so-called first generation models of cur-rency crises), during the last decade they also happened to economies enjoying a good reputation. This new experience started with the 1992 ERM crisis when the British pound and Italian lira were forced to be devalued. This was parti-cularly surprising in the case of the UK, the country, which went successfully through series of very ambitious econo-mic reforms in the 1980s. At the end of 1994 the serious currency crisis hit Mexi-co, and during next few months it spread to other Latin American countries, particularly to Argentina (the so-called Tequila effect). Although Argentina managed to defend its currency board, the sudden outflow of capital and banking crisis caused a one-year recession. Currency crises have not been the new phenomena in the Western Hemisphere where many Latin American countries served through decades as the textbook examples of populist policies and economic mismanagement. However, two main victims of "Tequila" crisis – Mexico and Argentina – represented a pret-ty successful record of reforming their economies and expe-rienced turbulence seemed to be unjustified, at least at first sight. Two years later even more unexpected and surprising series of financial crises happened in South East Asia. The Asian Tigers enjoyed a reputation of fast growing, macro-economically balanced and highly competitive economies, which managed to make a great leap forward from the ca-tegory of low-income developing countries to middle or even higher-middle income group during life of one genera-tion. However, a more careful analysis as done in this vol-ume could easlly the specify several serious weaknesses, particularly related to financial and corporate sector. Addi-tio- nally, as in the case of Mexico, managing the crisis in its early stage was not specially successful and only provoked further devaluation pressure and financial market panic. The external consequences of the Asian crisis became much more serious than those of the Mexican crisis. While CASE Reports No. 39 the later had a regional character only, the former affected the whole global economy and spread through other conti-nents. The Asian crisis started in Thailand in July 1997 and its first round of contagion hit Malaysia, Indonesia and the Philippines in summer 1997. The next wave caused serious turbulence in Hong Kong, the Republic of Korea, and again in Indonesia in the fall of 1997 and beginning of 1998. Singa-pore and Taiwan were affected to lesser extent. Asian developments also undermined market confidence in other emerging markets, particularly in Russia and Ukraine expe-riencing the chronic fiscal imbalances. Both countries, after resisting several speculative attacks against their currencies in the end of 1997 and in the first half of 1998, finally entered the full-scale financial crisis in August – September 1998. Following Russia and Ukraine, also other post-Soviet economies experienced forced devaluation and debt crisis. This relates to Moldova, Georgia, Belarus, Kyrgyzstan, Uzbekistan, Kazakhstan, and Tajikistan. Finally, Russian developments triggered eruption of currency crisis in Brazil in early 1999, and some negative contagion effects for other Latin American economies, particularly for Argentina (1999–2000). In the meantime, cumulated negative consequences of the Asian and Russian crises damaged confidence not only in relation to the so-called emerging markets but also affected the financial markets of developed countries. In the last quarter of 1998 the danger of recession in the US and worldwide pushed the Federal Reserve Board to ease sig-nificantly its monetary policy. However, some symptoms of the global slowdown such as a substantial drop in prices of oil and other basic commodities could not be avoided. The new crisis episodes stimulated both theoretical discussion and large body of empirical analyzes trying to identify the causes of currency crises, their economic and social consequences, methods of preventing them and effective management when a crisis already happened. On the theoretical ground, the new experience brought the so-called second and third generation of currency crises models. Both theoretical and empirical discussion started to put attention on the role of market expectations and multiple equilibria. Introduction by Marek D¹browski
- 8. 8 Marek D¹browski (ed.) In some extreme interpretations the role of the so-called fundamentals, i.e. soundness of economic policy, started to be neglected in favor of the role of collective psy-chology of financial market players (multiple equilibria, herd behavior, market panic, contagion effect). However, as the detailed analysis of the crisis episodes shows it would be hard to find any convincing case of currency crisis in "inno-cent" country. Although the role of multiple equilibria cannot be questioned, they can trigger a crisis only when funda-mentals are under question. This is convincingly document-ed in all the country studies presented in this volume. The same type of conclusions can be derived from dis-cussion on the role of globalization. Although increasing integration of product and financial markets make all coun-tries more mutually dependent and vulnerable to the exter-nal shocks, globalization itself cannot be blamed for causing crisis in any particular country. This volume presents six monographs of currency crisis episodes in two Latin American countries in 1994–1995 (Mexico and Argentine) and four Asian countries in 1997–1998 (Thailand, Malaysia, Indonesia, and Korea). The Asian part of this volume is supplemented with a short com-parative note, commenting these four monographs. All the studies were prepared under the research pro-ject no. OI44/H02/99/17 on "Analysis of Currency Crises in Countries of Asia, Latin America and Central and Eastern Europe: Lessons for Poland and Other Transition Coun-tries", carried out by CASE and financed by the Committee for Scientific Research (KBN) in the years 1999–2001. They were subjects of public presentation and discussion during the seminar in Warsaw organized by CASE on December 21, 2000, under the same research project. In the all analyzed cases currency crises were accompa-nied by other signs of financial turbulence such as (public and/or private) debt crisis or banking crisis. However, the limited scope of the conducted analysis forced the research team to concentrate on the currency crises and refer to banking and debt crisis only as the background or conse-quence of the currency devaluation. This collection of papers will be followed by another volume presenting episodes of currency crises in the Euro-pean transition economies. CASE Reports No. 39
- 9. 9 The Episodes of Currency Crisis in Latin... Part I. The Mexican Peso Crisis 1994–1995 by Wojciech Paczyñski 1.1. History of the Crisis In order to put into context the developments that final-ly led to the currency crisis of 1994/1995, it is useful to go back as far as mid-1980's. In December 1987, a set of reform policies was initiated aimed at "remaking the Mexi-can economy" [Lustig, 1992]. The shift in policies was acce-lerated after 1988 as they gained support from the newly elected president Carlos Salinas de Gortari [DeLong et al., 1996]. The reform package was successful and brought macroeconomic stabilisation. Inflation was reduced from nearly 160% in 1987 to the range of 18% – 30% in 1989–1991 and further down to less than 12% in 1992 and 8.3% in 1993. At the same time economic growth resumed reaching 3.5–4.5% pa. in the period 1989–1992. This result was quite remarkable given the record of failed reform attempts in previous years [Blejer and del Castillo, 1996]. The 1989 foreign debt restructuring left Mexico with relatively low and mostly long-term foreign debt (it accounted for some 19% of GDP at the end of 1993) [Sachs et al., 1995]. Public debt was substantially reduced from 67% of GDP in 1989 to 30% in 1993. From December 1990 onwards, foreigners were allowed to purchase short-term government peso-denominated debt instruments [Gil- Diaz, 1998]. After the restructuring, Mexico once again gained access to international financial markets. Private ca-pital inflows surged to an average of above 6% of GDP in the period 1990–1993 [IMF, 1995]. Economic policies during the period 1990–1993 result-ed in the implementation of important structural reforms in various fields. The authorities undertook major domestic financial sector reform and capital account liberalisation [Otker and Pazarbasioglu, 1995] and privatisation. One should also note the improvement of the regulatory frame-work governing economic activity in many sectors, e.g. in tourism, means of transport, petrochemicals, electricity, telecommunications, etc. [WTO, 1997]. Another important factor was trade liberalisation. This process had started CASE Reports No. 39 much earlier. In 1985 Mexico formally joined the General Agreement on Tariffs and Trade (GATT). The next major step was the signing of the North American Free Trade Agreement (NAFTA) in 1992 that stipulated the reduction in non-tariff barriers, liberalisation of investment laws, changes in competition law, etc. The NAFTA finally took effect in January 1994. The government followed a path of budgetary discipline. The operational budget of the public sector [1] was in sur-plus in the range of 2–3% of GDP in the early 1990's. Sev-eral social pacts were concluded between the government and labour organisations as well as business representatives [Blejer and del Castillo, 1996]. Among the issues agreed upon was the exchange rate policy that became the central anti-inflationary instrument. The question whether the exchange rate policy was appropriate and whether it result-ed in an overvaluation of the peso is one of the major issues raised in all analyses of the currency crisis of 1994/1995. This problem will be discussed later. In 1993, the overall economic situation deteriorated slightly. GDP growth slowed to only 0.6% and private consumption and investment fell in real terms. These developments are mostly attributed to the ongoing restructuring in the manufacturing sector, a tightening of credit conditions by monetary authorities, and a credit squeeze resulting from the deterioration in the quality of banks' loan portfolio [IMF, 1995]. There was also some uncertainty about the approval of NAFTA, which was final-ly resolved in November. Despite these setbacks, until 1994 Mexico was widely regarded as an example of a successful economic reform story. Some other views [Dornbusch and Werner, 1994] appeared among economists, but were not picked up nor were they considered important by investors. Suddenly, in the course of 1994, several events took place that turned out to be of considerable importance for Mexico's econo-mic situation. January witnessed the peasant rebellion in Chiapas – the first one of political events of 1994 that later turned out to have a significant impact on financial stability of the country. [1] Operational balance is defined as primary balance plus the real portion of the interest paid on public debt.
- 10. 10 Marek D¹browski (ed.) CASE Reports No. 39 In February, U.S. interest rates started to rise. In March the candidate in presidential elections Luis D. Colosio was assassinated. This came as a shock to investors and led to severe financial turbulence. The peso exchange rate increased from the bottom of the intervention band [2], where it stayed before, to the ceiling of the band, which constituted a nominal devaluation of ca. 10%. This was accompanied by a decrease in Central Bank reserves of around 9 billion USD. The monetary authorities followed a path of a rather loose monetary policy, boosting credit to the economy in order to prevent interest rate increase and to support weak commercial banks. Also, in the run-up to the presidential elections (the output slowdown could had been another factor) fiscal policy became more expansion-ary. The actions involved some tax cuts and increases in social spending [IMF, 1995]. In August presidential elections took place that gave a victory to Ernesto Zedilo. His victory, with a higher than expected margin, was considered a positive event from the point of view of foreign investors even though the elections were not carried out in a perfect way. In September, the secretary general of the ruling party was assassinated. Higher domestic interest rates (around 16% pa. from April until July as opposed to around 10% pa. in the first quarter) and the approval of a 6.75 billion USD short term credit line from NAFTA partners helped to ease the pres-sures from financial markets. The peso exchange rate stayed near the ceiling of the band, outflow of capital was stopped and reserves remained relatively stable from April until October. After July, interest rates began even to decline. Another policy action implemented by the authorities in order to increase the credibility of maintaining the exchange rate rule and to prevent increases in interest rates was sub-stituting short term peso-denominated government debt (Cetes) with dollar-indexed (but payable in pesos) short term bonds (Tesobonos). This started after the March events and continued in the following months. The out-standing stock of Tesobonos increased significantly – from 14 billion pesos in March 1994 to 63.6 billion in November. The whole operation within a very short period dramatical-ly changed the composition of short-term debt held by the private sector. While in the first quarter the share of Tesobonos in total Cetes and Tesobonos stock did not exceeded 10%, it reached almost 60% in July. The current account continued to deteriorate in the third quarter of 1994 reaching a record level deficit of 7.9 billion USD. In addition, both the stock of Tesobonos and its share in total short-term debt increased further. Heightened concerns about the sustainability of Mexico's external posi-tion led to intensified capital outflows. The reserves declined by 4.7 billion USD between October and Novem-ber and further 2.5 billion USD to 10 billion USD in mid- December. On 1 December president Zedillo took office and two days later the unrest in Chiapas intensified. Given the current situation, the authorities decided on 20 Decem-ber to widen the exchange rate band by 15%. This move was accompanied by the announcement of the authorities to support the peso at a rate of around 4 pesos to the U.S. dollar. This announcement was, however, not perceived as credible by investors, who put further pressure on the exchange rate. The Bank of Mexico lost around 4 billion USD within two days and was forced to freely float the peso on 22 December. Inflation was certainly one of the most important prob-lems that the authorities had to tackle after the devaluation. It jumped to the level of around 8% monthly, but in the sec-ond half of the year was reduced to the range 2–4% month-ly. The peak of 12-month inflation was recorded in Decem-ber 1995, when it stood at 51.97%. During the first quarter of 1995, the peso depreciated at a rather high rate reaching 6.82 in the end of March. It then regained some strength fluctuating between 5.8 and 6.4 pesos to the dollar until September, to fell further in the last quarter to 7.64 in the end of December. The crisis also resulted in a severe recession with GDP falling by 9.2% YoY in the second quarter and respectively by 8.0% and 7.0% in the third and fourth quarter of 1995 [INEGI, 2000]. Industrial production dropped sharply and the unemployment rate increased. In the second part of the year, the first signs of economic recovery became visible. These trends intensified in the last quarter, and since the second quarter of 1996 the Mexican economy returned to a path of fast growth (YoY rate of GDP growth reached 7.2% in the second quarter of 1996). The severity of the 1995 recession was caused by several factors, the most important probably being very high interest rates (lending rate stayed close to 70% in the first half of the year) that were used as an anti-inflationary measure. Other factors included a signi-ficant drop in capital inflows, and sudden reduction in cred-it in the economy. Domestic consumption was further repressed due to debt overhang and possibly substantial negative income and wealth effects resulting from the deva-luation [SHCP, 1995]. Gruben et al. (1997) point at sectoral fragmentation of severity of recession and the timing and strength of a rebound. During 1995, significant adjustment took place in exter-nal position of Mexico. Exports surged by 30% in compari-son to 1994 and imports contracted by around 9%. As a result, the trade balance improved from a deficit of 18.5 bil-lion USD in 1994 to a surplus of 7 billion in 1995. The cur-rent account deficit contracted from nearly 30 billion USD in 1994 to only 1.5 billion USD. A very important achieve- [2] Since November 1991 Mexico operated a moving band exchange rate system. This is discussed in more detail in section 1.2.4.
- 11. 11 The Episodes of Currency Crisis in Latin... ment of the authorities was the elimination of the short-term debt overhang and consequently regaining access to international capital markets. In particular Tesobonos were practically eliminated from the short-term debt stock during 1995. An access to credits from the foreign financial support package played an important role in managing the debt problem. Mexico used close to 12 billion USD of IMF cre-dits in 1995 in addition to around 14 billion of other excep-tional financing. A much-improved economic condition allowed Mexico to pay back these credits, in some instances ahead of schedule. Since 1996 Mexico has experienced re-latively stable economic growth. 1.2. In Search of the Causes of the Crisis: Macroeconomic Factors 1.2.1. Fiscal Policy The role of fiscal policy in the peso crisis has not been emphasised in most of the studies. This is because in the early 1990's Mexico achieved remarkable successes in near-ly balancing the public finances. The general public sector deficit declined from around 16% of GDP in 1986 to about 2% in 1993. In 1994 the result was not much worse – the deficit reached around 3.9% of GDP [3]. This fiscal perfor-mance was to a large extent due to reduced interest pay-ments during the period. One important observation is that fiscal policy was not tightened, and thus was not used as a tool for dealing with negative shocks of 1994. On the con-trary, fiscal policy was rather looser in the election year. The role of quasi-fiscal operations via development bank credits in 1994 is not very clear. Very soon after the crisis, some authors presented the view that fiscal expansion CASE Reports No. 39 through this channel could had played some role in the mix of bad policies that were implemented in 1994 [World Bank, 1995]. Most of the analyses show however, that develop-ment banks' credit was not an important factor. Sachs et al. (1995) argue that most of the activities of these banks do not belong in an economically meaningful definition of a budget deficit. An innovative way of looking at the role of fiscal policy in explaining the crisis is proposed by Kalter and Ribas (1999). They point out the role of the increasing magni-tude of go-vernment operations, rather than the size of government deficit, in affecting the relative price of traded to non-traded goods (i.e. the real exchange rate), the financial condition of the traded sector, and interest rates. They argue that the significant rise in government non-oil revenue collections measured in U.S. dollars or in terms of traded goods prices has had an effect on the tradable sec-tor analogous to that of a surge in export commodity prices (Dutch disease). The resulting deterioration of finances of traded goods sector was then passed to com-mercial banks' finances. While the arguments used by Kalter and Ribas (1999) are interesting and certainly add another dimension to the understanding of fundamental reasons behind the crisis, they do not provide an explana-tion for sudden events of December 1994. 1.2.2. Savings and Investment Balance In the period 1988–1994 Mexico witnessed a notice-able growth in investment and a decline in savings (see Table 1-2). Overall investment grew from 20.4% of GDP in 1988 to 23.6% in 1994. Interestingly, public sector investments remained relatively stable and were even reduced, while the growth was due to the private invest- Table 1-1. Public sector balances 1986–1994 (in percent of GDP) Financial balance Primary balance Operational balance 1986 -16.1 3.7 -2.4 1990 -3.3 7.6 1.8 1991 -1.5 5.3 2.9 1992 0.5 6.6 2.9 1993 -2.1 3.6 2.1 1994 -3.9 2.3 0.5 Notes: Financial Balance includes all public sector borrowing requirements. Primary Balance is defined as Financial Balance less interest paid on public debt. Operational Balance is defined as Primary Balance plus the real portion of the interest paid on public debt. Other sources provide slightly different data. Source: Sachs et al. (1995). [3] There is no consensus about the size of public sector surplus or deficit. Different authors use distinct measures. For example, Kalter and Ribas (1999) estimate the overall public sector deficit close to 1% in 1992–1993 and close to 2.5% in 1994.
- 12. 12 Marek D¹browski (ed.) Current account 1988 1.4 17.6 19 5 1 .4 520.4 -3.6 2.2 -1.4 1989 3.1 15.6 18.7 4.8 16.5 21.3 -1.7 -0.9 -2.6 1990 6.7 12.5 19.2 4.9 17 21.9 1.8 -4.5 -2.7 1991 7.5 10.3 17.8 4.6 17.8 22.4 2.9 -7.5 -4.6 1992 7.1 9.5 16.6 4.2 19.1 23.3 2.9 -9.6 -6.7 1993 6.3 8.9 15.2 4.2 17.8 22 2.1-8. 9 -6.8 1994 5 10.7 15.7 4.5 19.1 23.6 0.5 -8.4 -7.9 Source: Sachs et al. (1995) CASE Reports No. 39 Table 1-2. Saving and investment levels 1988–1994 (in percent of GDP) Saving Investment Net saving Public Private Total Public Private Total Public Private ment boom. This was accompanied by an even more apparent reduction in propensity to save. Total savings fell from 19% of GDP in 1988 to only 15.2% of GDP in 1993 and 15.7% in 1994. Private savings declined from 17.6% of GDP in 1988 to 8.9% in 1993 before starting to grow, albeit modestly, in 1994. This leads to the conclusion that the deterioration in the current account – the deficit reached 6.8% of GDP in 1993 and 7.9% in 1994 – was primarily caused by the level of private savings not match-ing the level of private investment [4]. 1.2.3. Private and Public Debt The role of Mexico's indebtedness in provoking the financial crisis deserves a more detailed analysis. First, it should be noted that Mexico had a history of problems asso-ciated with its foreign debt, including the crisis of 1982. In contrast to the past, the beginning of the 1990's was marked by a very significant improvement in this sphere. Public debt was reduced from some 64% of GDP in 1989 to 35% of GDP in 1993. Of this, 23% of GDP was foreign debt that as a result of 1989 restructuring had a favourable maturity structure (long term liabilities prevailed). Domestic debt accounted for 12% of GDP [IMF, 1995] [5]. These numbers were low in comparison to other developing and developed countries. Moreover, standard debt indicators such as the ratio of total debt to GDP or to exports or the ratio of inte-rests on debt to GDP or exports were improving in the early 90's. This clearly shows that the overall level of public debt did not play a big role in the loss of investors' confi-dence in 1994. What did matter, however, was the maturity and cur-rency structure of the domestic debt and the level of pri-vate borrowing. From 1989 to 1992, net credit to the private sector from the financial system expanded at an average annual rate of 66% in nominal terms, offsetting the decline in borrowing of the public sector resulting from the substantially improved fiscal position [IMF, 1995]. In 1993 net domestic credit of the banking system continued to expand at an annual rate of around 20%. Interestingly, this decomposes into a substantial reduc-tion in credit to the public sector (around 30%) and an expansion of credit to the private sector of the same magnitude (see table 1-3). This trend continued through the first half of 1994, while in the second half public sec-tor borrowing also started to rise, bringing the 12-month rate of growth of net domestic credit from the banking system to around 30%. It is also worth noting the faster expansion of credit to private sector from development banks than from commercial banks in 1994. Other inter-esting statistics are presen-ted in Gil-Diaz (1998) [6] which indicate that in the period from December 1988 to November 1994 credit card liabilities rose at an aver-age rate of 31% per year, direct credit for consumer durables rose at a yearly rate of 67% and mortgage loans at an annual rate of 47%, all in real terms. The above numbers, along with numbers cited in sec-tion 1.2.2 with regard to private investment and savings, show that it was mostly private sector borrowing that brought the current account deficit to the levels it reached in 1993 and 1994 (more than 6% of GDP). Such a level of the deficit seems quite high but was nevertheless easily financed in 1993, and from that perspective there were [4] Sources differ in calculations of the current account deficit in relation to GDP. For example Gurrha (2000) citing official Mexican data estimates the deficit to account for 5.8% of GDP in 1993 and 7.0% in 1994. [5] Also in this case different numbers are cited by other authors. For example Sachs et al. (1995) estimate the public debt at 67% of GDP in 1989 and 30% in 1993. According to this source this last number can be broken down to 19% of GDP of foreign debt and 11% of domestic debt. The ave-rage maturity of domestic debt was around 200 days. [6] The statistics were provided to the author (a former Vice Chancellor of the Bank of Mexico) by the Economic Research Department of Bank of Mexico.
- 13. 13 The Episodes of Currency Crisis in Latin... Table 1-3. Monetary sector – the expansion of credit 1990–1994 (twelve-month rates of growth, end of period) 1990 1991 1992 1993 1994 Broad money (M4) 46.4 30.9 19.9 25.0 17.1 Net domestic assets of the financial system 22.6 31.6 20.8 15.5 32.0 Net credit to public sector 3.9 -1.6 -31.7 -46.2 25.3 Net credit to private sector 63.2 53.3 57.126. 4 31.9 Net domestic credit of commercial 49.4 48.6 20.9 24.3 … banks Net domestic credit of development banks -10.0 18.8 23.4 47.4 42.2 Source: IMF (1995). reasons to believe that this could had happen again in 1994. Firstly, one should note that large capital inflows that resulted from markedly improved perception of the econ-omy by foreign investors and that were possible thanks to capital account liberalisation of 1990 were transferred to a significant increase of short term debt. The inflow was sterilised by issuing short-term government debt instru-ments (Cetes). As a result short-term indebtedness increased significantly with short term debt to total debt stock ratio rising from 21.9 in 1992 to 28.1 in 1994 [World Bank, 2000]. At the end of 1993 the value of Cetes alone reached 22.9 billion USD, i.e. it was very close to net international reserves of the Bank of Mexico (24.9 billion USD) (see Figure 1-1). This placed Mexico in a potentially vulnerable position. On top of that, during 1994 the cur-rency structure of short-term debt underwent a substan-tial change. After the March assassination and resulting turbulence in the financial markets, the authorities started exchanging peso-denominated bonds (Cetes) with dollar-indexed debt instruments (Tesobonos). This action was aimed at uphold-ing the investors confidence in the exchange rate regime after the peso depreciated by around 10% reaching the ceiling of the intervention band. It was also used as a tool to avoid further increases in interest rates. Werner (1996) estimates that the substitution from Cetes to Tesobonos was equivalent to an interest rates increase of around 8 to 11 percentage points. He argues that accounting for the currency composition of government debt gives more appropriate measures of currency risk premium in the period before the crisis. The scale of substitution from Cetes to Tesobonos was huge. By June 1994, the amount of Tesobonos and Cetes were roughly equal and in Decem-ber the amount of Tesobonos was around 5.5 times that of Figure 1-1. The composition of short term government debt (USD million) 30000 25000 20000 15000 10000 5000 0 1993M12 1994M6 1994M5 CASE Reports No. 39 1994M8 1994M7 1994M11 1994M9 1994M10 1994M12 1995M2 1995M1 1995M4 1995M3 Cetes Tesobonos BoM reserves Notes: Cetes – three month peso-denominated government debt. Tesobonos – three month dollar-indexed government debt. Source: own calculations based on IMF, IFS and World Bank (1995) data.
- 14. 14 Marek D¹browski (ed.) 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 Cetes. Another way to look at the process is to note that around 15 billion USD of private sector holdings in Cetes were swapped for Tesobonos from March till November. After the devaluation on 20 December, government borrowing was clearly not sustainable. Investors rushed to withdraw their investments and the government found itself unable to cover short term liabilities that led to a panic and the severe currency devaluation. Eventually, only a huge international support package helped to solve the problem. The shift to dollar denominated short-term pub-lic debt certainly contributed to the whole set of factors that provoked the crisis. Interestingly, however, risks asso-ciated with rapidly growing short-term dollar indexed debt of Mexico seem to had been underestimated, not to say unnoticed, by the international financial community until the devaluation took place. Sachs et al. (1996) present puzzling statistics on international press coverage of Mexi-co. The issue of Tesobonos was completely ignored by leading international financial papers with only one article mentioning it being published before December 1994 [7]. The problem of accumulated dollar denominated debt accompanied by depleted foreign reserves constituted an important factor in provoking the panic after the announcement of devaluation on 20 December (Sachs et al., 1995, see also section 1.2.5). 1.2.4. Exchange Rate Policy In the last decades, Mexico altered its exchange rate po-licy several times and had a history of several episodes of sig-nificant devaluations (e.g. 1976, 1982, and 1985). A fixed exchange rate regime that was introduced in 1988 and later corrected on several occasions played a major role in the anti-inflationary strategy of Mexican authorities. From Janu-ary 1989 until November 1991, a preannounced crawling peg was in operation (with two reductions of the rate of crawl), and from 11 November 1991 an exchange rate intervention band was used with several changes in the rate of crawl of both upper and lower bands. From 1991 until November 1994, the peso steadily and very slowly depre-ciated in nominal terms. Yet since the level of inflation was much higher than in the U.S., in real terms the peso appre-ciated by around 15% if consumer price indexes for Mexi-co and the U.S. are applied or close to 21% if the compari-son is based on wholesale price indexes [8]. The question whether the peso was overvalued at that time brought much attention, especially after the December crisis. One should note, however, that there were voices pointing to an overvaluation of the peso and the possible risks that it posed already in early 1994, the best known being Dornbusch and Werner (1994). In most analyses of the [7] The authors surveyed the Financial Times, the New York Times, and the Wall Street Journal. The number of such articles jumped to 6 in Decem-ber 1994 and 46 in January 1995. [8] In a new study Dabos and Juan-Ramon (2000) estimate the model of the real exchange rate in Mexico. Their results suggest that on the eve of the crisis the peso was overvalued by a number in the range of 12 to 25 percent. This result is consistent with the majority of previous studies. CASE Reports No. 39 Figure 1-2. Real exchange rate movements 1975–1995 0 RER_cons RER_prod 1975M1 1976M1 1977M1 1978M1 1979M1 1980M1 1981M1 1982M1 1983M1 1984M1 1985M1 1986M1 1987M1 1988M1 1989M1 1990M1 1991M1 1992M1 1993M1 1994M1 1995M1 Note: Rer_cons index is obtained using consumer price indexes in Mexico and the U.S., while Rer_prod is based on producer price indexes; 1 is an average value of respective indexes over the period. Source: Author's calculation based on IMF, IFS data
- 15. 15 The Episodes of Currency Crisis in Latin... crisis that appeared after 1994, the view that the peso has indeed been overvalued seems to gain rather strong support [World Bank, 1995]. The standard reasoning points to the fact that the exchange rate-based stabilisation under capital mobility has led to a large current account deficit and real appreciation of the peso that at some point became unsus-tainable and the correction of real exchange rate was need-ed [IMF, 1995]. It is, however, not clear whether this has played an important role in determining the crisis. In partic-ular, some authors concluded that the peso overvaluation is not at all useful in explaining the crisis [Gil-Diaz, 1998]. Also, as Sachs et al. (1996) point out, significant reduction in infla-tion in 1994 and 10% nominal depreciation from March to April 1994 certainly diminished the overvaluation problem. On 20 December, the upper limit of the intervention corridor was widened by 15%, but at that time the devalu-ation of that scale was widely regarded as insufficient. On the other hand, it undermined the confidence in the will and ability of the Mexican authorities to uphold the announced exchange rate policy. The continued pressure and a lack of possibilities to support the peso forced the authorities to let the peso flow freely on 22 December. In later months, the peso depreciated sharply hitting the rate of 6.82 pesos to the dollar in the end of March 1995, i.e. nearly twice as much as before 20 December. It recovered slightly in the next few months. 1.2.5. Monetary Policy During the whole of 1993, disinflation continued and interest rates on short term government papers were on a downward trend. Interest rates declined from around 17% at the beginning of 1993 to 13–14% in November. The approval of NAFTA in that month allowed for a further sig-nificant decrease below 10% in February and March 1994. Political turbulence at the end of March resulted in a surge in interest rates that stayed in the 16%–17% range from April until July. From August onwards, interest rates started to fall again and remained stable at around 13.7% from Sep-tember until November. Exactly the same pattern was fol-lowed by the real interest rates differential (i.e. real rate on Mexican papers compared to real rates on American Trea-sury bills). As Sachs et al. (1996) point out, such a behaviour of interest rates is markedly different from the one predicted by standard first generation crisis models (e.g. Krugman, 1979). This fact is presented as a main argument against the hypothesis that a speculative attack can be a mechanism used to describe the peso crisis. This point is perhaps some-what weakened when one takes into account interest rates differential adjusted using measures of currency structure of short term public debt [Werner, 1996]. One of the primary motives for substituting Cetes with Tesobonos was to avoid an adjustment via higher interest rates. This policy proved to be rather short sighted as dollar-indexed short-term debt very quickly reached high levels (see section 1.2.3). The policy of keeping interest rates low had imme-diate implications for the level of foreign exchange reserves in 1994. This basic yet important point is stressed by Sachs et al. (1995). Until March 1994, the Mexican private sector was selling securities to foreign investors at a rate that can roughly be estimated at around 20 billion USD yearly. This capital inflow financed Figure 1-3. Interest rates and inflation 1993–1995 80 70 60 50 40 30 20 10 0 -10 -20 CASE Reports No. 39 Cetes rate Real interest rate differential (Cetes vs. US T-Bills) Inflation (12month) 1993M1 1993M3 1993M6 1993M9 1993M12 1994M3 1994M6 1994M9 1994M12 1995M3 1995M6 1995M9 1995M12 Note: all numbers are percent per annum. Source: author's calculations based on IMF, IFS data.
- 16. 16 Marek D¹browski (ed.) CASE Reports No. 39 Figure 1-4. Components of monetary base January 1993 – May 1995 (millions of pesos) 120000 100000 80000 60000 40000 20000 0 -20000 -40000 the current account deficit. After March, interest rates demanded by foreign investors increased significantly yet the monetary authorities responded by trying to fix inter-est rates using credit expansion. The Central Bank simply offered to buy securities accepting low interest rates. This shows up in the Bank of Mexico accounts as domes-tic credit expansion to both private sector (mainly banks) and the government (mainly Tesobonos purchased from private investors). Such a behaviour did not provide any incentive to reduce the current account deficit and left no other way but to finance it from foreign reserves. What actually happened was that credits (issued in pesos) were converted into dollars to cover the trade deficit at the fixed exchange rate. Another way of looking at the mechanism is to note the identity decomposing the change in monetary base into the change of domestic credit and the change in reserves. With the monetary base being relatively constant, the expansion of domestic credit was mirrored by declining reserves (see Figure 1-4). The most common way of defending the policies of the central bank was to say that without providing credit, interest rates would have risen to levels that would se-riously affect the economy, and that the Central Bank was forced to act as a lender of last resort to commercial banks [Sachs et al., 1995; Gil-Diaz, 1998]. Carstens and Werner (1999) argue that, "in the case of Mexico during 1994 mo-netary policy had to defend the predetermined exchange rate, without affecting a weak banking system". Sachs et al. (1995) recommend, that the credit should be expanded moderately, while indeed the exchange rate should be allowed to depreciate. Still, some interest rates hike with all the adverse effects on economic growth seems to had been necessary anyway (and such a solution would proba-bly be less painful than the adjustment through a crisis). 1.2.6. Foreign trade Mexico experienced a substantial increase in private spending and trade deficits in 1988–1994. One should note that this kind of experience is similar to the one of many other countries that have undertaken exchange rate based stabilisa-tion programs. The trade deficit almost reached 16 billion USD in 1992, was somewhat reduced in 1993 and again rose to 18.4 billion USD in 1994. A deficit of that magnitude did not reflect weak performance of Mexican exports, which were growing at an average annual rate of above 10% between 1990 and 1994. The prospects for Mexican exports seemed to be very promising, especially after the final approval of the NAFTA in November 1993. One should note at that point that the U.S. was by far the most important tra-ding partner, accounting for more than 81% of exports and more than 71% of Mexican imports already in 1992. These shares have increased yet further in later years. 1.2.7. Balance of Payments The current account balance that was in surplus in 1987 soon turned to negative numbers. The size of deficits increased significantly after 1990. In 1991 it accounted for 4.6% of GDP, in 1992 and 1993 stayed at about 6.5% to widen still further to around 8% of GDP in 1994. As shown in section 2.2 this was primarily caused by private sector invest-ment exceeding its savings rather than imprudent fiscal poli- -60000 Monetary base International reserve Net domestic credit 1993M3 1993M5 1993M7 1993M9 1993M11 1994M1 1994M3 1994M5 1994M7 1994M9 1994M11 1995M1 1995M3 1995M5 Source: World Bank (1995)
- 17. 17 The Episodes of Currency Crisis in Latin... cies. The deficit was financed by high inflows of foreign capital to the private sector, majority of which was portfolio invest-ment. With capital inflows higher than the level of the current account deficit central bank's foreign currency reserves were gradually increasing from 6 billion USD in 1988 to 25.4 billion USD in 1993. In order to sterilise capital inflows the govern-ment issued large amounts of short-term peso- denominated treasury bills (Cetes) (see also section 1.2.3). The situation changed markedly after March 1994. The inflow of foreign capital fell abruptly. The capital account position from the balance of payment deteriorated from 11.8 billion USD in the first quarter to 3.7 billion in the sec-ond quarter. In turn, central bank's reserves fell from 29.3 billion USD at the end of February to 17.7 billion at the end of April, i.e. by 11.6 billion USD. The reserves remained rather stable at that level until November, when the next wave of reserve erosion took place. At the end of Novem-ber they stayed at only 12.9 billion USD. This provoked the final speculative attack against the peso. 1.3. In Search for the Causes of the Crisis: Microeconomic Factors In the years leading to the crisis major positive changes took place in the real sector environment. The economic program that was implemented starting in the late 1980s included several structural reforms. Substantial deregulation and privatisation took place along with trade liberalisation [Martinez, 1998]. The Mexican privatisation program was one of the most comprehensive in the world in terms of both the size and the number of companies privatised [La Porta and Lopez-de-Silane, 1999]. Privatisation was most intensive in the period 1989–1992 and by 1992 the govern-ment had withdrawn from most sectors of the economy with the exception of oil, petrochemicals and the provision of key infrastructure services. This constituted a major change to the situation from the early 1980s when the state was intensely involved in the economy through more than a thousand state-owned enterprises. The financial system, that until late 1988 was highly regu-lated, also underwent a quick and substantial liberalisation [Gelos and Werner, 1999]. All these factors contributed to a major improvement in perceived prospects of the Mexican economy and consequently, given the situation in world finan-cial markets, resulted in large capital inflows to Mexico in the period 1990–1993. There is no general consensus concerning the role of financial and real sector weaknesses in the peso crisis. Also, this channel is not very often thoroughly analysed, perhaps due to limited access to relevant data. It is clear that Mexico experienced a rapid expansion of credit to the private sector (see section 1.2.3). It is likely CASE Reports No. 39 that this was associated with poor screening of borrowers, and consequently, declining quality of credit [cf. Edwards, 1999]. Such a process is not unique to Mexico. Lidgren et al. (1996) highlight the problem of a lack of necessary credit evaluation skills in formerly regulated banks that are there-fore unable to use newly available resources more efficient-ly. Also, the notion that banks problems often precedes the financial crisis (devaluation) has strong support from other cross-country analyses [e.g. Kaminsky and Reinhart, 1996; Lidgren et al., 1996]. Gil-Diaz (1998) points to several caus-es of the rapid debt increase, the speed of which over-whelmed supervisors, e.g.: – speedy and not always well prepared privatisation of banks, sometimes with no respect to "fit and proper" criteria, either in the selection of new shareholders or top officers, – lack of proper capitalisation of some privatised banks and involvement in reciprocal leverage schemes, – lack of capitalisation rules based on market risk; this encouraged asset-liability mismatches that in turn led to a highly liquid liability structure, – loss of human capital in banks during the years when they were under the government; banking supervision capacity not meeting the requirements of increases in banks' portfolios. 1.4. Political Situation The Mexican peso crisis provides a clear and very inte-resting example of how political factors can contribute to a financial turbulence. The series of unexpected events in po-litics had a visible influence on the behaviour of economic aggregates and certainly played an important, through hard to measure, role in triggering the December crisis. The first of the series of events started on New Year's Day 1994 when peasants in the southern Mexican state of Chiapas began a rebellion by taking over six towns. Even though the uprising was rather quickly suppressed, it remained an issue in internal political life and occasionally flared up again. Especially before the August elections, the rebellion was again discussed and was recalled to question the extent of popular support for the government's eco-nomic program. On 23 March, Luis Donaldo Colosio, the presidential candidate of the ruling party, was assassinated. The causes of this murder were never actually revealed, and there were signs that it might had been associated with ten-sions within the ruling Institutional Revolutionary Party (PRI) [World Bank, 1995]. It should be noted that the PRI gov-erned Mexico since 1929 as a party organised around well connected political families. The assassination brought se-rious turbulence to the financial markets with the peso at the ceiling of the intervention band and sharply higher inte-
- 18. 18 Marek D¹browski (ed.) CASE Reports No. 39 rest rates. Improper response of authorities to the turmoil in the end of March and in April set the stage for the December crisis [Sachs et al., 1995]. The situation seemed to have calmed down when another candidate of the PRI, Ernesto Zedillo won the pre-sidential elections on 21 August. He received above 50% of votes that came as a surprise to many observers. It is indeed hard to verify whether the elections were free of fraud. On the other hand, 1994 elections were perhaps more demo-cratic and fair than many previous elections in Mexico. The outcome of the elections was generally considered positive from the financial stability point of view but may also have caused the authorities to believe that the worst of the insta-bility was over. One month after the elections, on 28 Sep-tember, PRI leader Jose Francisco Ruiz Massieu was assassi-nated. This murder remained a mystery too with several high officials of PRI possibly somehow implicated. All these developments certainly changed the position of Mexico's traditional ruling party. The process of reform in the political scene actually began slightly earlier, and, as Dornbusch and Werner (1994) point out, the rapid embrace of greater openness has shattered the PRI coherence, so that the old corporatism became unmanageable. It is thus clear that during the whole 1994 there were serious ten-sions within the ruling elite and the uncertainty about Me-xico's political future was a factor in the foreign perception of the country's financial stability. It should also be noted that the years of presidential elections have traditionally been associated with financial turbulence. This fact was recalled in many analyses of the 2000 presidential elections. Also, as many authors agree the long period between the voting and taking the office by the president elect has a negative impact on the quality of go-verning the country. President Zedillo took office on 1 December and it was followed by the intensified unrest in Chiapas. 1.5. Crisis Management While the policy mistakes of most of 1994 played an important role in triggering the December crisis, improper handling of the initial devaluation on 20 December probably exacerbated the crisis. The devaluation on 20 December was announced after weeks of assurances that the government was committed to the previous exchange rate system. The announcement was made by the Finance Minister on radio and television rather than through an official channel. As the World Bank (1995) stresses, such a way of publicising such a major policy change angered investors. Also, it turned out that business leaders were consulted before the devaluation, thus having the opportunity to make profits at the expense of unin-formed foreign investors [Krugman, 1997]. The devaluation was widely considered insufficient and the exchange rate immediately depreciated to the ceiling of the band, i.e. by 15%. The authorities had sacrificed the credibility without satisfying market expectations. The rush out of the country continued on the next day with Central Bank's reserves reportedly falling below 6 billion USD [World Bank, 1995]. President Zedillo affirmed the commitment to the new band, but on the next morning the government let the peso float. During the day it depreciated by a further 15%. On 26 December the planned press conference by the Finance Minister on the government anti-crisis plan was cancelled at the last moment. On the next day the peso depreciated to 5.45 pesos to the dollar. The auction of dollar denominated government bonds attracted almost no bids. Increasing prices made labour leaders to demand wage negotiations. On 29 December a new Finance Minister, Guillermo Ortiz Martinez was appointed, who within a few days announced a new economic program. On 3 January, Stanley Fisher, Acting Managing Director of the International Monetary Fund, made a statement expressing the Fund's support for this program and announcing the establishment of the Exchange Stabilisation Fund of 18 billion USD with contributions under the NAFTA from the monetary authorities of other major countries as well as from private investors. The talks on a stand-by credit from the IMF started a few days later and an 18-month credit of 17.8 billion USD was finally approved on 1 February. The Mexican authorities' program constituted of three main components: minimising the inflationary pressures of the devaluation, pushing forward structural reforms to sup-port and promote competitiveness of the private sector, and to address short term concerns of investors and establish a coherent floating exchange regime. To the end of the first objective, a National Accord was set among workers, busi-ness and government to prevent wages and prices hikes, the government spending were to be reduced by 1.3% of GDP, and cuts in credits from state development banks were to be implemented. In terms of structural reforms President Zedillo pledged to propose amendments to the constitution allowing for pri-vate investment in railroads and satellites, to open the telecommunication sector to competition, and to increase foreign participation in the banking sector. As the third objective is concerned, the co-operation with investment banks in order to address the issue of Tesobonos was announced, as well as creating a futures market in pesos, and commitment to a tight monetary policy. In the beginning of March, the finance minister announced a package of fur-ther measures aimed at strengthening the program. These included substantial increases in prices charged by public enterprises, VAT rate hike, and public expenditure reduc-tions, and were designed to allow the public sector to stay
- 19. 19 The Episodes of Currency Crisis in Latin... in surplus in 1995. Also, a further reduction of development banks was declared. The Mexican Rescue Plan was supported with what was then the biggest ever financial support package. The initial amount of 18 billion USD announced by Fisher at the begin-ning of January after further intense discussions rose to the range of 25–40 billion USD in mid- January and finally around 52 billion USD of loan guarantees and credits at the end of February. This package included 20 billion USD of loan guarantees from the U.S. government, 17.8 billion USD stand-by credit from the IMF (by-then the largest ever financial package approved by the IMF for a member coun-try both in terms of the amount and the overall percentage of quota, 688.4%), 10 billion USD from central banks via the BIS, and several billion dollars from other American gov-ernments [World Bank, 1995]. The unprecedented size of the support package brought about several controversies, especially in the U.S. The Clin-ton Administration was criticised heavily both for lack of action before the crisis, and for too much engagement in co-ordinating the support package. On 29 March 1995, Undersecretary of the Treasury, Lawrence Summers defended in closed hearings in the Senate the Administra-tion's failure to publicise a warnings on the situation in Me-xico. He admitted that the U.S Treasury lost the confidence in the peso before the dramatic devaluation took place but did not want to set off a market run on Mexico by making a public statement about the situation [Burkart, 1995]. The main arguments backing the support package pointed to the fact that Mexican economy was illiquid rather than funda-mentally insolvent [DeLong et al., 1996]. In retrospect it seems that this view was indeed right, even though the U.S. engagement in the package was to a large extent motivated politically, i.e. by fears of possible political destabilisation in the neighbouring country [Krugman, 1997]. 1.6. Conclusions The above presentation of several key factors and their possible role in explaining the crisis shows that there is still no clear consensus on the issue. Several aspects did play a role and only their joint impact led to the abrupt events of end of December 1994. Various models were used to describe the crisis. These were both models of the second generation type, pointing to the role of self fulfilling expec-tations and the political and economic constraints faced by the authorities, as well as modified first generation models stressing the importance of economic fundamentals. With respect to the causes of the crisis following general points can be made: – Private sector savings did not match the level of invest-ment. Mexico had easy access to credit as a result of the si-tuation CASE Reports No. 39 in the world financial markets, and liberalisation of the economy. Resulting credit expansion was not accompa-nied by proper credit screening. – Mexico was relying too heavily on foreign borrowing in 1993 and early 1994, having no easy escape route in the case this inflow would stop. – The exchange rate rule was perhaps not quite consis-tent with the developments in other spheres (overindul-gence of credit, excess of funds in international financial markets, fast growth of short-term debt, financial liberalisa-tion). The real overvaluation of the peso might also have played some role. – Mexico experienced a series of unexpected negative shocks during 1994 – the rise in U.S. interest rates coincid-ing with political tensions in the country. – Mexican politics in 1994 did play an important role in the crisis. – Lack of availability of timely and accurate information on the economic situation in the country might have played some role in the abrupt change of investors' attitude towards Mexico. – The policy response to the shocks of early 1994 was certainly inappropriate (this is an ex post diagnosis). – Neither fiscal nor monetary policy tools were used to adjust the economy to a worsening situation during 1994. – Allowing for the erosion of foreign reserves of that extent while building a large and rapidly growing stock of dollar indexed short term debt in the period March- December 1994 was an extremely risky strategy that did not work. This set the stage for the December crisis and then led to very high interest rates and, consequently, harsh consequences for the real sector. – Perceived risk of financial collapse played a role in both causing the collapse and making it very severe. – Inappropriate management of the devaluation and improper steps taken in the days following it led to a com-plete loss of confidence in Mexican policies and conse-quently to more severe consequences. An interesting feature of the Mexican crisis is the seve-rity of the recession that was caused by it. On the other hand, the crisis was relatively quickly overcome and the economy seems to have overcome its underlying causes. One of the possible explanations of such developments might be that the private sector was indeed heavily depen-dent on external financing. Then again, a relatively quick rebound of the economy could suggest that it was funda-mentally sound, and the crisis exposed it to the liquidity trap. In other words, given the abundance of credit, the pri-vate sector was using it heavily and possibly sometimes unwisely, but exposed to the dramatic change in the exter-nal environment was still able to become competitive again.
- 20. 20 Marek D¹browski (ed.) Appendix: Chronology of the Mexican Crisis – January 1994 – peasant rebellion in the Chiapas province – February 1994 – U.S. interest rates increase slightly – 23 March 1994 – assassination of the presidential can-didate of the ruling party – end of March – April – severe financial turbulence in Mexico: exchange rate depreciates by around 10% reaching the ceiling of the band, Bank of Mexico reserves shrink by 9 billion USD, interest rate rise significantly – April – December 1994 – government continues the process of substituting its short term peso denominated debt with dollar indexed debt – 21 August 1994 – Ernesto Zedilo wins the presidential elections; interest rates fall slightly – 28 September 1994 – assassination of the ruling party leader – October – November – capital outflow continues, Bank of Mexico reserves decline by further 4.7 billion USD – 1 December 1994 – president Zedilo takes office – 20 December 1994 – Finance Ministers announces the widening of the exchange rate corridor by 15% – 22 December 1994 – under pressure from financial markets the authorities announce free floating the peso – 29 December 1994 – appointment of the new Finance Minister, a few days later announcement of the government economic program – 3 January 1995 – IMF expresses its support for the program, announces the establishment of the Exchange Sta-bilisation Fund – 1st quarter 1996 – economic growth (0,1%) resumes to average at close to 7% during the next three quarters CASE Reports No. 39
- 21. 21 The Episodes of Currency Crisis in Latin... References Blejer M.I., G. del Castillo (1996). "Deja Vu All Over Again?" The Mexican Crisis and the Stabilization of Uruguay in the 1970s. IMF Working Paper WP/96/80. Burkart P. (1995). "Mexican Peso Devaluation: An Inves-tigative Report". web page: www.nafta.net/comuniq2.htm Carstens A.G., A.M. Werner (1999). "Mexico's Mone-tary Policy Framework under a Floating Exchange Rate Regime". Mimeo, Banco de Mexico. Dabos M., V.H. Juan-Ramon (2000). "Real Exchange Rate Response to Capital Flows in Mexico: An Empirical Analy-sis". IMF Working Paper, WP/00/108. DeLong B. et al. (1996). "The Mexican Peso Crisis. In Defense of U.S. Policy Toward Mexico". web page: http://econ161.berkley.edu (a shorter version of that paper appeared in May-June edition of Foreign Affairs). Dornusch R., A. Werner (1994). "Mexico: Stabilization, Reform, and No Growth". Brookings Papers on Economic Activity, 1:1994. Edwards S. (1999). "On Crisis Prevention: Lessons form Mexico and East Asia". NBER Working Paper WP 7233, Cambridge. Gelos R.G., A.M. Werner (1999). "Financial Liberaliza-tion, Credit Constraints, and Collateral: Investment in the Mexican Manufacturing Sector". IMF Working Paper, WP/99/25. Gil-Diaz F. (1998). "The Origin of Mexico's 1994 Finan-cial Crisis". The Cato Journal Vol. 17 No. 3, Winter, Cato Institute, Washington, D.C. Gruben W.C. (1997). "The Mexican Economy Snaps Back". Southwest Economy, March/April, Federal Reserve Bank of Dallas. Gurrha J.A. (2000). "Mexico: Recent Development, Structural Reforms, and Future Challenges". Finance & Development, Vol. 37, No. 1, March, International Mone-tary Fund. INEGI (2000). "Producto Interno Bruto Real 1991–1999". web page: www.quicklink.com/mexico/ IMF (1995). "World Economic Outlook". May. Kalter E., A. Ribas (1999). "The 1994 Mexican Econom-ic Crisis: The Role of Government Expenditure and Relative Prices". IMF Working paper, WP/99/160. Kaminsky G.L., C.M. Reinhart (1996). "The Twin Crises: The Causes of Banking and Balance-of-Payment Problems". International Finance Discussion Papers No. 554, Board of Governors of the Federal Reserve System. Krugman P. (1979). "A Model of Balance of Payments Crises". Journal of Money, Credit and Banking 11: 311–325. Krugman P. (1997). "Currency crises". (Paper prepared for NBER conference), web site: http://web.mit.edu/krug-man/ www CASE Reports No. 39 La Porta R., F. Lopez-de-Silane (1999). "The Benefits of Privatization: Evidence from Mexico". Quarterly Journal of Economics, November, p. 1193–1242. Lindgren C. J. et al. (1996). "Bank Soundness and Macroeconomic Policy". International Monetary Fund, Washington, D.C. Lustig N. (1992). "Mexico: The remaking of an econo-my". The Brookings Institution, Washington, D.C. Martinez G.O. (1998). "What lessons Does the Mexican Crisis Hold for Recovery in Asia?". Finance & Development, Vol. 35, No. 2, International Monetary Fund. Masson P.R., P.-R. Agenor (1996). "The Mexican Peso Crisis: Overview and Analysis of Credibility Factors". IMF Working Paper WP/96/6. Otker I., C. Pazarbasioglu (1995). "Speculative Attacks and Currency Crises: The Mexican Experience". IMF Work-ing Paper WP/95/112. Sachs J. et al. (1995). "The Collapse of the Mexican Peso: What Have we Learned?" NBER Working Paper WP 5142, Cambridge. Sachs J. et al. (1996). "The Mexican Peso Crisis: Sudden Death or Death Foretold?". NBER Working Paper WP 5563, Cambridge. SHCP (Secretariat of Finance and Public Credit) (1996). "Mexico: Quarterly Report; Second Quarter 1996". web page: http://www.shcp.gob.mx/english/docs/index.html SHCP (Secretariat of Finance and Public Credit) (1995). "Mexico: Quarterly Report; Fourth Quarter 1995". web page: http://www.shcp.gob.mx/english/docs/index.html Werner A.M. (1996). "Mexico's Currency Risk Premia in 1992–94: A closer Look at the Interest Rate Differentials". IMF Working Paper, WP/96/41. World Bank (1995). "Mexican Economic Crisis". World Bank Institute Policy Forum, web page: http://www.world-bank. org/wbi/edimp/mex/mex.html World Bank (2000). "Global Development Finance 2000". WTO (1997). "Trade Policy Review Mexico". web site: www.wto.org
- 22. 23 The Episodes of Currency Crisis in Latin... Part II. The 1995 Currency Crisis in Argentina by Ma³gorzata Jakubiak 2.1. Introduction This paper presents the economic developments that took place in Argentina at the time of the 1995 currency cri-sis. This financial turbulence resulted from the contagion of the Tequila crisis of the late 1994. And although the country maintained its commitment to a peg under the currency board arrangement, the reserves of the central bank were severely depleted and the consequences for the economy manifest. The study starts from a description of the economic reforms of the early 1990s that set the framework for the monetary and fiscal policies in place when the crisis hit. There is then a discussion on the macro- and microeconom-ic climate, followed by the description of policy responses to the crisis. The paper concludes with the assessment of whether the core factors that drove the crisis have been properly addressed by looking at the post-crisis situation. 2.2. Overview of Economic Situation Before and During the Crisis 2.2.1. Reforms of the Early 1990s and Crisis Developments The 1980s were marked by a series of economic and financial problems. One of them was chronic inflation and periodic hyperinflation which led to widespread dollariza-tion of the economy. Moreover, prevailing public sector deficits were crowding out private sector credit (Garcia- Herrero, 1997). A banking crisis erupted in 1980, and then developed into a severe currency crisis a year later. The next crisis started in 1985 and ended in 1987. The subse-quent economic plan aimed at lowering inflation resulted in its outburst in 1989 and the complete dollarization of the economy as investors' confidence weakened. Two huge devaluations took place in the late 1989 and in 1990. After another bout of financial turbulence, one more stabilization CASE Reports No. 39 plan failed. The exchange rate, which under earlier plans had been pegged, was allowed to fluctuate, and all price controls were removed (Choueiri and Kaminsky, 1999). The difficulties of the period 1989–1990 – mainly as a result of hyperinflation – allowed for the general recognition of the need for reforms. The changes, that put the country on a sustained growth path began with the Convertibility Plan of 1991. The Convert-ibility Law, which is still in operation, established the curren-cy board arrangement. Financial sector reforms followed. Argentina also eliminated, by 1993, restrictions on capital flows, relaxed or abolished barriers on imports and exports and deregulated trade and some professional services. Until 1994, roughly 90% of all state-owned enterprises was pri-vatized, bringing considerable gains to economic efficiency [IMF, 1998]. In 1994 Argentina recorded economic growth of 8%, managed to lower inflation to 4.2% (from over 10% in 1993 and 25% in 1992), and kept the budget deficit of the federal government at the level of 0.5% of GDP. The ban-king sector recorded growth in credits and deposits. From the early 1990s, capital started to flow in, as a result of investors' more favourable perception of the region and rel-atively high interest rates. At the time, interest rates in the United States were low, and net flows of portfolio capital to Argentina amounted to 33.7 billion dollars in 1993 and to 8.4 billion in 1994. These huge capital inflows led to an explosion of domestic credit, consumption, real estate and stock market booms, and lack of diversification of bank portfolios [Choueiri and Kaminsky, 1999]. The current account deficit deteriorated as a result of real exchange rate appreciation. As the U.S. and world interest rates soared in 1994, the Mexican peso was deva-lued in December 1994 and the capital outflows brought about balance of payment pressures, the rumors about abandoning the currency board spread out. The Argentine banking system suffered from a run on deposits, and the credit crunch followed. Between December 1994 and March 1995, the central bank (BCRA) lost 41 percent of its international reserves, defending the peso-dollar peg. The banking system lost 18% of its deposits in five months which generated liquidity problems. A number of prudential regulations were introduced in early 1995 in order to
- 23. 24 Marek D¹browski (ed.) CASE Reports No. 39 restore confidence in the banking system. By the end of the 1995, the deposits went nearly back to their pre-crisis le-vels and the monetary authorities managed to restore its gross international reserves [IMF, 1999]. The currency board was defended. However some smaller banks conti-nued to experience problems, and the crisis resulted in a severe recession. In 1995, real GDP went down by 4%. Among the core factors that allowed the transfer of external shocks to the real economy and leveraged the cri-sis were weak links with world financial markets, relatively underdeveloped domestic financial markets, labor market rigidities, systematic crowding out, and real exchange rate inflexibility imposed by the currency board regime [Caballero, 2000]. All these factors are addressed in the fol-lowing sections of the paper. 2.2.2. Monetary and Exchange Rate Policy From 1991 Argentina has followed a currency board exchange rate arrangement. The currency board is – after the classical monetary union – the second most rigid form of the exchange rate regime. In this orthodox form of a fixed exchange rate regime the role of monetary authorities is reduced to issuing notes and coins that are fully backed by a foreign reserve currency on demand at a fixed exchange rate. There is a minimum of 100% backing of reserve money by net foreign assets of the central bank, and cur-rency boards often hold excess reserves to offset against asset valuation changes. These excess reserves are related to the net worth of a currency board (Pautola and Backé, 1998), because seigniorage can be earned only from interest on reserves. The 1991 Convertibility Law and the 1992 Central Bank Charter created the basis for the functioning of the curren-cy board in Argentina. The exchange rate of the Argentine peso was fixed at one against U.S. dollar, and the central bank was required to keep 100% of its monetary base in international reserves. However, since 1995 1/3 of it may be kept in the safe dollar-denominated government bonds [Hanke, Schuler, 1999] while holdings of these securities cannot grow by more than 10% per year. These re-gula-tions eliminated the possibility of inflationary financing of the government deficit. Moreover, the charter restrains the central bank from financing provincial or municipal govern-ments, public firms, or private non-financial sector [Pou, 2000]. The central bank became fully independent from the legislative and executive branches of government, and set its principle goal at maintaining the value of domestic currency. As can be seen from the above description, the Argen-tine currency board is not the strictest form of a currency board where monetary authorities cannot intervene in the market, cannot act as a lender of last resort, and where interest rats are solely market-determined. Indeed, the Central Bank of Argentine Republic (BCRA) has some room for discretionary monetary policy, because its international reserves are not fully backed by the domestic currency and because it can set reserve requirements for commercial banks. This ability to retain some flexibility was used during the 1995 financial crisis. The international reserves of BCRA started to shrink quickly in January 1995 when the Argentine peso came under a pressure, and when the bank tried to rescue trou-bled commercial banks (see Figure 2-1). The reserves hit the lowest level in March 1995, which was around 2/3 of the monetary base, the minimum coverage requirement. BCRA Figure 2-1. International reserves and monetary base coverage, 1994–1996 17000 16000 15000 14000 13000 12000 11000 10000 9000 8000 7000 1.4 1.2 1 0.8 0.6 0.4 0.2 0 millions of USD Total reserves minus gold (foreign exchange+SDPs) Reserves/Reserve Money 1994M1 1994M3 1994M5 1994M7 1994M9 1994M11 1995M1 1995M3 1995M5 1995M7 1995M9 1995M11 1996M1 1996M3 1996M5 Source: own calculations on the basis of IFS data
- 24. 25 The Episodes of Currency Crisis in Latin... Table 2-1. Interest rates, 1994–1995 1994 1995 10 11 12 1 2 3 4 5 6 7 8 9 10 Prime deposit rate 8.27 8.72 9.55 10.65 11.64 19.38 19.07 15.54 10.83 10.24 9.17 9.21 8.92 Lending rate 9.83 10.00 13.56 18.06 19.06 34.05 26.45 22.13 16.19 14.57 13.29 13.26 12.55 Real interest rate 1.94 1.99 2.08 2.17 3.30 11.65 11.36 8.40 4.29 4.07 3.31 3.86 3.73 differential Source: IFS, own calculations Note: real interest rate differential is calculated as the difference between real Argentine deposit rates and real U.S. deposit rates was then buying dollar-denominated Treasury bonds, in order to mitigate the effects of the credit crunch. The cen-tral bank's holding of these government notes increased by 25% from 1994 to 1995, and declined sharply afterwards [Caballero, 2000]. This decline is reflected in the Figure 2-1 as the high indicators of foreign exchange reserves in 1996. As the central bank was depleting its reserves, interest rates rose sharply, reflecting a domestic liquidity squeeze and rise in the country risk premium. The real interest rate differential vs. U.S. deposit rate reached above 11 percent-age points, while during 1994, its value was close to 2. High interest rates induced the private sector to lower demand for credit [1] and reduce expenditures. Banks cut their cred-it lines and refinancing facilities. All these factors contributed to a decline in the economic activity in 1995 and to higher unemployment [Catao, 1997]. 2.2.3. Fiscal Policy Through the early Convertibility Plan years, the govern-ment was running budget deficits. The central government budget deficit averaged 0.5% of GDP during the years 1991–1994 – with a surplus only in 1993. The deficit of the consolidated public sector (including federal budget, provin-cial government budgets, and off-budgetary funds and pro-grams) averaged 1.8% of GDP, with 2.5% of GDP in 1994. The crisis year of 1995 was marked with the 1.5% deficit of central government budget, and 4.3% deficit of the whole public sector. According to the IMF calculations, fiscal policies were pro-cyclical in the early 1990s, with public sector deficits ris-ing faster than the cyclically adjusted public sector balance [IMF, 1999]. During this period (1991–1994), when output was growing above potential and fiscal impulse was expan-sionary, the lowering of inflation has been achieved by the nominal exchange rate anchor and the supply-side oriented reforms, such as change in the tax system, and elimination of distortionary tariffs. Only from 1995 onwards, as the coun-try slipped into recession, did the fiscal impulse start to have a negative impact on demand, thus contributing to the CASE Reports No. 39 reduction, and finally to the elimination of inflation [IMF, 1999]. Tax reform aimed at eliminating some taxes and shift-ing the relative tax incidence from production to con-sumption and incomes was implemented in the early 1990s. Many distortionary taxes, such as those on exports, bank debits and assets, with a yield about 3% of GDP, were removed. Some exemptions, mainly from VAT, as well as subsidies, were abolished. To improve labor market flexibility, the government significantly reduced the employer payroll tax in some sectors (reversed for a couple of months during the 1995 crisis). There was also a significant decline in the number of workers employed by the state as efforts to improve efficiency in the public sector were undertaken (the provinces started to be responsible for the health and education services [IMF, 1999]). New and stronger laws increased the govern-ment's ability to control tax evasion. Pension reform, aimed at shifting from the pay-as-you-go publicly funded system to a system combining public transfers and private capitalization, started in mid-1994. The reform resulted in the reduction of future liabilities of the public sector. However, the immediate costs for the budget are estimated to be around one percent of GDP per year, as the government pays the contribution to the private system for those who voluntarily opted to shift away from the pay-as-you-go scheme. In any event, these costs appeared first in the consolidated budget in 1996, exactly one year after the currency crisis, and thus the consequences of this reform for the fiscal sector are not explored further. 2.2.4. Private and Public Debt The reforms of the early 1990s allowed Argentina to return to the voluntarily financing of its external debt, which was rescheduled under the Brady Plan [Pou, 2000]. The developments in borrowing from the international capital markets, both public and private, have moved in the direc-tion of declining spreads, the lengthening of maturity, and [1] There were also other factors contributing to the decline of the private sector demand for credit, such as crowding out by government bor-rowing, which turned to domestic banks for financing its monetary interventions in 1995, such as providing troubled banks with fresh credit.
- 25. 26 Marek D¹browski (ed.) CASE Reports No. 39 the fixed rate nature of the debt (IMF, 1998). However, these developments were for some time reversed after the 1995 crisis. Total external debt rose sharply in 1995 and in 1996, and a large part of this change can be attributed to the rise of short-term debt. The debt of the public sector rose mainly because more bonds were issued in 1995 and 1996. Out-standing public debt attributable to the bond issues amount-ed to $12.4 billion in 1994, while in 1995 and 1996, the value rose to $14.8 billion and $23.5 billion respectively. Average maturity of the international public bonds issued in 1995 actually slightly increased in comparison with the pre-vious year (from 4.8 to 5.0 years). The sharp fall in the average maturity of bonds issued in 1995 shown in the private sector, where this maturity decreased from 4.5 years in 1994 to 2.6 years. The spread significantly increased. As the government was issuing more international bonds in 1995, the value of bonds issued by the private sector fell dramatically, to less than $1 billion, and returned to their pre-crisis value in 1996. However, the relatively small size of total private external debt suggests weak access to international financial mar-kets of the private sector. 2.2.5. Savings and Investment Typically for a fast-growing economy, Argentina was dependent on foreign savings to carry out investments necessary in order to sustain its economic growth. Domestic resource gap, which has been in place during the 1990s, resulted mainly from low domestic savings. The rate of investment, although significantly higher than sav-ings, was still lower than investment rates for Central Table 2-2. Main economic indicators, 1991–1997 1992 1993 1994 1995 1996 1997 Real GDP growth 5.7% 8.0% -4.0% 4.8% 8.6% Nominal GDP (millions of peso) 226 847 236 505 257 440 258 032 272 150 292 859 CPI inflation 24.9% 10.6% 4.2% 3.4% 0.2% 0.5% Unemployment Rate 7.2% 9.1% 11.7% 15.9% 16.3% Structure of GDP*: Agriculture 6.0% 6.7% 6.4% 7.0% 6.9% 6.6% Industry 30.7% 32.6% 32.3% 32.1% 32.1% 32.9% Services 63.3% 60.8% 61.2% 60.9% 60.9% 60.5% General Government Balance (as % of -0.2% 0.9% -0.5% -1.5% -2.4% GDP) Public Sector Balance (as % of GDP) -0.5% -0.9% -2.5% -4.3% -4.2% Broad money (M2) monetization 11.2% 16.3% 19.4% 18.8% 21.1% 24.0% Population (millions) 33.42 33.87 34.32 34.77 35.22 35.67 Source: IFS, WDI, own calculations based on the IFS and WDI data Note: * structure of GDP from 1992 is not fully comparable with later data Table 2-3. Argentine external debt, 1992–1996 1992 1993 1994 1995 1996 TOTAL EXTERNAL DEBT 68 345 70 576 77 434 83 536 93 841 In % of GDP 29.8% 29.8% 30.0% 32.4% 34.5% Total long-term debt 49 855 58 403 66 052 67 235 75 348 Public and publicly guaranteed 47 611 52 034 55 832 55 970 62 392 Private non-guaranteed 2 244 6 369 10 220 11 265 12 956 Total short-term debt 16 176 8 653 7 171 10 170 12 200 Total short-term debt (% of total 23.7% 12.3% 9.3% 12.2% 13.0% external debt) Source: Global Development Finance, 1998, and own calculations based on IFS and GDF Table 2-4. Savings and investment (in percent of GDP), 1994-1997 1994 1995 1996 1997 Gross national savings (% of GDP) 16.3 16.5 15.5 16.3 Gross national investment (% of GDP) 20.0 18.0 17.7 20.0 Source: IMF (1999)
- 26. 27 The Episodes of Currency Crisis in Latin... European countries in the mid-1990s, or than the invest-ment ratios of Southern European countries in the 1980s. It can be seen from the comparison of capital flows that the majority of this domestic resource gap was financed through foreign direct investment. 2.2.6. Foreign Trade One of the reforms of the Convertibility Plan involved the elimination of all tariffs on exports and the majority of non-tariff barriers on imports. Imports tariffs have been cut form over 40% average rate in 1989 [IMF, 1998] to 8.4% at the end of 1994, with a zero rate on capital goods and raw materials. This eliminated distortions in foreign trade. Argentina is an exporter of raw materials and some lightly processed primarily products. These primary and agro-industrial products account for 70% of all exports, and are subject to high fluctuations in world prices. Since 1990, there has been a quick rise of manufacturing exports to Brazil, following the creation of MERCOSUR (South American customs union). But still, in 1994 this number was below 25% of total value of exports. Generally, because of high commodity concentration, Argentine exports have been highly volatile. Following the removal of restrictions and high eco-nomic growth, Argentine imports between 1991 and 1997 was growing more than four times as fast as real GDP. Catao and Falcetti (1999) estimated that long-run income elasticity of imports was above 2 during this period, and that this elasticity showed some pro-cyclical behavior. It should be noted that between 1991 and 1995, the real exchange rate kept appreciating, which also explains the rapid growth of imports at this time, but not as much as booming economic activity. Over 50% of imports in 1994 came from the US, the EC and Japan. Trade deficits in place since the early 1990s were main-ly driven by high dependency on world prices for exports, soaring domestic demand and incomes, together with real appreciation toyical or most emerging markets. Trade sur-pluses of 1995–1996 were taking place because world prices for traditional Argentine exports increased, the economy stepped into one-year long recession and there was a fall in the real effective exchange rate during the Tequila crisis. 2.2.7. Balance of Payments Argentina has been systematically running current account deficits in the 1990s, which is a typical feature of an emerging economy. Even in the absence of trade deficits, as in 1995 when imports fell because of the reces-sion, the negative current account balance was caused by the outflows of the investment incomes. As 1995 showed, the current account deficit had to be financed by foreign borrowing, following changes in the investors' preferences and the sudden outflow of short-term capital. It should be noted that there have been large inflows of portfolio capital during the two years preceding the cri-sis. Net inflows of short-term capital in 1993 amounted to Figure 2-2. Foreign trade, 1993-1997 9000 8000 7000 6000 5000 4000 3000 2000 1000 0 CASE Reports No. 39 Imports Exports 1993Q1 1993Q2 1993Q3 1993Q4 1994Q1 1994Q2 1994Q3 1994Q4 1995Q1 1995Q2 1995Q3 1994Q4 1996Q1 1996Q2 1996Q3 1996Q4 1997Q1 1997Q2 1997Q3 1997Q4 Source: IFS
- 27. 28 Marek D¹browski (ed.) around 15% of Argentine GDP, which was at that time 16 times more than the net inflows of long-term investment. However, Foreign Direct Investment inflows have been growing systematically during the analyzed period, notwithstanding the 1995 decline in real output. The pri-mary reasons being prevailing international conditions, the implementation of structural reforms by Argentina, priva-tization, and the elimination of restrictions on foreign investors, as well as the removal of restrictions on capital transactions in general [IMF, 1998]. Since the beginning of the 1990s, there have been no capital controls on financial or commercial operations between residents and nonresi-dents [BCRA, 2000]. The significant rise in FDI flows in 1995 and in 1996 can be attributed mainly to the creation of the private pension funds, and to the sales of private firms to foreign investors CASE Reports No. 39 Table 2-5. Balance of payments (mil. USD), 1992–1997 1992 1993 1994 1995 1996 1997 Current account balance -5 521 -8 030 -10 992 -4 985 -6 521 -11 954 Trade balance -1 396 -2 364 -4 139 2 357 1 760 -2 123 Exports of goods 12 399 13 269 16 023 21 161 24 043 26 431 Imports of goods -13 795 -15 633 -20 162 -18 804 -22 283 -28 554 Non-factor services (net) -2 463 -3 221 -3 692 -3 326 -3 366 -4 178 Investment income -2 393 -2 931 -3 567 -4 529 -5 331 -6 089 Current transfers (net) 731 486 406 513 416 436 Capital and financial account 7 350 20 328 11 155 4 623 11 175 16 826 Direct investment (net) 3 218 2 059 2 477 3 818 4 922 5 099 Portfolio investment (net) 4 513 33 731 8 389 1 864 9 727 11 087 Other investment (net) -381 -15 462 289 -1 059 -3 474 640 General government -1 343 -10 196 969 1 197 -199 136 Bank 76 -570 761 2 570 -2 744 -1 615 Other sectors 739 -697 -1 423 -4 832 -567 2 130 Net error and omissions 54 -1 173 -872 -1 853 -1 316 -1 498 Overall balance 1 883 11 125 -709 -2 215 3 338 3 374 Financing -1 883 -11 125 709 2 215 -3 338 -3 374 Reserve assets -3 264 -4 279 -685 82 -3 875 -3 293 Use of IMF credits -73 1 211 455 1 924 367 -38 Exceptional financing 1 454 -8 057 938 209 170 -43 Source: IMF IFS Figure 2-3. Private and public net capital flows to non-financial sector, 1992–1997 4. 0% 3. 0% 2. 0% 1. 0% 0. 0% -1 .0% -2 .0% 1992 1993 1994 1995 1996 1997 % of GDP Public non-financial sector Private non-financial sector Source: own calculations based on the data from Argentine Ministry of Economy and IMF IFS
- 28. 29 The Episodes of Currency Crisis in Latin... [IMF, 1998]. Net FDI flows accounted for around 1% of GDP in 1992–1994 and for 1.7% on average, during the period 1995–1997. However, there was almost no change in the aggregate existing foreign investment stock, which averaged around $3.5 billion per year during the period 1992–1995. Its structure, though, has been changing, indi-cating faster growth of FDI non-related to the privatiza-tion opportunities. FDI grew most rapidly in the commu-nications and manufacturing sectors. Significant privatiza-tion- related investment flows were recorded also for the electricity, gas and water and for the petroleum industries. Around 40% of all FDI coming to Argentina during 1992–1995 originated in the USA. Despite the important role of huge inflows of portfolio investment in 1993, capital inflows (both private and pub-lic) constituted a relatively small fraction of GDP for a fast growing economy. During 1992–1994, while GDP was increasing by 7.8% per year, overall capital flows [2] amounted to 4.9% of GDP. This suggests that the link of Argentine financial market with international financial mar-kets was weak. As it was visible in 1995, this significantly constrained government ability to use international financ-ing when there was an external shock. Official capital flows depicted on the Figure 2-3, rose in 1995 supported by loans from IADB and the World Bank. 2.2.8. Real and Nominal Rigidities As there was little room for monetary policy, the bur-den of adjustment during 1995 fell on wages and prices. Argentina has a European-style labor market with centra-lized bargaining, high severance costs, and still high – on average – wage taxes. These rigidities – both nominal and real – amplified external financial shock by forcing a larger share of adjustment on output and employment. This was costly, since unemployment has stayed well above 10%, even in 1999. In addition, real exchange rate inflexibility, brought about the convertibility regime, combined with labor mar-ket rigidities and limited access to financial markets made the 1995 fall in output worse [Caballero, 2000]. 2.2.9. Banking System The introduction of the Convertibility Plan markedly changed the Argentine banking sector. High inflation and macroeconomic volatility of the 1980s, together with large capital flight, caused the demand for domestic money to decline heavily. The ratio of broad money (M3) bottomed at 6 percent of the GDP in 1990, the overall sum of deposits of the banking system was low, and real interest rates on deposits were negative. Thus, the banking system reforms of the early 1990s focused primarily on the strengthening of the whole system and on removing obsta-cles to financial intermediation [IMF, 1998]. The introduction of the simplified system of reserve requirements began in 1991. The rates on both foreign and domestic currency transactions were unified, and the adherence to these requirements was made more effec-tive. New capital adequacy requirements, incorporating lending interest rate risk factor, were put in place, and they have been gradually tightened. The loan classification, portfolio risk rules and provisioning rules were introduced in 1994. The supervisory role of the financial superinten-dence was reinforced in order to verify compliance with all the prudential standards. The overall result of these regulations was visible in the fall of outstanding central bank credits to financial institutions [IMF, 1998]. The risk-weighted capital to assets ratio rose to 11.5% in January 1995, well above the 8% Basle standard [Pou, 2000]. The reserve requirements averaged 17.5% of deposits [IMF, 1998]. The reforms of the early 1990s removed the barriers of entry and increased competition between banks. They were aimed at ensuring the safety of individual banks and the whole banking system. They also intended to reduce moral hazard. These issues were very important, since under the currency board, the BCRA role as the lender of last resort was very limited. Basically, the central bank was not allowed to provide liquidity to the banks in finan-cial trouble. The BCRA was able to conduct intervention through repo operations, but this was limited to the smoothing of fluctuations in the interbank market. The Convertibility Law allowed extending credit to financial institutions in the emergency situation only [IMF, 1998]. 2.2.9.1. Developments during December 1994- March 1995 As the fears about the possible abandoning of the peg, triggered by the Mexican devaluation, intensified, and as Argentine debt prices kept falling, banks experienced a run on deposits. This situation is shown on the Figure 4, describing effective growth rates of loans and deposits [3]. As the depositors started to withdraw their funds, it initially affected wholesale banks, whose loan portfolios were composed mainly of government bonds, then spread to the entire banking system. In 1995, the BCRA [2] Together with banking sector flows. [3] Time series of deposits and loan growth rates, each less their respective real interest rate CASE Reports No. 39
- 29. 30 Marek D¹browski (ed.) Figure 2-4. Effective growth rate of deposits and loans, 1994–1997 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% -5.0% -10.0% -15.0% -20.0% 1994M6 1994M9 1994M12 1995M3 1995M6 1995M9 1995M12 started to use its reserves to provide funds to banks. The central bank lost nearly 20% of its monetary base cove-rage [IMF, 1998] [4], which was the limit of the interven-tion allowed as a response to crisis situation. Despite these efforts, the loss of deposits resulted in a huge cre-dit crunch. It should be noted that an important feature of the Argentine banking system at the time were the problem loans, accounting for more than 10% of the total loan portfolio in 1994. Moreover, they have not been uni-formly distributed [IMF, 1998]. This factor was responsi-ble for some of the changes in the banks' assets that hap-pened in the early 1995. Another important factor was the lack of official deposits insurance, which, coupled with the restricted role of the BCRA as a lender of last resort, intensified the perception of deposit risk. During the early phase of the run of deposits, there has been a visible shift to quality. First of all, this meant that depositors started to convert peso deposits to dollar deposits, expecting the devaluation. This is shown on Fi-gure 5. While the peso deposits of the whole banking sys-tem have been falling since December 1994, the dollar deposits were still growing in February 1995. Then the fall in the dollar deposits was less pronounced, and after mid-1995 they quickly started to build up again. At the same time, peso deposits stayed at a relatively unchanged level until the end of the year. We can see that the reco-very of deposits in the Argentine banking system towards 1996M3 1996M6 1996M9 1996M12 1997M3 the end of 1995 was attributable mainly to the increase in the amount of foreign currency deposits. Secondly, the non-uniform distribution of the problem loans meant that small public provincial banks had a dis-proportionately bigger share of non-performing loans than the larger banks. When the run on deposits started, these small banks suffered more, as their depositors start-ed to move funds to the larger banks. Public provincial banks lost their market share (from 12.8% of total assets in 1994 to 9.6% in 1995), while private banks gained. While 56.7% of total assets belonged to the private banks in 1994, the share increased to 58.6% at the end of 1995. The market share of the large national banks remained relatively unchanged at around 30% of total assets [Bur-disso et. al, 1998]. It is claimed that although part of the small banks market share was lost as a result of privatiza-tion following the crisis, an important fraction was gone due to a change in the market perception of their credit risk. Another way of looking at the situation of the small banks during the Tequila crisis is to examine their indica-tors of profitability. Return on total assets of the 20 largest banks remained relatively stable and generally positive during the years 1994–1997, while the profitability of total retail banking sector was much more volatile, often nega-tive, and fell sharply during the first months of 1995 [Bur-disso and D'Amato, 1999]. This situation increased the contagion, as some banks have virtually found themselves CASE Reports No. 39 [4] Around 30% of usable foreign exchange reserves according to the author's calculations (see Figure 2-1). 1997M6 1997M9 1997M12 Loans Deposits Note: The term "effective" refers to the annual growth rates of credit and deposits less the real lending and the real deposit rates, respectively (after Caballero, 2000). PPI year-over-year inflation was used in the construction of real interest rates. Source: own calculations based on IFS
- 30. 31 The Episodes of Currency Crisis in Latin... Figure 2-5. Peso and dollar deposits of the deposit money banks, 1994–1997 70000 60000 50000 40000 30000 20000 10000 0 short of liquid assets what affected even a number of apparently solvent institutions [IMF, 1998]. The full-scale run on banks started in February 1995. 2.2.9.2. Role of Private, Public and Foreign Ownership Although the privatization process started in early 1993, following the remonetization of the Argentine econ-omy and the redefinition of the role of the public sector, only after the Tequila crisis did this process gain momen-tum. Until early 1995, only 3 banks had been privatized [Burdisso et. al, 1998]. At the end of the 1994 there were 135 private banks – both foreign and domestically owned – in Argentina. They accounted for over 50% of the whole banking system assets. 33 public banks were on average larger: they had 30% (national) and 13% (provincial) of the market share [Burdisso et. al, 1998]. Following the removal of restrictions on foreign investment and capital flows of the early 1990s, the num-ber of foreign banks in Argentina increased. However, in 1994 they still accounted for under 20% of system assets only [Goldberg et. al, 2000]. When comparing their loan portfolios to those of the state-owned banks, it shows that the foreign banks had lower mortgage shares and higher shares of commercial, government, private and other lending. Foreign banks were similar under this respect to the domestic private banks. However, the for-eign banks were perceived as generally safer and health-ier. The loan growth rates of the foreign banks were sub-stantially higher than the respective rates of the domes-tic banks in 1994, and they continued to grow faster even during the crisis period [Goldberg et. al, 2000]. According to the research of D'Amato et. al (1997), bank "fundamentals" – such as their profitability and level of interest rates – as well as the overall macroeconomic situation were very important in driving the dynamics of deposits. However, there is evidence of contagion effects in the group of small and medium-sized banks on which public information was poorer. 2.2.10. Domestic Financial Market The financial markets in 1994–1995 had tenuous links with the international markets. As already mentioned, the relatively small inflows of capital relative to the size of the economy bear this out. Another argument in favor of the weak link with the world markets is the fact that foreign capital focused mainly on large enterprises, and that the smaller companies had difficult access to the international markets. During the early months of 1995, the volatility of the stock index for prime companies and the overall stock index substantially differed. The volatil-ity of stock index for prime companies (MERVAL) increased significantly, while the volatility of the total market index remained relatively unchanged [Caballero, 2000]. Notwithstanding volatility, values of both MERVAL (see Figure 2-6) and total stock market index have been relatively low from December 1994 until the end of 1995. There was also a large spread-premium on Argentine sovereign bonds relative to the U.S. throughout this time, which further confirms the country risk-premium and its weak ties with international financial markets. Moreover, Argentine markets are still underdeveloped by interna- CASE Reports No. 39 peso deposits dolar deposits 1994M1 1994M5 1994M9 1995M1 1995M5 1995M9 1996M1 1996M5 1996M9 1997M1 1997M5 1997M9 nillion of peso Source: own calculations based on data provided by A. Ramos from the IMF, and on the IFS
- 31. 32 Marek D¹browski (ed.) Figure 2-6. Stock market indicators, 1992–1998 30% 25% 20% 15% 10% 5% 0% Dec 92 Dec 93 Dec 94 Dec 95 tional standards. The broad money, loans, and the stock market capitalization (see Table 2-2 and Figure 2-6) expressed as a fraction of GDP are relatively low [Caballero, 2000]. Stock market capitalization reached over 50% of GDP in Chile, more than 25% in Mexico, around 40% of GDP in Spain and Portugal, and about 100% in the U.S. in 1997. The figure in Argentina was well below 20% of GDP during the period of 1993–1996. Broad money monetization, which stayed during the years 1993–1996 at around 19% of GDP in Argentina, had on average val-ues around 25% of GDP in Brazil and Mexico. Leaving aside well developed, leading markets, Argentine per-formed poorly even when compared to other countries in the region. The consequences of the weak financial links and the sub-development of the domestic market became visible during the 1995 crisis. When the country found itself near the limit of access to international markets, this hampered the swift allocation of resources. 2.2.11. Private and Public Sector Prior to the crisis, the enterprise sector was almost entirely private, as during 1991–1994 around 90% of all state-owned enterprises were privatized [IMF, 1998]. This move increased productivity, brought significant gains in 800 700 600 500 400 300 200 100 the reallocation of resources as the public sector shrank, but also boosted unemployment. Despite a large and expanding private sector, there were considerable differences in the growth prospects between larger enterprises and agricultural producers together with small industrial enterprises located in the countryside. Diffi-cult access to bank credit may serve as an example. Although the size of the private sector has became more and more important since the beginning of the 1990s, one of the problems of the post-crisis period was the lack of recovery of private sector credit. It is claimed that its growth has been repressed by huge government borrow-ings from the domestic market, which took place during 1995. As the government turned to domestic banks for financing its monetary interventions during the Tequila crisis (Figure 2-7) while facing external constraints, the private sector credits have been crowded out [Caballero, 2000]. The fast consolidation process in the banking system also enhanced this trend [5]. 2.3. Political Situation and Management of the Crisis It is believed that the run on deposits of March 1995 was also caused by the bad perception of the current polit- [5] As many of the local branches of the wholesale and cooperative banks disappeared after 1995, the information concerning their clients' credit-worthiness was not available (there was no countrywide credit rating system). There is evidence that the surviving banks were unwilling to "screen" their potential clients and did not want to lend to unknown borrowers from the countryside. Larger share of resources was then used to buy govern-ment CASE Reports No. 39 bonds and improving liquidity position [Catao, 1997]. Dec 96 Dec 97 Dec 98 0 stock market capitalization as % of GDP MERVAL (stock exchange intex for prime companies) Source: National Securities Commission
- 32. 33 The Episodes of Currency Crisis in Latin... Figure 2-7. Net public borrowing from domestic banks relative to private sector credit 35% 30% 25% 20% 15% 10% 5% 0% 1994Q1 1994Q2 1994Q3 1994Q4 1995Q1 1995Q2 1995Q3 1994Q4 1996Q1 1996Q2 1996Q3 1996Q4 1997Q1 1997Q2 1997Q3 1997Q4 1998Q1 1998Q2 1998Q3 1998Q4 ical and economic policy related situation. There were uncertainties concerning short-term fiscal policy, enhanced by the incoming presidential elections (on the May 14th). As the election system changed, the incumbent president needed over 50% of votes to avoid the second round. There was no IMF program in place at this time [D'Amato et. al, 1997]. Towards the end of April, the central bank charter has been changed slightly to allow a more flexible use of redis-counts in order to help banks. This move was misinter-preted as a relaxation of the currency board regime. Moreover, there have been spreading rumors about the possibility of suspending convertibility of bank deposits [D'Amato et. al, 1997]. The monetary operations of the BCRA linked with the announcement of the new fiscal package after an agree-ment with the IMF, stopped the fall of deposits during the period between March and May 1995. The IMF, as well as other international institutions promised a significant amount of financing [6] [D'Amato et. al, 1997]. And indeed, there has been net inflow of $1.9 billion until the end of the year. The dynamics of the recovery of deposits after May 1995 varied according to their types. First of all, this recovery can be attributable mainly to the dollar deposits, as the time was needed to restore confidence in the domestic currency. Secondly, the quickest response came from the deposits of the large national public banks, and then from the foreign banks. These groups of banks were the first to start regaining their deposits. The cooperative banks and interior banks still suffered the largest fall in deposits in mid-1995. The immediate response of the monetary authorities to the crisis situation was primarily directed towards improving the liquidity of the banking system. The authorities lowered reserve requirements and provided troubled banks with fresh credits (thus depleting the reserves of the BCRA in April-May 1995) through swaps and rediscounts of prolonged maturity. This was allowed due to the already mentioned modification of the central bank charter. The BCRA acknowledged rediscounts beyond 30-day window in case of systemic liquidity prob-lems [IMF, 1998]. Two Fiduciary Funds were created: one to facilitate mergers and acquisitions within the private banking sector, and the other to foster the privatization of both provincial banks and firms [Burdisso et. al, 1998]. The temporary safety net redistributing the liquidity within the system and controlled by the largest national banks was created, as well as a privately managed deposit insurance scheme. This deposit insurance system was founded with compulsory contributions of financial institutions as a surcharge on deposits. The scheme has CASE Reports No. 39 Net claims on government in % of private sector credit Source: own calculations based on IFS data [6] On April 6, 1995, the IMF approved the fourth year extension of the extended Fund facility (EFF) for Argentina. The three-year EFF, approved initially in 1992, was supposed to support Argentina's medium-term economic and financial program. With the extended program, in April 1995, about US$1.6 million was immediately available to Argentina, and the rest (US$1.2 million) have been disbursed in three quarterly installments. One year later, the IMF approved a stand-by credit for Argentina of about US$1 million over the next 21 months, in support of the government 1996–1997 econom-ic and financial program [IMF, 1995 and 1996].
- 33. 34 Marek D¹browski (ed.) an upper limit per depositor in order to hamper moral hazard [IMF, 1998]. As the confidence in the banking system was restored, deposits kept mounting, and the interest rate spreads started to decline, the government commenced the intro-duction of new prudential measures. Generally, these measures were aimed at further raising the liquidity of financial institutions, capital to assets ratios, addressed still existing information asymmetries in the credit market, and improved the existing payment system [IMF, 1998]. In 1995, the reserve requirements were still being replaced by liquidity requirements, with rates depending of the residual times of maturity [7] [BCRA, 2000]. These requirements have been gradually tightened over time by increasing their rates [8] and extending their applicability to other types of bank liabilities. The purpose of this move was to improve the public perception of individual banks' liquidity position and limit imprudent lending policies as there is evidence that the public discriminated against "good" and "bad" banks on the basis of their perceived liq-uidity position during the Tequila crisis. Between 1996 and 1997, the authorities further strengthened banks' capacity to withstand liquidity short-ages by the creation of a contingent repo facility between the BCRA and a group of 13 major international banks. This agreement allows to swap a collateral – Argentine government securities owned by the central bank or by domestic financial institutions – for up to $ 7.3 billion in cash [IMF, 1998]. Capital to assets ratio was further increased by the incorporation of a new weighting system that takes into account market risk factors, as well as by the increased role of the regulator of banks. There were also steps towards the improvements in the information about debtors available to financial institutions (addressing adverse selection problem), as well as about individual banks. There were also significant improvements in the functioning of the payment system, which today consists of a real-time gross settlement scheme and three auto-mated clearinghouses [Pou, 2000]. It is claimed that the "second set" of prudential measures significantly improved the liquidity of the system, as there was no system-large run on banks during the Brazilian crisis. 2.4. Post-Crisis Developments Domestic financial markets have grown visibly since 1994. Some indicators, such as monetization or stock mar-ket capitalization, show gradual improvement. Neverthe-less, the rise of private sector credit has been constrained for a long time. Credit to the private sector has basically did not recover until only very recently. The banking system increased significantly, and became generally healthier. Regulations implemented after the 1995 reduced its exposure to external shocks. As a result of consolidation within private banks, privatizations and closures, the number of banks declined to 119 in 1999. Several smaller, provincial banks, which were not trans-parent and suffered during the 1995 run, have been priva-tized. Non-performing assets in the banking sector decreased significantly. They accounted for 8% of total assets in private banks and for 13% in public banks in 1997. There is a need for further fiscal adjustment. The sup-ply- side rigidities should be addressed, so that external shocks will not affect the real economy as quickly as in 1995. The public sector is still running large deficits which is a problem. It is argued that around 1% of GDP of these deficits per year can be attributed to the pension reform, effective since 1996, and thus should improve future effi-ciency. Nevertheless, as the external conditions hardened again in 1998–1999, public sector deficits went up to over 4% of Argentine's GDP. The country risk still remains high and was increasing during the Asian, Russian, and Brazilian crises. The ratio of broad money to GDP – although growing over time – is low by international standards. Similar to the Argentine country risk, also interest rates were rising in response to the recent currency crises. Although the increases, which took place in 1998 and 1999, were significantly lower than during 1995, the rates stayed high for a long period, and finally soared in 2000. The high interest rates in the period of 1999–2000 reflect significant decreases in consumer and business con-fidence, and the progressive hardening of the borrowing conditions on international financial markets [9] (and hence the suppressed access to foreign borrowing for Argentine investors). The slow recovery from recession affecting Argentine economy since mid-1998, has also been attributable to the impact of fiscal tightening on domestic demand, the political uncertainties (new govern-ment taking office in 1999) reflecting doubts about the course of economic policy, and the downturn of trading partner demand. Argentine GDP fell by 3.4% in 1999, and according to the preliminary data, by 0.2% in 2000 [IMF, 2001]. There has been consumer price deflation, and the significant fall in CASE Reports No. 39 [7] Higher for shorter residual time of maturity. [8] To reach 20% of most banking liabilities in 1998; the rate substantially higher than in other countries of the region. [9] High U.S. interest rates, and reduced access to international financial markets for emerging economies in general.
- 34. 35 The Episodes of Currency Crisis in Latin... export volume. Trade balance improved in 2000, but main-ly as a result of suppressed imports. Domestic investment fell to 16% of GDP, and domestic savings – to less than 13% of domestic production. External public debt reached 32% of GDP in 2000, and public finances deteriorated. To ease the government financing constraint, the authorities have secured a financial support package of about US$39 billion, including an augmented stand-by agreement with the IMF, credits from the Inter-American Development Bank and the World Bank, and a loan from Spain. The authorities plan to enforce a program aimed at promoting private and public sector investment, ensuring fiscal sus-tainability, and reduce the public debt burden in the medi-um term. GDP is projected to grow above 2% in 2001. On the positive side, there was no full-scale run on banks during the recent episodes of financial turbulence in the emerging markets. And, in aggregate, there was neither a loss of deposits, nor a loss of credits when looking at the annual data – although their growth has slowed markedly. Even though these recent crises precipitated the recession, there was neither international, nor domestic capital flight from the banking system. The result of a financial stress test indicates that Argentinean banks appear to be well insulat-ed from the interest rate risk. In April 2000, they were found able to withstand a flight of deposits of an amount twice as large as in 1995 [IMF, 2000]. Coming back to the Tequila crisis, the main factors responsible for spreading out the 1995 crisis seem to be, in their majority, addressed. However, there is still a need to further deepen financial markets, as well as to broaden the role of the private sector. As the developments that took place in 1999–2000 showed, fiscal-side reforms need to be addressed, if the country wants to emerge on a sus-tained growth path while maintaining the currency board regime. Similarly, the labor market reforms should be implemented to allow for more flexibility. 2.5. Conclusions There is evidence that the 1995 Argentine financial crisis coexisted with weak credibility of the currency board and risk-averse investors. A calibration of a model of contagious currency crisis to Argentine data done by Choueiri (1999) shows that if we are to believe that investors were sufficient-ly risk-averse, this financial turmoil could be attributed to the Tequila effect from the Mexican devaluation alone. Moreover, the economic fundamentals of Argentine economy did not matter in triggering the crisis [Choueiri, 1999]. Although there were speculative attacks on the peso, Argentina did not devalue the currency. Instead, the mon-etary authorities depleted its exchange reserves, and real interest rates rose. This situation matches the definition of a currency crisis given by Eichengreen, Rose and Wyplosz (1994). The currency crisis occurred, although there was no change in the nominal exchange rate. As a result of cur-rency pressures, the bank runs followed. The characteristics of the 1995 Argentine crisis are perfectly captured by the second-generation theoretical models of the currency crises. Speculations about the pos-sible devaluation of the Argentine peso increased the probability of this devaluation. The authorities were then facing a choice between short-term and long-term eco-nomic goals. The government had reasons both to aban-don the peg and to defend it, but the latter only at a cer-tain cost. Finally, the currency pressure was not directly related to economic fundamentals. The interesting question is what would happen if the monetary authorities decided to devalue the peso. First of all, it should be noted that Argentina, with its high exter-nal debt stock, was financially fragile. Large real deprecia-tion would have surely risen country risk premium. From this point of view, if the devaluation had taken place in 1995, it would have had destabilizing effects, to the finan-cial system in particular. On the other hand, as the real exchange rate was depre-ciating gradually following the crisis, real interest rates stayed high for long, thus adversely affecting investment and output. It is then tempting to assume that if a country had decided to devalue, the current output and employment would not have fall as much as they did in 1995. However, is it a possible outcome? Firstly, it should be remembered that when the crisis hit, it aggravated the already existing problems in the banking sector, and in the whole economy in general. As the experience of some of the East Asian countries indicate, devaluation does not have to be an immediate remedy, when the real-economy problems lie in the lack of transparency of the banking and enterprise sec-tor, even when domestic financial market is relatively well developed. There is also a problem of an initial overshoot-ing. And output may fall as well in such situation. Secondly, Argentine GDP started to rise during the year following the crisis, and its growth rate has been impressive. It should be remembered, that the Argentine economy has been grow-ing, on average, by 4.7% during the period 1991–1999 notwithstanding two recessions [Pou, 2000]. To sum up, it is hard to believe that the Argentine eco-nomic performance would have been better if the authori-ties devalued the currency in 1995. The credibility of the anti-inflationary policy would have been destroyed and the country risk would have been much higher, indicating high-er vulnerability to subsequent external shocks. Besides, there were positive changes in the Argentine banking sys-tem brought about by the crisis and by the decision to con-tinue the existing exchange rate policy. It is also doubtful, whether the real economy response would be much differ-ent than it was in 1995. CASE Reports No. 39
- 35. 36 Marek D¹browski (ed.) CASE Reports No. 39 Appendix: Chronology of the Argentinian Crisis December 1994 Devaluation of Mexican peso December 1994 - February 1995 International: growing perception about Argentina country risk by international investors; outflow of portfolio capital; prices of the Argentine debt start falling Domestic: shift to q uality in the b anking sector (deposit portfolio reallocation towards dollar deposits and larger banks); fears that the fixed exchange rate regime may be abandoned February 1995 BCRA (central bank) slightly changes its charter to allow more flexible use of rediscounts to aid banks End of February 1995 Rumors that the authorities are contemplating suspending convertibility of deposits; Full scale run on deposits 1-22 March 1995 All banks losing deposits The authorities more actively helping banks – BCRA losing 41% of its foreign reserves, reducing the monetary base coverage by 20% Creation of the two Fiduciary Funds to facilitate mergers and acquisitions between private banks and to facilitate privatization of small provincial banks April 1995 Amendment in the central bank charter which allows more flexible help in providing liquidity to troubled banks in an emergency situation March-May 1995 Agreement with IMF about significant financial support and the announcement of a new fiscal package Deposits of largest banks stop falling May-December 1995 BCRA rebuilds its exchange reserves and introduces new set of prudential measures Falling interest rate spreads Mergers and acquisitions in the banking system Gradual recovery of deposits
- 36. 37 The Episodes of Currency Crisis in Latin... References Antczak, R (2000). Theoretical Aspects of Currency Crises. CASE Studies and Analyses Vol. 211. BCRA (2000). Main Features of the Regulatory Frame-work of the Argentine Financial System. Banco Central de la Republica Argentina: Buenos Aires. Burdisso, T., and D'Amato, L. (1999). Prudential Regu-lations, Restructuring and Competition: the CASE of the Argentine Banking Industry. BCRA Working Paper No. 10 Burdisso, T., D'Amato, L., and Molinari, A. (1998). The Bank Privatization Process in Argentina: Towards a More Efficient Banking System? BCRA Working Paper No. 4. Caballero, R. J. (2000). Macroeconomic Volatility in Latin America: A View and Three Case Studies. NBER Working Paper No. 7782. Catao, L., and Falcetti, E. (1999). Determinants of Argentina's External Trade. IMF Working Paper No. 99/121 Catao, L. (1997). Bank Credit in Argentina in the After-math of the Mexican Crisis: Supply or Demand Con-strained? IMF Working Paper No. 97/32. CEP (1999). Review of the Real Economy No. 22 (Jan-uary- February). Centre for Production Research at the Industry, Trade, and Mining Secretariat. Buenos Aires: Min-istry of the Economy and Public Works and Services. Céspedes, L. F., Chang, R., and Velasco, R. (2000). Bal-ance Sheets and Exchange Rate Policy. NBER Working Paper No. 7840. Choueiri, N. (1999). A Model of Contagious Currency Crises with Application to Argentina. IMF Working Paper No. 99/29. Choueiri, N., and Kaminsky, G. (1999). Has the Nature of Crises Changed? A Quarter Century of Currency Crises in Argentina. IMF Working Paper No. 99/152. D'Amato, L., Grubisic, E., and Powell, A. (1997). Con-tagion, Bank Fundamentals or Macroeconomic Shock? An Empirical Analysis of the Argentine 1995 Banking Prob-lems. BCRA Working Paper No. 2. Eichengreen, B., Rose, A., and Wyplosz, Ch. (1994). Speculative Attacks on Pegged Exchange Rates: An Empir-ical Exploration with Special Reference to the European Monetary System. NBER Working Paper No. W4898. Garcia-Herrero, A. (1997). Banking Crises in Latin America in the 1990s: Lessons from Argentina, Paraguay, and Venezuela. IMF Working Paper No. 97/140. Goldberg, L., Dages B.G., Kinney, D. (2000). Foreign and Domestic Bank Participation in Emerging Markets: Lessons from Mexico and Argentina. NBER Working Paper No. 7714. Hanke, S., and Schuler, K. (1999). A Dollarization Blue-print for Argentina. Friedberg's Commodity and Currency Comments Experts' Report. Special Report, February 1st. Toronto: Friedberg Mercantile Group. CASE Reports No. 39 IMF (2001). IMF Approves Augmentation of Argentina's Stand-By Credit to US$14 Billion and Completes Second Review. Press Release No. 01/3. January, 12. IMF (2000). Argentina: 2000 Article IV consultation and First Review Under the Stand-By Agreement, and Request for Modification of Performance Criteria – Staff Report and Public Information Notice Following Consultation. Staff Country Report No. 00/164. December. IMF (1999). IMF Concludes Article IV Consultation with Argentina. Public Information Notice No. 99/21. IMF (1998). Argentina: Recent Economic Develop-ments. Staff Country Report No. 98/38. IMF (1996). IMF Approves Stand-By Credit for Argenti-na. Press Release No. 96/15. April, 12. IMF (1995). IMF Approves Extension, for Fourth Year, or EFF Credit for Argentina. Press Release No. 95/18. April, 6. Jakubiak, M. (2000). Design and Operation of Existing Currency Board Arrangements. CASE Studies and Analy-ses No. 203. Pautola, N., and Backé, P. (1998). Currency Boards in Central and Eastern Europe: Past Experiences and Future Perspectives. Focus on Transition No. 1. Pou, P. (2000). Argentina's Structural Reforms of the 1990s. Finance and Development. IMF Quarterly Maga-zine, Vol. 37, No. 1 (March). Ramos, A. (1998). Capital Structure and Portfolio Decomposition During Banking Crisis: Lessons from Argentina 1995. IMF Working Paper No. 98/121. Data Sources Argentine Statistical Office at www.indec.mecon.ar Banco Central de la Republica Argentina at www.bcra.gov.ar Bolsa de Comercio de Buenos Aires at www.bcba.sba.com.ar IMF (2000). International Financial Statistics CD-ROM IMF (1999). Public Information Notice No. 99/21. National Securities Commission at www.nsc.gov.ar The World Bank (1999). World Development Indica-tors CD-ROM. The World Bank (1998). Global Development Finance. World Bank: Washington.
- 37. 39 The Episodes of Currency Crisis in Latin... Part III. The 1997 Currency Crisis in Thailand by Ma³gorzata Antczak 3.1. Introduction The financial turmoil that erupted in Thailand in 1997 did not fit into any group of models of financial crises exist-ing in the economic literature at that time. It is just recent-ly when researches tried do develop the so-called third generation models. The Thai experience is an example which confirms that financial crises occur when macroeco-nomic as well as microeconomic fundamentals experience vulnerabilities. This paper seeks to explore the country-specific factors lying behind the Thai financial crisis. It provides analysis of macroeconomic and microeconomic roots of the crisis. It shows that while macroeconomic imbalances played an important role, the close relationship among banks, corpo-rations and the government created additional problems, which resulted in many bankruptcies and led to a sharp and unexpected economic downturn. The financial crisis con-tributed to a sharp contraction in domestic demand and activity. Also, the paper describes the sequence of the crisis and its management. Having relatively strong macroeconomic fundamentals (excluding a deteriorating current account balance, falling investments and some foreign exchange reserve indicators) the Thai authorities did not face any dramatic external shock as in the second-generation models. In Thailand, as in all Asian countries there was a boom-bust cycle in segments of the asset market (stocks, land prices, and real estate) preceding the currency crisis. Starting from the late eighties, prudent macroeconomic policies have supported a period of rapid economic growth and price stability in Thailand. However, in recent years the combination of a fixed exchange rate (which was linked to a basket of other currencies but with a strong dominance of the U.S. dollar), an increasingly open capital account, and impres-sive economic growth, attracted short-term capital inflows. These inflows were often channeled to over-invested sectors CASE Reports No. 39 due to weak prudential regulations in the banking sector enhanced by risky investments and poor corporate gover-nance [1]. The huge amount of short-term investments left the economy vulnerable to sudden shifts and external shocks. These began to materialize in 1996 as a sudden drop in exports led to a high current account deficit. At the same time, slowing economic activity and a weakening in the financial position of banks and finance companies led to debt-servicing difficulties and an increase in non-performing loans (NPLs). Starting from May 1997, the Thai currency market was destabilized by a series of currency attacks of increasing intensity. The Thai authorities attempted to defend the baht by increasing short-term interest rates and intervention in the market. As a result, the Bank of Thailand reserves were depleted, to significant extent, and the baht depreciated sharply. From the beginning of the crisis, economic policies have been progressively strengthened through: suspension of unlivable finance companies, expenditure cuts in the central government budget, and an increase in the central bank interest rates. Probably the most important action was a change in the Thai exchange rate regime, effective on July 2, 1997, from the so-called fixed but adjustable peg to a managed float. Building on these steps, the govern-ment developed a comprehensive medium-term econom-ic policy package, which was implemented with the help of the IMF. It was focused on the stabilization of the curren-cy and strengthening of the financial system. 3.2. The Way to the Crisis Simplifying the classification, the economic fundamentals can be divided into two broad categories: macro and micro-economic. At the onset of the crisis, macroeconomic fun-damentals in Thailand remained relatively sound and did not show many signs of vulnerability. [1] Recent literature, which emphasizes weak banking and financial sectors as one factor in currency crisis, includes Chau-Lau and Chen (1998), Chang and Velasco (1998), Krugman (1998), and Marshall (1998).
- 38. 40 Marek D¹browski (ed.) CASE Reports No. 39 Table 3-1. Basic macroeconomic indicators for Thailand In the mid 1980s, Thailand's economy embarked on a decade of rapid economic growth. From 1981 through 1986 growth had averaged 5.5 percent. But from 1987 through 1995 the growth rate almost doubled, averaging close to 10 percent per annum. The acceleration of eco-nomic growth was primarily investment-led and the Thai economy experienced a significant shift in the composition of production. Thailand became a more industrial econo-my while the agricultural share of GDP fell by half from 1980 to 1996. Manufacturing production and non-tradable sector of construction, finance and real estate offset this. The expansion of investment provided the counterpart for these changes. In late 1980s, investment growth rates exceeded 20 percent per annum, almost doubling the growth rate of the economy. In the 1990s investment growth rates were rising more in line with overall GDP rates of growth and the share of investment stabilized at the level of 40 percent of GDP. In the early 1990s, inflation measured by CPI was under control and stabilized at the level around 5.8 per-cent per annum. The price stabilization led to a gradual decline in nominal interest rates. Demand for high-pow-ered money in Thailand was relatively stable and broad money monetization was increasing (Table 3-1), amounting to 86 percent of GDP in 1997. The central budget indicat-ed a surplus of 2.4 percent of GDP in 1996. The unem-ployment rate was very low at 1.1 percent during 1993–97. Investment and saving rates were high, averaging at the level above 30 percent of GDP. The exception to the favorable economic outlook was a deteriorating current account deficit, which rose to 7.8 percent of GDP in the years 1995–96 and was mostly covered by short-term portfolio investments. The deficit reflected private-sector demand for foreign capital. In the 1980s Thailand's priority was large net capital inflow promotion, through tax and institutional reforms (see below) while concurrently developing its financial markets [2]. Large positive interest rate differentials and a pegged exchange rate supported this policy. This regime provided a guarantee to short term investors that they can make a quick exit at little or no cost. Authorities' measures to attract foreign capital included: – Elimination of restrictions on foreign investments, – Elimination of most barriers on foreign ownership of export oriented industries [3], – Granting tax incentives to foreign mutual funds and investments in the stock market – Creation of closed-end mutual funds, [2] In 1992, the authorities approved the establishment of the Bangkok International Banking Facility (BIBF), which greatly eased access to foreign financing and expanded short-term inflows. [3] Some limitations on foreign ownership were retained in non-export oriented industries and on the maximum foreign ownership of companies listed on the stock exchange. 1981-1994 1994 1995 1996 1997 1998 1999 Real GDP Growth 8.5 9.9 8.9 5.9 -1.8 -10.0 4.0 CPI Inflation average 3.8 5.1 5.8 5.8 5.9 8.5 5.8 Fiscal Balance to GDP Ratio* -0.6 1.9 3.0 2.4 -0.9 -2.4 -1.1 Private Sector Credit to GDP Ratio 90.9 97.5 100.0 122.5 115.1 Current Account to GDP Ratio 5.3 -5.5 -7.8 -7.8 -2.0 12.7 9.0 Financial Account to GDP Ratio 8.4 13.0 10.5 -11.3 -13.0 Gross National Savings to GDP Ratio 35.5 35.6 33.2 31.9 31.1 30.1 Gross Domestic Investments to GDP Ratio 41.38 43.3 43.7 33.6 19.0 20.4 Broad Money Monetization** (in percent) 71.0 72.2 75.4 85.9 99.0 Source: Own calculations on the basis of data from the IMF, The World Bank * Central budget balance (percentage of fiscal-year GDP) ** Monetization of an economy is defined as a ratio of a measure of money to an annualized value of GDP at current prices. Table 3-2. Net capital inflow to GDP Ratio (in percent) and nominal interest rate differential in Thailand 1992 1993 1994 1995 1996 1997 1998 Net Capital Inflow 8.5 8.4 8.4 13.0 10.5 -11.3 -13.0 Differential of Interest Rate 5 5.5 4 4.5 4.8 3.4 -0.4 Source: Own calculations on the basis of data from IFS
- 39. 41 The Episodes of Currency Crisis in Latin... – Establishing rules for foreign debenture issues by Thai companies, – Reduction of taxes on dividends remitted abroad. The promotion of capital inflows combined with a rapid-ly growing economy contributed to very substantial net cap-ital inflow to Thailand in the range of 9–13 percent of GDP between 1989 and 1995. Between 1991–1996, net capital inflows amounted to 85 billion U.S. dollars. The composition of capital inflows evolved, as a growing proportion of the net inflows had short-term nature (port-folio and other investments), reaching 95 percent of the total in 1995 (Figure 3-1). Net direct investment inflows played a bigger role at the beginning of 1990s reaching its peak in 1993 at the level of 15 percent of total inflows. However, net portfolio inflows became more important in 1994, as a result of the mentioned subsequent reforms of the Thai stock markets and the large interest rate differen-tial. The contribution of foreign direct investments to a total capital inflow stabilized at the level of 5–7 percent in 1994–1996. The continuation of short-term capital inflow kept the overall balance of payments in surplus helping to fuel investments and economic growth. Net capital inflow used to be partially sterilized. In practice, the sterilized intervention maintained high domestic interest rates and a large wedge between domestic and international interest rates. It attracted foreign capital even more. 3.2.1. Macroeconomic Signs of Vulnerability There were already signs of Thailand's vulnerability before the crisis. The main macroeconomic indicators sig-naling a crisis were: the level of official international reserves, deterioration in investment, high current account deficit and excessive credit expansion. Although the official foreign exchange reserves increased by 183 percent between 1990 and 1996, they were not suf-ficient to protect against speculative attacks in the context of an increasing current account deficit and short-term external debt payments. Traditionally, three months of imports' coverage is con-sidered a minimum threshold of official foreign exchange reserves. The East Asian countries, apart from Korea, recorded relatively safe reserves to imports ratio in the 1990s (Table 3-3). This relates particularly to Thailand. CASE Reports No. 39 Although the level of reserves did not indicate the danger of a currency crisis, the reserves to short-term debt ratio was much less favorable. In this respect, Thailand represented one of the weakest records [Jakubiak, 2000], with interna-tional reserves below the country's short-term debt obliga-tions. Three months before the crisis, the ratio of reserves to short-term external debt indicated the 1.1 coverage of short-term obligations, two months before the crisis it fell below 1, and in July 1997 it amounted only to 0.7. The value of this indicator in Thailand showed that reserves did not exceed official and officially guaranteed short-term debt in the pre-crisis period. Another important indicator is the ratio of international reserves to base money. Thailand recorded relatively safe levels of backing in the years prior to the crisis, but substan-tially lower during the last months preceding the crisis. The ratio of reserves to base money was falling from about 2.5 in July 1996 to 1.5 in the time of crisis in July 1997 what indi-cated the increasing financial fragility of the economy. Over-investment and excessive capital accumulation accompanied the years of rapid growth in Thailand. In 1996, investment growth slowed, falling to little less than 7 per-cent compared with an average of more than 10 percent growth per annum during the previous five years. Private investment grew by only 3 percent, reflecting signs of excess capacity and earlier over-investment. In 1997, the overall investment to GDP ratio declined to 33 percent from 43 percent of GDP in the previous year. The declining invest-ment contributed to the output contraction during the onset of the crisis. An excessive expansion of the non-tradable sector, par-ticularly in the real estate and construction activities played a crucial role in the investment break down. In the pre-cri-sis period all sectors of the economy were growing rapidly. Private investment in construction grew rapidly during 1990–94 ant it took up to 50 percent of total fixed invest-ments. The public investment in construction grew at 25 percent on average, twice higher than the private invest-ment in construction. Much of the office construction in the commercial sector was built not by professional property developers, but by companies itself and for their own use. In the pre-crisis period, overall private credit was grow-ing rapidly (Figure 3-4). In particular, this related to loans to the real estate and housing projects carrying out by financial companies. Many of these loans were turn to non-perform- Table 3-3. Reserves in months of imports in Asian Countries 1992 1993 1994 1995 1996 1997 1998 Indonesia 3.3 3.0 2.4 2.6 3.1 5.5 5.4 Korea 1.9 2.0 1.7 2.0 2.4 3.1 4.1 Malaysia 3.3 3.4 4.3 3.3 3.3 4.3 3.7 Thailand 4.5 4.1 4.0 4.8 6.8 8.0 6.8 Source: Own calculations on the basis of data from IFS
- 40. 42 Marek D¹browski (ed.) CASE Reports No. 39 ing, and financial companies lost their solvency, what indi-cated clear evidence of over-expansion of property sector credit. In the period of 1985–95, Thai exports grew on average by 23 percent per annum. For much of this time, the growth exceeded the average of its regional competitors (Indonesia, Korea, Malaysia, and the Philippines). However, rapid export growth came to an abrupt halt in 1996 – it was a seri-ous warning that the Thai economy was vulnerable to a dis-ruption. Export growth rates declined sharply and turned negative in both value and volume terms. The main external factors behind the fall in exports in 1996 were declining competitiveness, slower demand growth in partner coun-tries, and real exchange rate appreciation starting in early 1995. The appreciation of CPI based real exchange rate of baht versus U.S. dollar was not very much visible, because of the peg to a dollar and CPI inflation in Thailand was not significant in the pre-crisis period. In 1996, however, the yen depreciated strongly in nom-inal terms against the U.S. dollar while Japan played a major role in Thai trade (first place in imports and second in exports). Thus, the baht appreciated by 8.5 percent in real terms between end-1994 and end-1996 against the yen which reflected a loss in international competitiveness in the major export market. The result was sharp contraction of Thai exports and a further increase in the current account deficit, to almost 8 percent of GDP in 1996. However, structural (internal) factors also played a role in the export slowdown, including slow adjustment toward more capital-intensive and high-tech products. Thailand lost market share in labor-intensive products, such as garments and footwear, to lower-wage countries, including China and India. The labor-intensive exports declined by 21 percent in 1996. Meanwhile, high technology products such as com-puters, electronic motors faced increasing competition of Korea, Singapore, Taiwan, Malaysia, and Hong-Kong, show-ing greater convergence in their export structures. Export volumes of these goods fell by 8 percent in 1996. The large current account deficit, coupled with changes in the export structure, and the perception that a currency is overvalued led to a balance of payments-type crisis. The industrialization strategy implemented in Thailand fuelled by the financial system liberalization and massive cap-ital inflow resulted in rapid increases in private sector bor-rowing. Most loans were short-term and denominated in foreign currency. The Thai borrowers preferred borrowing in U.S. dollars at short-term interest rates, even to finance long-term projects because it was cheaper than borrowing in baths. Thailand's foreign debt rose to the level of 50 per-cent of GDP, of which 80 percent was private-sector bor-rowing. The public sector borrowing played a minor role. From 1995 to 1996, the growth rate of private credit averaged well above 15 percent, or twice the rate of real GDP growth. In the middle of 1997, the private credit expansion accelerated, reaching its pick in January 1998 of 25 percent per annum. The fact that raised loans were invested in the risky busi-ness of declining rate of return (for discussion on efficiency of investments and profitability of the corporate sector – see the next section) led many of them to become non-per-forming, putting an extraordinary burden on the banking sector. 3.2.2. Microeconomic Signs of Vulnerability Apart from macroeconomic indicators, the weakness of both the financial and corporate sectors appeared to be cru-cial in determining the crisis development in Thailand. The difficulties faced by Thai corporations have their roots in the over-investment that took place in the years leading to the crisis. From 1987 to 1995, growth of real fixed investment averaged almost 16 percent, as compared with a real GDP growth rate averaging 10 percent. As noted in the previous section, the acceleration in investment took place in the late 1980s when investment growth rates increased at the rate of 20–30 percent per annum. The result was a rise in the investment-to-GDP ratio to around 40 percent. In the first-half of the 1990s, investment growth moved in line with a real GDP growth. However, with the capital-output ratio steadily increasing, it became inevitable that diminishing returns to capital would put under question the sustainability of this particular investment-led growth strategy. One clear symptom of this was the decline in capacity utilization before the crisis. This picture of over-investment and declining real rates of return could also be seen in the financial statements of Thai corporations. From 1994 to 1997, the value of assets grew significantly in non-tradable sectors such as construc-tion, communication, and property development. However, Table 3-4. Performance of non-financial private corporations in Thailand 1994 1995 1996 1997 1998 1999 Q2 Total Loans of Firms Billion Bahts 776 1038 1333 2092 1816 1780 Profits* over Interest Expenses (%) 6.1 4.4 3.5 1.0 1.3 1.9 Profits* over Liabilities (%) 24.3 18.9 15.3 7.4 9.5 13.6 Debt to Equity Ratio 1.5 1.7 2.0 4.6 2.8 2.9 Source: Stock Exchange of Thailand. Merrill Lynch * Profits are defined as earnings before interests, taxes, depreciation, and amortization.
- 41. Table 3-5. Non-performing loans at domestic commercial banks 1995–99 (percent of total loans) [4] These numbers exclude debt instruments such as bills of exchange and commercial papers, and are calculated using an end-1997 exchange rate 43 The Episodes of Currency Crisis in Latin... the growth in asset values was not accompanied by equiva-lently high growth of earnings. The return on assets fell by roughly one-third from 1994 to 1996, and as a result stock prices started to go down in the second half of 1996. The consolidation of companies' ownership in Thailand was very strong. In the ten largest non-financial private sec-tor firms, the top three shareholders owned on average as much as 45 percent of the outstanding shares. The desire of the owners to retain control of their conglomerates led them to use debts to finance their expansion. Large capital account liberalization facilitated this expansion, by increas-ing the supply of funds to corporations. As a result, by end- 1997 the corporate sector held debts of approximately 153 billion U.S. dollars (more than 150 percent of GDP), where 123 billion U.S. dollars was financed by domestic banking system, and 30 billion U.S. dollars from abroad [4]. The result of this debt-financed expansion was an increase of debt-equity ratio of corporations from 1.5 in 1994 to more than 2 in 1997. Several years of strong economic growth – underwrit-ten by rapid credit expansion and large capital inflows – exposed underlying structural weakness of the banking sec-tor, especially under a poor regulatory framework. The financial sector grew rapidly in the 1990s, driven by expan-sion of finance companies and the banking sector. The investment-led growth of the Thai economy was largely debt financed, which was reflected in the rapid growth of banks' assets. Simultaneously, softening of licensing require-ments for finance companies contributed to their expan-sion. Finance companies tended to focus more on con-sumer and real estate financing, while banks leaned more toward investment financing, particularly in the manufactur-ing sector. Banks recorded high profit and their share prices boomed in the period of 1993–97. The interest rate spreads averaged about 6.3 percentage points in this period, which supported bank profitability with return on assets averaging 1.6 percent in this period. However, belying this positive picture, indicators of underlying weakness in the finance sector started to appear. Both banks and finance companies were heavily exposed to the property sector but the exposure was most acute in the case of finance companies. In 1996, investments of finance companies in real estate and construction amounted to 35 percent of the total credit, while commercial banks invest-ed around 20 percent of their total credit in real estate and construction. This was particularly worrisome in light of the increasing evidence of over-investment in the property sec-tor. Additionally, substantial share of finance companies' credit was being channeled into the stock exchange, leading to a rapid growth of risk. Already in 1996, finance companies started to exhibit liquidity problems, illustrated by increasing strains of accrued interest. Although the level of overall non-per-forming loans was relatively low (12 percent of total at the end of 1996), accrued interests in several banks were high-er than average and growing, suggesting that the true NPLs were actually higher and increasing. The first clear sign of trouble occurred in March 1997 when the Bank of Thailand and the Ministry of Finance announced that ten, as yet unknown, finance companies would need to raise capital. In early 1997, more severe liquidity problems emerged as the economy slowed. Public confidence in finance com-panies eroded as solvency problems occurred. In May 1997, the Bank of Thailand suspended 16 insolvent finance com-panies and announced that its creditors are expected to bear part of their losses. During the spring of 1997 the finance sector began to experience a large-scale deposit withdrawals, which lead to massive and secret liquidity sup-port from the authorities to 66 finance companies. This sup-port peaked in August 1997, reaching altogether about 10 billion of U.S. dollars (about 8 percent of 1997 GDP). The deposit withdrawal represented a flight to quality by house-holds and businesses moving their savings from finance companies to large commercial banks. The banking sector in Thailand also started to show weaknesses. Bank capital was substantially overstated, reflecting reliance on collateral of uncertain value. Anecdo-tal evidence suggest that banking practices focused heavily on "name" based lending, relying on personal guarantees and collateral to secure loans. These transactions were mostly valued not by independent appraisers what had its picture in bank balance sheets and income. Indeed, while reported NPLs of banks amounted to 11.6 percent of assets, this figure largely included loans that had been non-performing for one year and over, and did not capture the most recent deterioration in asset quality. Many private market analysts estimated NPLs to be at least 15 percent of total banks' loans at that time. The subsequent deterioration of the corporate balance sheets and adverse effects of the depreciation led to a rapid of 1 U.S. dollar = 47 Bahts CASE Reports No. 39 1995 1996 1997 1998 1999Q2 NPLs 8 10 22 48 51 Source: Thailand: Selected Issues. IMF Staff Country Report No. 00/21
- 42. 44 Marek D¹browski (ed.) build-up in non-performing loans and decapitalization throughout the whole fragile financial system. 3.3. The Crisis In 1996 economic growth, exports and investment dete-riorated in the face of an appreciating real exchange rate. The current account was in deficit, interest rates were high, and inflation was increasing. Moreover, serious weakness appeared in the financial system due to exposures to the property sector and inadequate loan provisioning. High interest rates to counteract capital outflows aggravated the solvency and liquidity position of many banks and finance companies and resulted in intervention by the authorities to support the financial system. In early 1997, the baht came under pressure as traders began to doubt the viability of its peg to the dollar. The Thai currency was subject to several speculative attacks in the first-half of 1997 and the central bank intervened actively on foreign exchange markets and imposed capital controls in May 1997. As a result, the official international reserves fell by almost 13 billion U.S. dollars in the first eight months of 1997 (by 38 percent). The fall was continued and the low-est level was recorded in February 1998 when reserves reached the 1994 level. On July 2, 1997, faced with a banking crisis, a run on the currency, and large reserve losses, the Bank of Thailand floated the baht. The currency fell 10 percent immediately and then weakened further. The bath depreciated by an additional 22 percent against the U.S. dollar during July. On July 28, 1997 Thailand formally sought IMF assis-tance. On August 20, 1997 the IMF announced an assistance package of 4 billion U.S. dollars and established a list of reforms that the country was obliged to implement [5]. 3.3.1. Managing the Crisis. The IMF Intervention in Thailand On August 20, 1997, the IMF's Executive Board approved a 34-month Stand-By Agreement with Thailand, amounting to 4 billion U.S. dollars (equivalent of SDR 2,900 million or 505 percent of quota). The adjustment program was aimed at stabilizing the exchange rate and reducing the current account deficit through control of domestic credit, and limiting the rise in inflation. Key elements of the policy package included fiscal policy measures, and financial sector restructuring, including closure of insolvent financial institu-tions, consolidation of banks, and non-performing loan man-agement. The program provided for three reviews to be complet-ed during the first year (program targets for September 1997, December 1997 and June 1998). Thereafter, the pro-gram was to be subject to two twice-yearly reviews (pro-gram targets for end-December 1998, end-June 1999, and end-December 1999). Upon approval of the program, Thai-land drew 1.2 billion U.S. dollars from the IMF and received a further 4 billion U.S. dollars from bilateral and multilateral sources. Additional financing was pledged by the World Bank and the Asian Development Bank (2.7 billion U.S. dollars), which also provided extensive technical assistance. Financial sup-port by Japan and other interested countries (10 billion U.S. dollars) was pledged at a meeting in August, hosted by Japan. Bilateral financing has been disbursed in parallel with the purchases from the IMF. Total official financing of the sta-bilization program amounted to over 17 billion U.S. dollars (Table 3-6). In the second half of 1997, the baht continued to depre-ciate as the contagion in Asia began. While macroeconomic policies were on track and nominal interest rates were raised, market confidence was adversely affected by delays in the implementation of financial sector reforms, and poli-tical uncertainty. By the time of the review under the special emergency procedures (on October 17, 1997), there were also signs that the slowdown of economic activity would be more pro-nounced than anticipated. And in fact it was. The further depreciated exchange rate put pressure on increase in inter-est rates, and it resulted in a much sharper decline in private investment and consumption than originally anticipated. A new government took office in mid-November 1997. The new economic team headed by Prime Minister Chuan reconfirmed the commitment to the adjustment program. To help stabilize the foreign exchange market, the program Billion U.S. dollars Percent of GDP IMF 4.0 3.0 Asian Development Bank and World Bank 2.7 2.0 Other 10.5 7.0 Total package 17.2 12.0 Source: "IMF-Supported Programs in Indonesia, Korea, and Thailand. A Preliminary Assessment". IMF, Washington DC 1999 CASE Reports No. 39 Table 3-6. Official financing of stabilization program [5] A detailed chronology of the crisis in financial sector is given in the Appendix 1
- 43. 45 The Episodes of Currency Crisis in Latin... was strengthened at the first quarterly review (on Decem-ber 8, 1997). The new government was determined to take a number of additional measures to support the policy pack-age. With weakening economic activity, constraining rev-enues, additional fiscal measures were introduced to achieve the original fiscal target for 1997/98. Reserve money and net domestic assets of the Bank of Thailand were to be kept below the original program limits. As a result, indicative interest rates were raised and a specific timetable for financial sector restructuring was announced. In early February 1998, the baht began to strengthen against the U.S. dollar as improvements in the policy setting revived market confidence. Growth projections, however, were marked down further. Contracting domestic demand helped to keep inflation under control and contributed to a larger-than-expected adjustment in the current account. The stabilization program was revised significantly at the time of the second quarterly review (on March 4, 1998). Under the revised program, monetary policy continued to focus on the exchange rate, with interest rates to be main-tained at high levels until evidence of sustained stabilization emerged. Fiscal policy shifted to a more accommodating stance. In addition, the program included measures to strengthen the social safety net, and broaden the scope of structural reforms to strengthen the core banking system and promote corporate restructuring. The third quarterly review took place on June 10, 1998 and a marked strengthening of the baht during February- May 1998 was noticed (some 35 percent vis-a-vis the U.S. dollar from the low in January). The revised program was on track, but with real GDP projected to decline 4–5 per-cent in 1998 and inflation subdued, further adjustments were made to allow for an increase in the fiscal deficit tar-get for 1997/98 from 2 percent to 3 percent of GDP. Mon-etary policy continued to focus on maintaining the stability of the baht. While the reductions of interest rates since late March 1998 was viewed as consistent with exchange mar-ket developments, it was understood that interest rates would be raised again if necessary. Additional measures to strengthen the social safety net were planned, and the pro-gram for financial sector and corporate restructuring was further specified. The exchange rate weakened during June-July 1998 amid growing concerns about the growth outlook, and renewed signs of strains in the financial sector, where grow-ing difficulties in the corporate sector complicated restruc-turing of financial institutions. Fiscal and monetary policies had been tighter than programmed, economic activity was weaker than expected, and exports had failed to pick up. The large adjustment in the current account (projected to amount to over 10 percent of GDP) reflected a sharp con-traction of imports. The fourth quarterly review (on September 11, 1998) focused on adapting the policy framework to support the recovery without sacrificing stabilization gains. With output now projected to decline by 6–8 percent in 1998, efforts were stepped up to utilize the scope of fiscal easing provid-ed under the program. Foreign exchange market conditions were relatively stable (in spite of the Russian crisis), provid-ing room for further lowering of interest rates. The pro-gram for financial and corporate sector restructuring was broadened significantly, and the structural reform agenda in other areas (privatization, foreign ownership, and social safety net) was strengthened. As of October 19, 1998, 12.2 billion U.S. dollars from the total financing package for Thailand (17.2 billion U.S. dollars) had been disbursed, including 3 billion U.S. dollars from the IMF and 9.2 billion U.S. dollars from other multi-lateral (World Bank and Asian Development Bank) and bilat-eral sources. During 1999 there were several additional quarterly reviews of the stabilization program. All of them were focused on revitalizing of domestic demand and on the social safety net. The overall public sector deficit was grad-ually set at the higher level (5 percent of GDP in the fiscal year 1998/99 and 7 percent in the fiscal year 1999/00. These fiscal targets accommodated reductions in revenues (of about 0.5 percent of GDP in 1998/99 and in 1999/00) from the impact of lower-than-expected nominal GDP. The flexible use of interest rate policy to maintain baht stability was reaffirmed, monthly interest rates were lowered and inflation was falling. Growing confidence has allowed inter-est rates to fall below pre-crisis levels without compromis-ing exchange rate stability. The overall balance of payments outcome was stronger than expected, as a higher current account surplus, reflecting weak domestic demand, carried over to higher than projected reserves. On May 8, 2000 the Executive Board of the IMF com-pleted the ninth, and final review under Thailand's Stand-By Arrangement. To date, under 17.2 billion U.S. dollars official financing package, Thailand has drawn 14.3 billion U.S. dol-lars from bilateral and multilateral contributors, including 3.4 billion U.S. dollars from the Fund. 3.3.2. Macroeconomic Environment after the Crisis The persistence and widening of the current account deficit, over-investment, declining rates of return on capital, and the over-expansion of the non-tradable sector pointed on macroeconomic reasons of the crisis and the need for deep adjustment in exchange rate. In the aftermath of the crisis, Thailand's real effective exchange rate depreciated by 35 percent in the second half of 1997, but subsequently recovered. As of end-1999, the cumulative real exchange rate depreciation was 25 percent. After depreciation Thai-land's external current account balance shifted from a deficit CASE Reports No. 39
- 44. 46 Marek D¹browski (ed.) of eight percent of GDP in 1996 to a surplus of more than 12 percent of GDP in 1998. Due to the fall in Thailand's terms of trade and falling dollar export prices, export vol-ume growth exceeded 8 percent per annum in 1997–98. Conversely, import volumes fell dramatically by more than 40 percent from 1996 to 1998, reflecting the weakness in domestic demand and the relative price effect of the deval-uation. Once the full extent of the weaknesses in Thailand's economy became known, including the underlying problems in the financial sector and the collapse of Thailand's interna-tional position, financial market confidence vanished. Thai-land's pre-crisis problem of persistent and excessive capital inflows was transformed into one of managing major capital outflows, with creditors refusing to rollover short-term debt. As indicated before, investment was falling sharply [6] in the first quarter of 1997 what influenced output contrac-tion. In 1998, the recession widened and real GDP declined by 10 percent. Private consumption also fell markedly. Con-sumer durables were particularly hard hit, with car sales falling to around one quarter of their pre-crisis levels. As a result of high interest rates and the reduced availability of credit, consumption fell by 13 percent in 1997 and 1998. Another factor responsible for decline in consumption was lowering personal incomes. With the general collapse in domestic demand, unemployment increased and wages declined. Since January 1998, the rate of private credit growth (adjusted after correcting for changes in valuation due to exchange rate fluctuations) declined steadily (Figure 3-4). Later the growth of credit continued to decrease and was even negative at the rate of 10 percent per annum in Janu-ary 1999. In 1999 the situation started to improve. And in the end of 1999, the credit started to grow at the rate of 5 percent per annum. However, the volume of credit was still 20 percent below the peak reached in late 1997. This sug-gests two possibilities. First, that credit-intensity of firms has fallen as firms started to rely increasingly on retained earn-ings and other sources of non-bank financing or, second, that there has been a shift in the allocation of credit [7]. After the sharp contraction in 1998, the recovery start-ed in 1999. Manufacturing production, which had already bottomed out by the middle of 1998, grew at double-digit rates through much of 1999, and by September 1999 it had surpassed its pre-crisis peak. On the demand side, lower interest rates and improving recovery prospects have stim-ulated private consumption. This trend was supported by a temporary VAT reduction, which took effect in early 1999. In 1999, Thailand's economy reached the growth rate of 4 percent, which was much more modest than in the past. In 2000, the recovery was strengthened and real GDP was expected to increase by 4–5 percent. If 2000 trends con-tinue, Thailand will recover pre-crisis levels of output and consumption per-capita by the end of 2002 (the IMF esti-mates). The major contributors to growth continued to be exports (6.3 percentage points) and private consumption (3.4 percentage points). Total factor productivity (TFP), which began to decrease well before the crisis, became negative in 1996, and bot-tomed out in 1998. However, since 1999 TFP appeared to start to grow again, helped by structural reforms and cycli-cal bounce back, and was estimated to grow from 1 to 1.4 percent in 2000 (the IMF estimates). Exports have done well and have been a key driver of the recovery. In 1999, they grew by close to 9 percent, and were set to grow by 6.8 percent in 2000 (IMF estimates). The U.S. and EU markets contributed to the pick-up in Thai exports. More recently, the exports to Japan and ASEAN countries recovered, and accounted for over 30 percent of total exports in 1999. Nonetheless, there were concerns regarding the competitive weakness of the Thai industry. Skill-intensive activities complained of shortages of high level skilled manpower, and technology-intensive activities remained largely confined to the final assembly stage of operations. More recently, technology intensive exports have increased. Recent data suggest that the growth in Table 3-7. Contribution to economic growth in the year 2000 (percent) Growth Contribution to Growth Real GDP growth (%) 4.5 4.5 Private consumption 6.4 3.4 Public consumption 4.9 0.5 Private investment 11.0 1.2 Public investment 4.5 0.5 Exports 11.0 6.3 Imports 17.0 -7.5 Source: World Bank. Thailand Economic Monitor. June 2000 [6] While investment fell across the board, investment in construction was especially badly hit, its share fell to 35 percent of the total investment CASE Reports No. 39 from 50 percent before the crisis. [7] Some firms have suspended servicing their loans, thereby "obtaining credit" by generating NPLs.
- 45. 47 The Episodes of Currency Crisis in Latin... Table 3-8. Contributions to GDP growth (in percent) imports has started to slow down. Imports fell by 17 per-cent between December 1999 and January 2000 (IMF esti-mates). Given projected growth rates of imports and exports, the current account balance was expected to generate a surplus of 7.7 billion U.S. dollars in 2000 (5.3 billion US$ of the trade balance surplus). On the capital account side, repayments by the private sector were expected to fall from 15.5 billion in 1999 to 9.5 billion U.S. dollars in 2000. This created a cushion to support potential weakness in portfolio flows and foreign direct investment. Gross official reserves increased to 34 billion U.S. dollars at the end of 1999. But in the middle of 2000, the country's foreign reserves stagnated at around 32 billion U.S. dollars [8]. This level was sufficient to cover more than three times the cash in circulation at that time (which approximated 9 billion U.S. dollars and about 400 billion Thai bahts). This level of reserves was an equivalent to 200 percent of debt maturing in the next 12 months, and 6 months of imports. A second key driver of recovery was private consump-tion. In 1999, total private consumption grew by 3.5 per-cent, recovering from sharp contraction in 1998 (-12.3 per-cent). This growth was broadly consistent with a return of consumer confidence, reflected in an increase in aggregate disposable income resulting from rising wages and higher levels of employment. Inflation remained under control in 1999 and did not create a risk for the economy. The gov-ernment set an inflation ceiling of 3.5 percent for the year 2000, what was possible to reach. However, employment data show that the recovery is still fragile. The crisis did not appear to have affected the trend in a significant way. After the onset of the crisis employment was expanding gradually but the unemploy-ment rate was falling at a very slow pace. The February unemployment rate fell from 5.4 percent in 1999 to 4.8 per-cent in 2000. While the unemployment rate appeared to be modest when compared to European countries, Thailand has no unemployment insurance system and welfare impact can be severe. However, a review of the adjustments in the labor market showed that wage reductions among less edu-cated workers were less severe compared to the educated workers, suggesting that labor markets protected the less well off (WB Monitor). Looking back over the two and half years under the Fund-supported program, the successful implementation of macroeconomic policies can be observed. All above-men-tioned indicators show that the main objectives of the pro-gram have been met. Over the medium term, the key chal-lenge will be to sustain economic recovery in the context of a heavily indebted corporate sector and continued weak-ness of financial system. The results of corporate debt restructuring are not satisfactory and this process still has some way to go. In order to complete reforms in the finan-cial sector, the speeding up of corporate debt restructuring process is also necessary. 3.4. Conclusions The reasons for the Thai financial crisis were almost exclusively internal [9]. The Thai experience shows that financial crises can erupt not only when macroeconomic but also when microeconomic indicators express vulnerabilities. The Thai crisis can be classified as representing a kind of third generation model, which theoretical backgrounds is still questionable [10]. In early 1997, Thailand faced a canon-ical balance of payments crisis when a structural misbalance between the deficit in current account and capital and finan-cial account (sources of financing) occurred. In the second quarter of 1997 the authorities defended the baht and inter-national reserves diminished what together with microeco-nomic problems led to a currency crisis. Once the full CASE Reports No. 39 Capital Quality adjusted labor TFP GDP 1995 6.2 1.0 1.7 8.9 1996 5.4 1.8 -1.8 5.4 1997 3.4 2.7 -7.9 -1.7 1998 1.8 1.3 -13.0 -10.0 1999 1.8 1.4 0.8 4.2 2000 1.9 1.6 1.0 4.5 Medium term 2.0 1.6 1.4 5.0 Source: World Bank. Thailand Economic Monitor. June 2000 [8] Chase, International Fixed Income Today, 13 September 2000. [9] The important external factor of the crisis was yen/U.S. dollar exchange rate developments. [10] Antczak (2000)
- 46. 48 Marek D¹browski (ed.) extend of the weakness in Thailand economy became known, including underlying problems in the financial sector, excessive capital account liberalization led to massive with-drawal and to full-fledged financial crisis. The Thai crisis led to a contagion effect in Asia. These developments changed investors' perception of the Asian Tigers of early 1990s. It contributed to external shocks and the Asian flu infected Korea, Philippines, Malaysia, and Indonesia. The second broad conclusion from this analysis is a fundamental need for an integrated approach to capital lib-eralization and financial sector reform. The Thai stabilization program was successful and the economy recovered. The authorities have made progress toward resolving the problems in the financial sector. Banks have raised substantial amounts of new capital, the core banking system remained in private hands, and foreign entry should stimulate competition and improvements in the technology and service. But despite the positive changes there is still a lot to be done, especially in the area of struc-tural reforms. CASE Reports No. 39
- 47. 49 The Episodes of Currency Crisis in Latin... Figure 3-1. Net Capital Flows in Millions of U.S. Dollars 25000 15000 5000 -5000 -15000 -25000 1991 1992 1993 1994 1995 1996 1997 1998 Direct investments Portfolio investments Other investments Source: own calculations on the basis of data from the IMF IFS Figure 3-2. Total Reserves in U.S. Dollars and Nominal Exchange Rate Developments in Thailand 40000 35000 30000 25000 20000 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 CASE Reports No. 39 60 55 50 45 40 35 30 25 20 Total reserves Exchange Rate (right scale) Source:
- 48. 50 Marek D¹browski (ed.) 25 20 15 10 5 0 -5 -10 CASE Reports No. 39 Figure 3-3. Current Account Structure 20000 15000 10000 5000 0 -5000 -10000 -15000 -20000 -25000 1993Q1 1993Q4 1994Q3 1995Q2 1996Q1 1996Q4 1997Q3 1998Q2 1999Q1 Exports of goods Export of services Imports of goods Import of services mln USD Source: own calculations on the basis of data from the IMF IFS Figure 3-4. Private Credit Growth Before and After the Crisis 6000 5500 5000 4500 4000 3500 3000 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 TBt -15 % NDC percentage change Claims on private sector (right scale) Source: own calculations based on IMF IFS
- 49. 51 The Episodes of Currency Crisis in Latin... Figure 3-5. Commercial Bank Profitability, 1993–98 (percent) 1993 1994 1995 1996 1997 1998 5% 4% 3% 2% 1% -1% -2% -3% CASE Reports No. 39 Return on Assets Net Interest Yield 0% -4% Source: Thailand: Selected Issues. IMF Staff Country Report No. 00/21
- 50. 52 Marek D¹browski (ed.) CASE Reports No. 39 Appendix 1: Chronology of the Thailand's Currency Crisis March 1997 First explicit sign of trouble. BoT and MoF announce that 10 as yet unknown finance companies would need to raise capital. March-June, 1997 Public confidence in finance companies erodes. Deposit withdrawals. Massive and secret liquidity support from the authorities to 66 finances companies. June 1997 BOT suspends 16 finance companies and announces that their creditors are expected to bear part of companies’ losses. July 2, 1997 The baht is floated, then it depreciates by 32 percent against U.S. dollar during July. In the context of IMF program negotiations, BoT and MoF issue a joint statement detailing measures to strengthen confidence in the financial system. Additional 42 finance companies have their operations suspended (altogether 58 out of 91 finance companies) and are given 60 days to present rehabilitation plans to the authorities. August 1997 Government announces blanket guarantee to banks and remaining finance companies backed by unlimited FIDF support (in baht). August 14, 1997 First Thailand’s IMF Letter of Intent. August 20, 1007 The IMF Executive Board approves a three-year Stand-By Arrangement, amounting to 4 billion U.S. dollars (505 percent of quota). October 17, 1997 Emergency Financing Procedures by the IMF. November 25, 1997 Second Thailand’s IMF Letter of Intent. MoF announces a closure of 56 finance companies. December 1997 BoT intervention at Bangkok Metropolitan Bank - capital of existing shareholders is written down, management is changed, and the bank is recapitalized by authorities vie debt-equity swap. December 8, 1997 First quarterly review of the policy package. Strengthening of the program, implementation of additional fiscal measures. Indicative range for interest rates is raised, and a specific timetable for financial sector restructuring is announced. First Bangkok City Bank and Siam City Bank are intervened and dealt with the same fashion as BMB in the previous month. These three banks account for about 10 percent of banking system deposits. A new state-owned commercial bank, Radanasin Bank, is established in order to take control over the higher-quality assets. January 1998 A majority stake in Thai Danu Bank is acquired by foreign investors (Development Bank of Singapore). Baht begins to strengthen against the U.S. dollar as improvements in the policy settings revived market confidence. Contracting domestic demand helps to keep inflation in check and contributed to larger-than-expected adjustment in the current account. February 24, 1998 Third Thailand’s IMF Letter of Intent. February – May 1998 Strengthening of baht. Agreement on compensation reached with creditors of 42 finance companies under March 1998 rehabilitation program. Cautious reduction of interest rates viewed as consistent with exchange rate developments. March 4, 1998 Second quarterly review of the policy package. Under the revised program, monetary policy continues to focus on the exchange rate stabilization, with interest rates to be maintained high until evidence of a sustained stabilization emerged. The program includes measures to strengthen financial sector. March – April 1998 Banks start to recapitalize with many foreign deals. New loan classification and provisioning rules are introduced. May 1998 Additional 7 finance companies are intervened and merged with KTT (a large government owned finance company). May 26, 1998 Fourth Thailand’s IMF Letter of Intent. June 1998 Bank of Asia acquired by ABN-AMRO Bank.
- 51. 53 The Episodes of Currency Crisis in Latin... June 10, 1998 CASE Reports No. 39 Third quarterly review of the policy package. International reserves strengthen in the larger-than expected scope, but recession deepens. Adjustment in fiscal policy allows for an increase in the fiscal deficit target for 1997/98 from 2 percent to 3 percent of GDP. June –July 1998 The exchange rate weakens. Fiscal and monetary policies have been tighter than programmed, activity is weaker than expected, and exports fail to pick up. The large adjustment in current account reflects a sharp compression of imports. Growing difficulties in corporate sector. Union Bank of Bangkok and Laem Thong Bank are intervened. Laem Thong Bank is merged with Radanasin Bank. Union Bank of Bangkok together with 12 intervened finance companies merged with Krung Thai Thanakit (KTT), the state owned finance company and subsidiary of the state-owned Krug Thai Bank (KTB). First Bangkok City Bank is merged with KTB. August 1998 Introduction of financial sector restructuring package.. August 25, 1998 Fifth Thailand’s IMF Letter of Intent. September 11, 1998 Fourth quarterly review of the policy package. Foreign exchange market conditions are relatively stable (in spite of the Russian crisis), provide room for interest rates lowering to pre-crisis level. October 19, 1998 As of this date, 12.2 billion of U.S. dollars of total financing package for Thailand (17 billion of U.S. dollars) has been disbursed, including 3 billion of U.S. dollars from the IMF and 9.2 billion of U.S. dollars from other multilateral and bilateral sources. December 1, 1998 Sixth Thailand’s Letter of Intent. March 23, 1999 Seventh Thailand’s Letter of Intent. April 1999 Establishment of Bank Thai from merger of Union Bank of Bangkok and 12 finance companies. May 1999 Siam Commercial Bank raises over 1.5 billion of U.S. dollars in new capital. July 1999 Nakomthon Bank is intervened. August – November 1999 Auctions and further asset of finance companies sales to the state owned Asset Management Company. September 1999 Nakomthon Bank is sold to Standard Chartered Bank. September 21, 1999 Eighth Thailand’s Letter of Intent. November 1999 The sale of Radanasian Bank to United Overseas Bank of Singapore is finalized. May 8, 2000 The IMF completed Final Review of the Thai stabilization program.
- 52. 54 Marek D¹browski (ed.) References Antczak R. (2000). "Theoretical Aspects of Currency crises", Studies and Analyses, No. 211, CASE Warsaw. Blaszkiewicz M. (2000). "What Factors Led to the Asian Financial Crisis: Were or Were not Asian Economics Sound", Studies and Analyses, No. 209, CASE Warsaw. Chau-Lau J.A., Z. Chen (1998). "Financial Crisis and Credit Crunch as a Result of Inefficient Financial Intermedi-ation – with Reference to an Asian Financial Crisis". IMF Working Paper, 98/127, (Washington: International Mone-tary Fund). Chang R., A. Velasco (1998, "Financial Crisis in Emerging Markets: A Canonical Model", Working Paper 98-10, Reserve Bank of Atlanta. Demirguc-Kunt A., E. Detragiache, P. Gupta. "Inside the Crisis: An Empirical Analysis of Banking Systems in Distress". IMF Working Paper 00/156, Washington D.C. Fisher Jr. R., R.P. O'Quinn (1998). "The United States and Thailand: Helping a Friend in Need". The Heritage Founda-tion Backgrounder, March, Washington, D.C. Jakubiak M. (2000). "Indicators of a Currency Crises: Empirical Analysis of Some Emerging and Transitional Economies", Studies and Analyses, No. 218, CASE, Warsaw. Johnson B.R., S.M. Darbar, C. Echeverria (1997). "Sequencing Capital Account Liberalization: Lessons from the Experiences in Chile, Indonesia, Korea, and Thailand". IMF Working Paper, 97/197, (Washington: International Monetary Fund). Krugman P. (1998). "What Happened to Asia?". (January), http://web.mit.edu/krugman/www/disinter.html Marshall D. (1998). "Understanding the Asian Crisis: Sys-temic Risk as Coordination Failure". Economic Perspectives, 3rd Quarter, Federal Reserve Bank of Chicago, p. 13–28. Nogayasu J. (2000). "Currency Crisis and Contagion: Evi-dence from Exchange Rates and Sectoral Stock Indices of the Philippines and Thailand". IMF Working Paper, 00/39, (Washington: International Monetary Fund), June. O'Driscoll Jr. G.P. (1999). "IMF Policies in Asia; a Critical Assessment". The Heritage Foundation Backgrounder, March 1999, Washington, D.C. Stone M.R. (1998). ”Corporate Debt Restructuring in East Asia. Some Lessons from International Experience”. IMF Paper for Analyses and Assessment, PPAA/98/13, Octo-ber, Washington, D.C. "Thailand Economic Monitor", June 2000, The World Bank, http://www.worldbank.or.th/monitor. Thailand Selected Issues, IMF Staff Country Report No. 00/21, (Washington: International Monetary Fund). Thailand Letters of Intent and Memorandums of Eco-nomic Policies of: August 14, 1997, November 25, 1997, February 24, 1998, May 26, 1998, August 25, 1998, December 1, 1998, March 23, 1999, September 21, 1999, http://www.imf.org/external/np/loi CASE Reports No. 39
- 53. 55 The Episodes of Currency Crises in Latin... Part IV. The Malaysian Currency Crisis, 1997-1998 by Marcin Sasin 4.1. Overview 4.1.1. Introduction Malaysia, a country in Southeast Asia situated in the Malaysian Peninsula, gained its independence from Britain in 1957. Since then, the main political force in the country – the multiethnic National Front (Barisan Nasional) – has won all 10 elections. The key component of the National Front is the United Malays National Organization led by Mahathir Mohamad, who has also been a Prime Minister since 1981. The population of Malaysia, standing at 23 million is 60% Malay, 30% Chinese and 10% Hindu by origin. The econo-my and politics of Malaysia operates principally along racial lines. The Malays have monopolized the country's politics, they occupy key posts in the administration, military, police, 15 10 5 0 -5 CASE Reports No. 39 constitutional bodies etc. Chinese descents have dominated the country's economy and exercise control over 40% of all the nation's economic wealth [1]. The Hindu population is predominantly visible in the labor force. After some dra-matic racial tensions in the 1960s, the authorities have introduced so called New Economic Policy (NEP) – its pri-mary goal has been "accelerating the process of restructur-ing Malaysian society to correct economic imbalance so as to ... eliminate the identification of race with economic function... and the creation of a Malay commercial and industrial community... Within two decades, at least 30 per-cent of the total commercial and industrial activities ... should have participation by Malays... in terms of ownership and management" [2]. Uniform political leadership and social consensus grant-ed Malaysia a stable environment and allowed it to enter the track of rapid economic growth. The government has been involved actively in large infrastructure projects and other Figure 4-1. Malaysia: GDP growth (% p.a.) -10 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: IMF, IFS [1] Foreigners control over 30% - based on ownership of share capital of KLSE listed companies. [2] At the times of the formulation of this strategy Malays controlled around 10% of national wealth. The strategy proved partly successful, in 1997 the Malays' share increased to around 20%.
- 54. 56 Marek D¹browski (ed.) CASE Reports No. 39 100 80 60 40 20 public enterprises. Industrialization, export growth and investment have been promoted. In the late 1980s, it became clear that purely state-owned enterprises were not efficient enough for further growth – they have been there-fore privatized by "corporatization", listing on Kuala Lumpur Stock Exchange and by selling large parts of government shares. Foreign direct investment inflows have been suc-cessfully encouraged – between 1989 and 1995 they aver-aged around 7% of GDP. Fast growth was possible thanks to the impressive rate of investment, one of the highest in the world – in 1997 43% of GDP was invested [3]. The rate of savings was also remarkable – in 1997 it amounted to 39% of GDP, private and public saving had more or less equal shares in total sav-ings. In 40 years time, Malaysia managed to transform itself from an underdeveloped, third world commodity exporter to a modern and industrialized country. The share of prima-ry sector (agriculture, mining, fishing, etc.) in GDP was only 19% in 1997. Secondary sector (manufacturing, construc-tion) had a share of 40%, while tertiary sector (services) constituted 41% of GDP. 4.1.2. The Public Sector In Malaysia, the (non-financial) public sector consists of the federal government, 13 state governments (+2 federal territories), 148 local governments as well as statutory bod-ies and non-financial public enterprises. The federal govern-ment is endowed with revenue collection power over its most important sources (corporate, petroleum and person-al income taxes and custom duties) and with expenditure responsibilities over strategic domains (state expenditures, infrastructure, etc.). As a result it acquires over 80% of all public revenues and spends around 60% of all the expendi-tures. Local (and state) governments have narrowly defined responsibilities, i.e. provision of essential civic services to local communities. It derives financing from agricultural and urban taxes, non-tax revenues (like asset sales), and trans-fers from federal government. The size and influence of rather inelastic local government on fiscal situation is negli-gible compared to the federal government, which is the only body responsible of implementing fiscal policy objectives. Figure 4-2. Malaysia: GDP by sector of origin 0 1960 1965 1970 1980 1990 1995 Agriculture etc. Industry Services Source: IMF [3] There are suggestions that official measures of investment rate are likely to be upward biased, because part of investment in Asian economies is actually a disguised form of consumption. Figure 4-3. Malaysia: Domestic demand componentes in 1996 Investment public 12% Consumption private 43% Investments private 32% Consumption public 13% Source: IMF
- 55. 57 The Episodes of Currency Crises in Latin... 4 2 0 -2 -4 Generally speaking, the public sector is relatively small. In 1997, the federal government raised 24% of GDP and spent 21.4% of GDP. For the broad public sector (general government), these figures are 28.7% of GDP and 25.3% of GDP respectively. Direct taxes provide 40% of revenues, the same amount comes from indirect taxes, remaining 1/5 are derived from other sources. On the expenditure side, the budget is divided into current (operative) and develop-mental parts. Total direct development expenditures in 1997 amounted to 25% of the total budget or 6% of GDP. The importance of public sector in the economy is obscured by widespread use of off-budget accounting, through which significant part of government investment is done – espe-cially in the case of big infrastructure projects. Data related to such actions is limited. In 1998 off-budget spending amounted to around 1.1% of GDP, for 1999 this figure was 1.8% of GDP. Government engagement in large firms (usu-ally previously state-owned) effects in other quasi-bud-getary activities like tax concessions and exemptions, lend-ing on favorable terms, etc. with rather unclear conse-quences for public finance. Until the mid-1980s, the main purpose of fiscal policy was the implementation of government development objec-tives. The government became extensively involved in many large-scale non-financial public enterprises – infrastructure CASE Reports No. 39 projects, industrial investments, etc. This resulted in exces-sive budgetary deficits that eventually became unsustainable (reaching as much as 16.6% of GDP) and led to the eco-nomic crisis and recession in 1985. Afterwards, the conduct of fiscal policy was altered, with emphasis placed on macro-economic stability and sustainable growth promotion, and more attention paid to short-term aggregate demand man-agement. Expected budgetary revenues started to be calcu-lated realistically what resulted in somehow strange perma-nent record of revenue and overall balance underestimation [4]. As a result of the economy overheating from 1992–93, the authorities adopted a rather conservative fiscal policy in order to try to slow down and stabilize domestic demand arising from large capital inflows. As a consequence, from 1993 till 1997, public sector consecutively recorded sizeable budgetary surpluses. Thanks to consecutive surpluses, the Malaysian authori-ties succeeded in containing domestic debt – its volume was steadily falling in years preceding the crises. In 1996, federal government debt stood at decent level of 32% [5] and was held primarily in government securities. The broad public sector debt was somehow larger and totaled around 50% of GDP. The external debt of federal government was rela-tively small at end-1996, and amounted 4.1 billion USD, or around 4% of GDP. Figure 4-4. Malaysia: Federal budget fiscal position (% of GDP) -6 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: IMF Table 4-1. Federal Government Debt end-1996 (% of GDP) Domestic Foreign Total 31.7 4.2 35.9 out of which Gov. securities 25.1 Market loans 2.1 Source: IMF, Malaysian authorities [4] For 1990 the budget deficit has been overestimated by 1% of GDP. The same figures for 1991–1994 stand at 2%, 4%, 5% and 3% of GDP. [5] As a memorandum: central government debt amounted to more than 100% of GDP in 1987.
- 56. 58 Marek D¹browski (ed.) CASE Reports No. 39 14 12 10 8 6 4 2 Around half of this came from international institutions and foreign governments, half from market loans. To get a clearer picture of general government external liabilities the non-governmental public sector debt of 11.5 billion USD must be also be included – a considerable part of this debt carried federal government guarantees. All the public debt has been medium or long-term. 4.1.3. Monetary Policy and the Financial Sector In the 1990s, the short-term operational target of the Bank Negara Malaysia (BNM), i.e. Malaysian central bank was the one-month interbank rate. At the same time, central bank was monitoring money and credit growth and the exchange rate of Malaysian currency, the ringgit (MYR). The main instruments used were reserve requirements, direct lending and borrowing from the interbank market and sales of Bank Negara bills (introduced in 1993). Measures taken by the BNM were successful in containing inflation. In spite of rising aggregate demand, inflation stood below 4% in the years pre-ceding the crisis. Interest rate policy was rather liberal and remained in line with the corporate sector needs for low-cost financing and authorities' objective of fast development and economic growth. In 1994, the nominal interest rate was around 4%, which means that the real interest rate was about 1%. From 1994 until 1997, the nominal interest rate was steadily rising up to around 7% – it was a response to a large credit expansion. Nevertheless, the liquidity of the banking system remained high. The BNM has also been responsible for exchange rate management. As the interna-tional trade constituted a very large portion of the Malaysian Figure 4-5. Malaysia: Money market interest rate (%) 0 1994M1 1994M4 1994M7 1994M10 1995M01 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 Source: IFS Figure 4-6. Malaysia: Ringgit exchange rate (MYR/USD) 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 96-01-02 96-03-08 96-05-13 96-07-18 96-09-22 96-11-27 97-02-01 97-04-08 97-06-13 97-08-18 97-10-23 97-12-28 98-03-04 98-05-09 98-07-14 98-09-18 98-11-23 99-01-28 99-04-04 99-06-09 99-08-14 99-10-19 99-12-24 00-02-28 00-05-04 00-07-09 00-09-13 Source: Bloomberg
- 57. 59 The Episodes of Currency Crises in Latin... Table 4-2. Banking system indicators, third quarter 1997 economy, the primary objective was to maintain exchange rate stability in order to eliminate unexpected fluctuations and exchange rate risk. Although direct foreign exchange market interventions were not an instrument in the conduct of the monetary policy, this goal was achieved: in the 1990s (up to the financial crisis in 1997) the ringgit exchange rate stood always in the vicinity of 2.5 MYR/USD. The financial sector in Malaysia consisted of three types of institutions authorized to take deposits: commercial banks, finance companies, and merchant banks. Commercial banks were engaged in retail and wholesale banking and were the only institutions that could accept demand deposits. Foreign banks had a long tradition in Malaysia. Despite regulation lim-iting foreign ownership in the banking system, in 1999 foreign banks controlled around 25% of assets of all the commercial banks. Finance companies offered lending and other types of credit to consumers and small business deriving funds mainly from time and saving deposits. Merchant banks were engaged in other, usually fee-based activities such as loan syn-dication, corporate advisory work, securities underwriting and portfolio management. In recent times, merchant banks turned as well to the provision of loans and deposit taking, but the BNM tried to discourage such activities. In addition to the above-mentioned institutions, there were other participants of the financial market, like pension and insurance funds and stock brokerages. The Malaysian financial system was characterized by the relative importance of non-bank financial intermediaries – the combined assets of the BNM and all commercial banks constituted, in 1996, only about 53% of total financial system asset. Another feature of the system was a high leverage of financial enterprises. Over-all, the 1992–96 average debt-to-equity ratio amounted to 239%, and for insurance companies and stock brokerages 592% and 452% respectively. Since the late 1980s, the Malaysian authorities pur-sued a policy of financial system liberalization. The mea-sures included liberalization of interest rates, reduction of credit controls, enhancement of competition and efficien-cy, and deregulation of a banking system. In 1989, the bar-riers between different types of financial institutions were removed, finance companies were allowed to participate in the interbank market. In 1991, lending rates were lib-eralized (being previously pegged to lending rates of two main banks). As the result of authorities' effort to deepen the financial market, the interbank money and foreign exchange market developed rapidly. The 1989 Banking and Financial Institution Act placed all banking entities under the BNM supervision and oblig-ed them to meet tight prudential regulation such as disclo-sure requirements, limits on large exposures, loan classifi-cations, capital adequacy ratio, and other responsibilities based on the Basel capital framework. Indeed, Malaysia had a well-developed supervisory and regulatory frame-work – one of the best in the region. As a result, the qual-ity of assets in the financial system, according to the BNM, CASE Reports No. 39 Share of loans to broad property sector Risk weighted capital ratio % of total banking system assets (1999) Commercial banks 31.7 11 74 (dom:57, foreign:17) Finance companies 22.8 10.6 20 Merchant banks 31.9 13.3 7 Source: IMF, Malaysian authorities Table 4-4. Financial system indicator ("leverage" and short term debt), 1996 debt/equity ratio % of short term debt Financial system 239 90.6 Commercial banks 154 98.8 Finance companies 202 94.6 Insurance 592 62.2 Stock brokerage 452 95.9 Source: IMF Figure 4-7. Malaysia: Credit to private sector/GDP 200 150 100 50 0 1992 1993 1994 1995 1996 1997 1998 Source: IMF
- 58. 60 Marek D¹browski (ed.) 35 30 25 20 15 10 5 was relatively good. The proportion of non-performing loans in total loans outstanding was 4.1% (end-1996). The risk weighted capital ratio of 10.6% was also reasonable comparing to 8% minimum requirement. The side effect of liberalization was the emergence of a large number of small and undercapitalized banks. To reme-dy this problem, the BNM actively encouraged mergers in the banking system. The banking sector took advantage of the liberalization and engaged in rapid expansion of lending activity, which in years preceding the 1997 crisis took a form of a lending boom. The domestic credit growth averaged about 25% per annum. As a percentage of GDP, lending amounted to about 160% – one of the highest intermedia-tion level in the world. Lending to the broad property sector, against equity and other assets, thrived – the share of loans to this sector in total banking system portfolio (risk exposure) exceeded 30% and was the highest among merchant banks. The col-lateral valuation was also very high – ranging from 80 to 100% – which in the presence of real estate price inflation added much risk to this portfolio. Together with financial sector deregulation, the authori-ties pursued a policy of capital account liberalization. For the highly open and emerging Malaysian market, this liberalization was an important way of facilitating international trade and providing investment capital. In the years before the crisis, the capital control regime could be described as liberal. The ring-git was externally convertible – was allowed as a currency of trade settlements, which relieved resident importers and exporters from the need to hedge against exchange risk. There were relatively few restrictions on the ringgit transac-tions with nonresidents. As a consequence, the system of external accounts [6] and offshore market for the ringgit developed, mainly in Singapore. There were no restrictions on the source of funds placed in the external account as well as on the transfer of funds into or out of the external account. Malaysian banks were allowed to provide forward cover against the ringgit to nonresidents. Over-the-counter trading of the Malaysian stocks and bonds took place in Singapore and Hong Kong. Borrowing abroad by authorized dealers as well as their foreign exchange lending activities were only subject to prudential regulations. Foreign currency borrowing by residents was allowed, provided the applicant could prove its earning in foreign exchange. Portfolio capital inflow was unrestricted into all Malaysian financial instruments. FDI inflows were actively encouraged (e.g., through tax exemptions); repatriation of nonresident invest-ment and profit was completely free. This policy, together with a stable, growth-oriented envi-ronment and wide investment opportunities effected in con-stant inflow of foreign direct investments, which in 1997 amounted to 6.7 billion of USD. On the other hand, a favor-able MYR-USD interest rate differential and remarkable stabil-ity of exchange rate could not go unnoticed by the short-term investors. In response to a surge in capital inflow, that began to get out of hand in 1994, the authorities decided to introduce temporary inflow controls on portfolio transactions – these restrictions were indeed soon lifted. Short-term inflows quick- CASE Reports No. 39 Figure 4-8. Malaysia: Private sector credit growth in % p.a. 0 1992 1993 1994 1995 1996 1997 1998 total to property for consumption Source: IMF [6] External account is defined as a ringgit account maintained with a financial institution in Malaysia, where the funds belong to a nonresident indi-vidual or corporation.
- 59. 61 The Episodes of Currency Crises in Latin... Figure 4-9. Malaysia: Market capitalization (% of GDP at year end) 400 350 300 250 200 150 100 50 0 1992 1993 1994 1995 1996 19971 998 Source: IMF ly rebounded and in 1996 totaled over 4 billion of USD. As a result of these inflows, foreign reserves of the BNM increased steadily. 4.1.4. The Corporate Sector During the 1990s, Malaysia witnessed a rapid develop-ment of the capital market. The stock market expanded sig-nificantly – fuelled by the privatization and listing of large state-owned companies, establishment of the Securities Commission (1993) and credit agencies, improvement in trading and settlement system as well as strict prudential supervision. The capital market became a major source of funds for the corporate sector – in 1997 the total net amount raised there by the private sector equaled to 30.4 40 30 20 10 billion MYR (around 10 billion USD): half in new shares, half in debt securities. In 1997 Malaysia had the biggest per capi-ta market capitalization among ASEAN-4 countries (more than 3 times its GDP). Following the brief downturn in 1994, there was a strong upsurge in stock prices. At the end of 1996, annual increase of the Kuala Lumpur Stock Index (KLSI) exceeded 25%. However, in the first quarter of 1997 the trend was reversed and the KLSI started to decline. Equity price increases were not the only stock exchange phenomenon, there was also general asset price inflation – most notably and importantly in the property and real estate sector. This happened primarily due to large capital inflow and domestic credit expansion. The corporate sector in Malaysia was characterized by rapid growth. In the 1990s and before the crisis, a number 1400 1200 1000 800 600 400 200 CASE Reports No. 39 Figure 4-10. Malaysia: Net funds raised by the private sector in the capital market (bln MYR) 0 1992 1993 1994 1995 1996 1997 1998 Source: IMF Figure 4-11. Kuala Lumpur Stock Index (KLSI) 0 1992M1 1992M5 1992M9 1993M1 1993M5 1993M9 1994M1 1994M5 1994M9 1995M1 1995M5 1995M9 1996M1 1996M5 1996M9 1997M1 1997M5 1997M9 1998M1 1998M5 1998M9 1999M1 1999M5 1999M9 Source: Bloomberg
- 60. 62 Marek D¹browski (ed.) of listed companies grew by average 14% per annum, and total market capitalization of companies from main and second board by 40%. Since the late 1980s, the evolution of corporate sector was boosted by the privatization of large state-owned companies and big investment projects. Another reason for private sector expansion was the sta-ble macroeconomic situation and growth-oriented fiscal and monetary policies encouraging investment and capital inflows. For financial enterprises, the average growth rate of assets amounted to 40% per annum between 1992 and 1996, for non-financial enterprises it averaged 31%. Growth was financed mainly through debt issues and bor-rowing, however, equity also became a significant source of fund rising. The important feature of private sector financing was a heavy dependence on short-term debt, which constituted 90% of the total debt of financial insti-tutions and 60% in the case of non-financial enterprises. Fortunately for Malaysia, the corporate debt was primar-ily domestic. Resident companies were banned from tak-ing foreign exchange denominated loans unless they could prove hard currency revenues. The corporate sector also became highly leveraged due to a long period of high growth and a policy aimed at rapid asset accumulation and sales expansion. Malaysian corporations had a rather com-plicated interdependence structure such as cross-hold-ings, "double leverages", pyramid structures, etc. 28% of market capitalization was in 1997 controlled by 15 fami-lies, which was a very high degree of ownership concen-tration. Many large companies, especially those created as a part of industrialization strategy or previously state-owned, had close links to the government. Total external debt of the broad private sector at the end of 1996 amounted to 23 billion USD, 60% of that medium and long term. Remaining 40%, i.e. about 10 bil-lion, was short-term owed mainly (6.7 billion USD) by the banking sector. Non-bank financial institutions and corpo-rate sector accounted for 1/3 (3.2 billion USD) of total short-term debt. There was a rather rapid build up in short-term private debt in 1996 – it rose more than 50% from 1995 level, compared to 15% increase between 1994 and 1995. Total external debt of an economy as a whole amounted to 38.6 billion USD, or about 40% of GDP. 74% of this debt was denominated in USD. 4.1.5. The External Sector The Malaysian economy is the most open in Southeast Asia. The measure of openness, i.e. average cross-border trade ((export+import)/2/GDP) amounted 110% in 1997. The once popular import substitution was, since the 1970s, gradually abandoned and switched to export pro-motion. This process gained momentum in the mid-1980s when Malaysia embarked on the process of gradual liber-alization of trade and exchange regime designed to boost the export-oriented manufacture sector. Measures were taken to promote export competitiveness (like the nomi-nal ringgit devaluation) and facilitate the import of essential capital goods and intermediate inputs. Rapid economic development and industrialization sifted Malaysia's export from commodities to manufactured goods. The share of rubber in export fell from 55% in 1960 to 2% in 1997. The share of manufactured goods in export rose from 16% in 1960 to 80% in 1997. Malaysia's outward orientation intensified during the 1990s. Non-tariff barriers were gradually eliminated in line with Malaysia's commitment under the WTO. The effec-tive import-weighted tariff rate was lowered from 11.2% in 1992 to 9.4% in 1997. CASE Reports No. 39 Table 4-4. External debt at end-1996 in bln USD Total 38.7 public 15.7 Long-term private 13 public 0 Short-term banking 6.8 private non-banking 3.2 Source: BNM Figure 4-12. Direction of Malaysian export – Singapore 20% Japan 13% Other 27% US 19% Other ASEAN 6% EU 15% Source: BNM, IMF
- 61. 63 The Episodes of Currency Crises in Latin... Figure 4-13. Malaysia: Export of goods composition Singapore was the major Malaysian trade partner in 1997 (20% of total export), followed by the US (19%), European Union (15%) and Japan (13%). Trade links with other ASEAN countries were not as close as one would expect given their geographical proximity. The most important item in commodities' export in 1997 was palm oil (3.8 billion USD), crude petroleum (2.5 billion USD) and rubber (1 billion USD). Total major commodity export made up about 15% of the total export. The main com-ponents of manufactured export were semiconductors (14.5 billion USD), electronic equipment (14.2 billion USD), and electrical appliances (13.6 billion USD) amount-ing together to around 55% of total export. Intermediate goods (51 billion USD) and capital goods (15 billion USD) dominated Malaysian import. Consumption goods (5 bil-lion USD) constituted only 6% of total imports. The over-all trade balance was usually positive through the 1990s. Service settlements – primarily freight, insurance and investment income payments – decided about negative current account balance. In the analysis of external sector development in Malaysia before the 1997 Asian financial crisis, there are two important issues that have to be addressed: real exchange rate overvaluation and current account deficit. There are well known problems with measuring real exchange rate misalignment but there is a consensus that prior to the crisis the Malaysian ringgit was overvalued at least 5% (CPI-based). Other methods (PPI-based, export-unit- value-based) produce figures around 20–25%. Nom-inal exchange rate was kept in rather narrow range of 2.4–2.8 MYR/USD. This virtual peg to the USD was expected to facilitate external financing of domestic pro-jects and promote international trade. The real overvaluation of the ringgit emerged from two sources. First, the sharp appreciation of the US dollar relative to Japanese yen and to European currencies led to deterioration of cost-competitiveness in Malaysian trade with these countries. Second, the surge in private capital inflows – notably portfolio and foreign direct investment – that took place in the 1990s led to continuous upward pressure on the ringgit: from 2.73 MYR/USD in early 1994 it appreciated to 2.48 MYR/USD in 1997. Real apprecia-tion led to a current account deficit – movements of the real exchange rate were, to high degree, correlated with a current account balance. The other factor worth mention- 120 100 80 60 40 20 CASE Reports No. 39 – semicond. 19% other 1% other manufacturing 26% minerals 5% electronic equipement 21% electrical appliances 17% agriculture products 11% Source: BNM, IMF Figure 4-14. Malaysia: CPI-based real effective exchange rate 0 1994M1 1994M4 1994M7 1994M10 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 Source: IFS
- 62. 64 Marek D¹browski (ed.) 30 20 10 0 -10 ing was the fall in demand for semiconductors – the main Malaysian export product – in the years preceding the 1997 crisis. On the other hand, the above trade balance explanation might not be sufficient – most of the current account deficit originated in service account, notably in investment income payments. The surge in foreign direct investment resulted in a rather inelastic demand for intermediate and capital goods import. This would point to large long-term capital inflows as a principal reason of current account imbalances, and suggests that these imbalances had a structural character. However, Malaysia had a good record of current account financing – it was in almost 100% covered by foreign direct investments. Nevertheless, Malaysian authorities took steps to contain this deficit – mainly through restric-tive fiscal policy. To some extent, they proved to be effec-tive: from a record 10% of GDP in 1995, the current account deficit decreased in 1997 to 5%. 4.2. The Crisis 4.2.1. Introduction The direct cause of the Malaysian financial crisis was the contagion effect from Southeast Asian neighbors. The mechanism of crisis development has been similar to these countries. A gradual deterioration in the macro-economic situation was making international capital mar-ket anxious about further profitability of investment in Malaysia. Some short-term capital was withdrawn, the downward pressure on currency confronted the authori-ties with the necessity to take one of the alternative deci-sions. First, if the authorities have sufficient foreign reserves they can defend the currency through raising interest rates. Such a measure has a well-known shortcoming – it slows down the economy. This could be acceptable as a temporary response – longer debt maturities are usually not affected. It becomes more serious if a large portion of domestic debt represents short-term maturity. A pro-longed period of high short-term interest rates can force some cash-short companies to postpone investments. This means a danger of recession. Some companies being very cash-short may default on their obligations. One default can cause another default – the economy will experience a credit crunch with serious contractionary consequences. Matters can become much more serious if companies are highly leveraged – they not only become insolvent but they can go bankrupt. That happens more often if they invest in assets such as stocks or real estate, as prices usually decrease substantially in the crisis. The second option is letting the exchange rate depre-ciate. However, if the economy has a large stock of exter-nal debt, especially in the private sector or if firms borrow in foreign currency, getting their receipts in domestic ones (and remaining usually unhedged after a prolonged period of exchange rate stabilization) companies can default with all the above-described consequences. Additionally, defaults on external debt damage the country's interna-tional reputation. Moreover, after devaluation the increased price of import puts pressure on domestic price level, inflation accelerates which in turn fuel depreciation expectations and the vicious circle will close. If there is a conjunction of negative factors such as a large stock of short-term debt, excessive leveraging of companies and large unhedged external debt (combined with other ingredients like shortage of foreign reserves or political instability), the authorities might have no degree CASE Reports No. 39 Figure 4-15. Malaysia: Current account balance and composition (% of GDP) -20 1994 1995 1996 1997 1998 current account trade balance service balance Source: BNM, IFS
- 63. 65 The Episodes of Currency Crises in Latin... of freedom to maneuver. Such situation falls under the term "vulnerability". When this becomes common knowl-edge, speculators join capital outflow and sell the domes-tic currency short. At this point, it is virtually impossible to end the crisis without paying a high price. Malaysia was not an exception to the mechanism described above. However, the extent of the crisis was moderate compared to Indonesia, Thailand or Korea. Malaysia managed to avoid the widespread bankruptcies, bank runs, social unrest, and downward spiral of inflation and recession. The key questions are: was Malaysia "vul-nerable", if yes – why, and why matters did not go such a disastrous way as in other Southeast Asian countries? 4.2.2. Malaysian Vulnerability Analysis We start the analysis of Malaysian "vulnerability to cur-rency/ financial crises" in the end of 1996 with a review of the academic research findings related to currency crisis predictions. The most popular approach is called "early warning system" and is based on the fact that usually some macroeconomic indicators (called "leading indicators") endowed with above-average predictive power exceed their usual values and, therefore, issue a warning signal on the possible crisis. The main drawback of this system is that after each crisis researchers revise the set of leading indicators and come up with the new ones that seem to have a better predicting power. Afterwards, a new crisis unfolds and a new revision takes place. However, there is a consensus [see, for example, Edison, 2000] that the most useful indicators are as follows: real exchange rate, level of foreign reserves, current account balance, the level of short term debt, domestic credit growth, fiscal and monetary expansion, short-term capital flows as well as other, hardly measurable features like the extent of moral hazard ("the incentive structure of financial system"), exposure to contagion, etc. Essentially all the currency crises in the 1990s had a short-term capital outflow and currency speculation as their direct cause, and so were with Malaysia. From the point of view of short-term international investor holding ringgit assets, the most important parameters were the expected MYR/USD exchange rate and domestic interest rate. In order to defend a currency against speculative pres-sures the central bank needs to have the sufficient inter-national reserves. Malaysian foreign exchange reserves stood at 27.7 billion USD at mid-1997. The first question is whether the debtor had enough foreign currency to pay the interest and amortization due. Total external debt service in 1996 amounted to 6.4 billion USD or 23% of the international reserves, so they were sufficient to service the debt. In fact, Malaysia had the best position in the region with respect to this issue. Sometimes the debt service is presented as a ratio of export – again the ratio of 8.2% was the best indicator in the region. The second question is: once all short term creditors would like to withdraw their funds at one moment, would reserves be sufficient enough to meet their demand? Again, the ratio of the total external short-term debt plus external debt service to foreign reserves stayed at 70%, the best among ASEAN-4 [7] countries plus Korea. Figure 4-16. Malaysia: Foreign reserves (mln USD) 35000 30000 25000 20000 15000 10000 5000 0 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 Source: IFS [7] Malaysia, Indonesia, Thailand, Philippines. CASE Reports No. 39
- 64. 66 Marek D¹browski (ed.) However, there was some confusion about the defini-tion of short-term debt. The above used data, coming from the BNM, revealed (as mentioned above) that 27% (around 10 billion out of 38.6 billion USD) of the total external debt had a short-term character. When we take into account the data of The Bank of International Settle-ments whose members controlled 26 billion USD (2/3) of the total Malaysian debt the picture looks somehow differ-ent and less comfortable. In the BIS debt sub-sample the proportion of short-term debt to total debt amounted to about 50%. Hence, all the above indicators should be adjusted (almost twice) accordingly. International investors can also be afraid that in the case of financial panic they will not get their hard currency back. From the theoretical point of view, they can feel secure when central bank liabilities (reserve money) are covered by international reserves. So the ability of the cen-tral bank to completely cover its liabilities with foreign exchange reserves is a good sign for the solvency of the system. The ratio of reserve money to official reserves at end-1996 equaled exactly 134% [8]. However, BNM acted as a lender of last resort to the banking system. So, in the event of financial panic all liquid money assets could poten-tially be converted into foreign currency. Hence, the ratio of M2 to foreign reserves was another leading indicator of possible distress. In Malaysia this ratio amounted to 480%, which could be regarded as potentially dangerous but still it was the best among ASEAN-4 [9]. On the other hand, the informative content of this indicator is not very clear, as M2/FX ratio is, to large extent, country specific and reflects rather the development of domestic banking sys-tem. Malaysian system in the 1990s was always character-ized by a high degree of financial intermediation. The importance of external short-term debt was already discussed above – Malaysia had actually a very decent record with respect to that. Almost equally impor-tant was the share of short time debt in private sector financing. The overdependence of the Malaysian corporate sector on the equity market and short-term debt securities pointed to the underdevelopment of long-term loan mar-ket. The high share of short-term liabilities meant that the authorities would restrain as much as possible from inter-est rate hikes in the event of capital outflow and increased pressure on the foreign exchange market. Instead, they would seek to sterilize these outflows by direct interven-tions on the interbank money market. The BNM policy of stabilizing the interest rate meant that facing possible devaluation expectation the central bank would hesitate to compensate (with high interest rates) investors for the devaluation risk. Devaluation expectations could possibly come from concern about overvaluation of the the real exchange rate. The exchange rate was assessed to be overvalued about 5% based on CPI-measure [10]. Based on a real-export-unit- value, the real exchange rate was overvalued by about 20–25%. Still this could reflect a temporary decline in semiconductors and commodities prices. Although differ-ent sources proposed different estimates there was a con-sensus that the Malaysian currency was over its parity in 1996. The prime source of concern in 1996 was a wide cur-rent account deficit (around 10% in 1995 and around 5% in 1996). All through the 1990s, the balance was perma-nently negative giving rise to anxiety about its sustainability [11]. The notion of "sustainability" is, however, hard to define – what is sustainable today can become unsustain-able tomorrow. The principal issue is CA deficit financing. In the case of Malaysia, it was covered through inflows of capital, notably FDI. The inflows of FDI in the 1990s was just sufficient to cover the current account deficit. But in the years preceding the crisis, FDI/CA deficit ratio was rather on the long-term decline, and there were expecta-tions that this trend might continue. One way to make "sustainability" operational is to introduce non-increasing foreign debt to GDP ratio. The current account is sustainable if it doesn't cause an exces-sive build-up of foreign debt. By taking an arbitrary 1% difference between long-run interest rates and long-run growth rate, Corsetti, et.al. (1998) show that a "sustain-able" current account in case of Malaysia equals about 2.3% of GDP. In practice, the external debt/GDP ratio stood almost unchanged since 1993 till the crisis. Generally, there is nothing wrong with a current account deficit as long as it reflects a consumption smooth-ing process. Milesi-Ferretti and Razin (1996), argue that Malaysian deficit development matched this pattern rather closely. But the consumption smoothing theory predicts that at some moment in future there will be a switch into trade surplus. From a political and economic point of view, the deficit will be "sustainable" if it can be reverted into surplus without a crisis or drastic policy change. However, in 1996–97 the current account imbalance seemed to have [8] The money(M1)/foreign reserves indicator was 145%. [9] It is worth to notice, that in November 1994, just before Mexican crisis M2/foreign reserves has been 9.1 in Mexico, and 3.6 in Brazil and CASE Reports No. 39 Argentina. [10] This method has, however, many drawbacks. First, inflation is usually closely monitored by authorities and they attempt to contain the increase in the price of the most weighted components – so a change in inflation do not necessarily reflect the same change in competitivenes. Second, the so-called Balassa-Samuelson effect should be taken into account – due to faster productivity growth in the tradable goods sector the exchange rate seems overvalued (in CPI-terms) while it is actually not. [11] The authorities were perfectly aware of the problem and tried to cool down the economy and reduce CA deficit in 1997 budget.
- 65. [12] Some examples included: huge dam in remote Borneo, which rationale was questionable, new national airport, costing 9 billion US$ (almost [13] The issue of capital productivity is closely linked to ongoing and yet unresolved debate about the causes of Asian miracle, namely whether the fast growth of Asian countries resulted just from abundance of capital and labor force or from productivity growth. The first view was advanced by Yong in the beginning of the 1990s and popularized by Krugman (1994, 1998). They argue that the total factor productivity in East Asia was significant-ly lower (sometimes even close to zero) than the rate of GDP growth. There were also studies that contradicted this view. Sarel (1997) found that TFP in Malaysia in 1978–1996 grew on impressive rate of 2% per year, while Claessens et.al. (1998) remarked that the return-on-assets indicator increased in Malaysia from 5.5% between 1988–1994 to 6.3% in 1995–1996. [14] It is worth mentioning that prices of some most weighted consumer goods in the CPI were effectively regulated and controlled in order to [15] Sarno and Taylor (1999) conduct a formal test and cannot reject the hypothesis that asset prices took divergent (bubble) path in Malaysia before 67 The Episodes of Currency Crises in Latin... Figure 4-17. Malaysia: FDI/current account deficit (%) 300 250 200 150 100 50 0 1991 1992 1993 1994 1995 1996 1997 Source: Corsetti et. al. (1998) a structural character. The perspective of its financing was closely connected to a "sustainability" of capital inflows which, in turn, was connected to growth sustainability and investment opportunities. Malaysia's economy grew at the average rate of 8% per annum in the 1990s. It was possible thanks to sizeable cap-ital inflows, and vice versa the capital was attracted by the expected high rates of growth and investment opportuni-ties. However, the abundance of capital, high investment rate and development promotion by the authorities wors-ened the quality of investment projects. The government engaged itself in "mega-projects" and massive infrastructure constructions [12], many of them reflecting political or pro-pagandist considerations, rather than efficiency justification. Private sector, also facing increased supply of capital turned to more risky projects. Indeed, the incremental capital out-put ratio increased from 3.7 in 1987–92 to 4.8 in 1993–96, indicating a sharp decline in investment efficiency and prof-itability [13]. In such a situation it would be overoptimistic in early 1997 to hold expectations that in short future Malaysia would sustain 8–9% rates of growth without some adjustment or structural reforms. The signs of overheating were evident in end-1996 and, eventually, optimistic growth expectations was revised in early 1997 when the data revealed a sharp fall (60%) in investment from both foreign and domestic sources. The economic slowdown seemed inevitable, for what Malaysian growth-oriented companies with ambitious fund rising aspi-rations were not prepared. The authorities denied the overheating problem pointing to low inflation. However, rapidly expanding domestic credit fueled not only consumer prices [14] but the asset market too. The asset market development seems to be the main factor of Malaysian vulnerability. Facing diminishing returns in corporate sector investment, the banking system switched from lending to manufacturing (growth in lending fell from 30% in 1995 to 14% in 1996) to lending for equi-ty purchases (and growth in loans granted for share pur-chases rose to 20%, from 4% in 1995). Such loans were granted mainly by finance companies and merchant banks. As a result of such a policy and the wide availability of property loans, asset prices (shares, real estate) increased rapidly. A surge in the asset market was consistent with (speculative) overinvestment story described above. The KLSI gained about 25% in 1996 and a property and equity related sector index rose over 50% [15]. The BNM made an effort to slow down the domestic credit expansion by rising reserve requirements and intro- 10% of GDP), etc. keep "inflation" down. 1997. CASE Reports No. 39
- 66. 68 Marek D¹browski (ed.) ducing controls over consumer lending for cars and hous-es in October 1995. It was also gradually raising interest rates. In March 1997 the BNM tried to halt the asset mar-ket bubble by placing restrictions and ceilings on property and equity lending [16]. But this intervention probably came too late. Recognizing that these measures and BNM attitude would eventually put an end to the property and stock boom, investors started to withdraw their funds. This only reinforced the downward trend on KLSE and on the prop-erty market, which already started in the beginning of 1997. Few days after the restrictions were announced the index was 17% lower than its heights month before. In the first-half of 1997, the capital market witnessed some spec-tacular failures in fund raising and initial public offerings. Many companies were forced to suspend their invest-ments. Still, there were no crisis expectations at the time. Economists and analysts were only talking about a "slow-down". The Malaysian economic fundamentals seemed strong, and the BNM had a good reputation. In mid-May the ringgit came under speculative pressure but a few days of high interest rates (overnight rose as high as 18%) was sufficient to counter this pressure and fend off the specu-lation with virtually no impact on the exchange rate. Malaysian securities had a high rating -rating agencies failed altogether to anticipate the crisis. 4.2.3. Crisis Development Malaysia encountered serious problems on July 2, 1997 when the Thai baht peg to the USD collapsed. Immediate-ly, market confidence to Southeast Asian economies was reassessed – unexpected devaluation of the baht meant that any country with similar economic structure, compa-rable state of fundamentals and export structure was like-ly to give up its exchange rate policy under similar circum-stances. A prolonged period of fierce speculative pressure has started. The yield curve inverted dramatically. On July 8, 1997, the BNM was forced to heavily defend the ringgit. Short-term rates reached the level of 50%. On July 11, 1997, the Philippines gave up supporting its peso. The ring-git collapse was a matter of days. After the weekend, on July 14, the BNM abandoned its ringgit peg to the USD. Ten days after, Prime Minister Mahathir publicly blamed "rogue speculators" as responsible for the crisis. Later on, on several occasions, he suggested that currency specula-tion should be banned. Such statements further under-mined investors' confidence in Malaysia. On September 4, the ringgit broke the psychological barrier of 3 MYR/USD and continued to depreciate. In January 1998 it hit his bot-tom ever of 4.5 MYR/USD. The actual direct cause of the crisis was a rapid rever-sal of short-term capital flows that finally turned into finan-cial panic. Malaysia, which in 1996 experienced inflows of over 4 billion USD and expected billions more in 1997, lost Figure 4-18. Malaysia: Capital flows (bln USD) 10 5 0 -5 -10 1994 1995 1996 1997 1998 total long term short term Source: IMF [16] The restrictions introduced limited further loans to 15% (30%) for commercial (merchant) banks and ceilings on existing outstanding stock of loans, which actually was lower than the proportion of property loans in banks portfolio. That implicitly meant that banks have to contract their lend-ing, CASE Reports No. 39 cease to roll-over pending credit, so new loans wouldn't actually be granted soon.
- 67. 69 The Episodes of Currency Crises in Latin... 4 billion USD in 1997 and about 5 billion in 1998. The equi-ty and property asset bubbles burst. KLSI in 1997 lost 50% of its value, while its property sector subindex lost almost 80%. Vacancy rate in central business district rose from 10% in June 1997 to 17% in June 1998. Prices and rents of office and residential property fell. The crisis brought a 7% decline in Malaysian GDP in 1998 (instead of 8% growth as expected). It affected the Malaysian economy via the following transmission channels. The stock exchange crash was a shock in a country with such a heavy reliance of the corporate sector on the capital market. Many projects had to be halted, companies had to postpone their investments. There was also a widespread uncertainty about the direction of the economy that con-tributed to the investment slowdown. This sharp decline in investment activity was the major source of a drastic fall in aggregate demand, which in turn became responsible for a recession. Private investment demand fell by 58% in 1998. Rapid depreciation of asset prices brought massive nega-tive wealth shock. Individuals and the economy as a whole Table 4-5. Domestic demand collapse components private public Source: BNM Figure 4-19. Malaysia: Percentage of nonperforming loans in portfolio 40 35 30 25 20 15 10 5 0 1994Q1 1995Q1 1995Q2 1995Q3 1995Q4 1996Q1 1996Q2 1996Q3 1996Q4 1997Q1 1997Q2 1997Q3 1997Q4 1998Q1 1998Q2 1998Q3 1998Q4 1999Q1 Source: IMF CASE Reports No. 39 postponed their consumption. Private consumption, a second major factor in aggregate demand decline, contracted by 12%. As explained below, the overall contribution of public sector to aggregate demand change was also negative. A sharp ringgit depreciation brought a rapid improve-ment in the current account balance. From a deficit of 5% of GDP in 1997, it turned to a 13% surplus in 1998. As export and service balances held steady, all the current account adjustment happened by a squeeze in imports which declined by 18%. The positive net external demand partially (in half) offset domestic demand contraction, so the overall aggregate demand fall by 25%. Exchange rate depreciation triggered inflationary pres-sures in the economy. From the end of 1997, inflation started to rise and at its peaks reached about 6% in mid of 1998. Afterwards, however, thanks to the stabilization of the exchange rate, inflation began to fall and reached pre-crisis level in the beginning of 1999. Inflation was substan-tially lower than expected [17]. The consumer price index revealed only a part of a price increase process – a large Change from 1997 to 1998 in % consumption -10.3 investment -57.8 consumption -3.5 investment -10 commercial ban ks finance co mpanies merchant banks banking system [17] Malaysian authorities estimated that 1% of exchange rate depreciation would have the impact of 0.176% on inflation. Given around 40% depreciation of the ringgit, inflation should have risen at least the additional 6–7 percentage points.
- 68. 70 Marek D¹browski (ed.) part of the impact of ringgit depreciation was not passed on to consumers but instead absorbed by firms. As opposed to other countries in Southeast Asia, Malaysia did not have a large burden of private sector external debt. Its vulnerability to exchange rate deprecia-tion was therefore of less concern and did not pose a seri-ous threat to the economy. Movement in exchange rate also had little impact on banks portfolio quality because of BNM strict prudential supervision over foreign currency borrowing. The collapse of asset prices caused heavy losses of usually highly leveraged financial institutions engaged in stock market gambling, real estate lending, or other col-lateral- based credit activities. The banking system became insolvent, the ratio of non-performing loans soared, and the economy entered a credit crunch. The main source of concern became merchant banks and financial companies – exposed the most to asset market risk. The banking system portfolio quality deteriorated slowly during 1997. After a crash of the stock and prop-erty markets this process accelerated rapidly. The pro-portion of non-performing loans in the total credit vol-ume increased from 6% in December 1997 to 24% in March 1999. The worst was the situation of merchant banks, which had 37% of their total assets in non-per-forming status. In 1998, the banking system reported 2.2 billion MYR pre-tax losses, in sharp contrast with 7.6 bil-lion MYR profit in 1997. Losses belonged only to finan-cial companies and merchant banks, as commercial banks managed to maintain positive return on assets [18]. Despite losses the capital base of the banking sys-tem increased by 1.2 bln MYR in 1998 but only thanks to 4.6 bln MYR injection of fresh capital provided by the authorities. As a result of an asset meltdown through 1998, at the end of this year 9 out of total 78 financial institutions ceased to meet minimal capital criteria and other basic regulatory requirements. In addition to bank-ing sector problems, there were also widespread prob-lems in the shortly indebted corporate sector. Most of the companies' short-term debt was to domestic banks, establishing the link between financial sector problems and corporate sector distress. Some companies went insolvent or illiquid because their banks were unable to roll over their short-term debt. Similarly, the lack of cap-ital and falling asset prices caused many corporate defaults, increasing the proportion of bad loans in the financial sector. 4.3. Response to the Crisis 4.3.1. Introduction Hostile towards international institutions, as well as those who tolerated currency speculation which suppos-edly caused the financial crises in sound and solvent emerging markets, the Malaysian authorities rejected the help of the International Monetary Fund (together with its restructuring and reform directives), and decided to cope with the crisis themselves. 4.3.2. Fiscal Policy Response The 1998 budget presented in October 1997 was designed to deal with large current account imbalance and overheating through further fiscal contraction. The fiscal surplus of 3% of GDP was planned, up from over 2% of GDP in 1997. In addition to these measures, the Prime Minister finally postponed several multibillion construction and infrastructure projects. In November-December 1997, it became obvious that Malaysia was going to experience a significant slowdown and the authorities announced dramatic policy tightening (fiscal and monetary as well) to counteract the crisis. Bud-get spending were cut by 18% (partly to avoid a possible deficit), regulation of big import was introduced. Projec-tion of budgetary surplus was revised down to 1.5% of GDP in anticipation of lower tax revenues. These steps came far too late as the crisis spread to other parts of the economy. The authorities realized this in the beginning of 1998 and decided to stimulate the economy. From March 1998, the policy stance was gradually changed to the expansionary one. In March 1998, the fiscal package of 3 billion MYR (1% of GDP) was announced. At the same time a drastic revision of federal budged projected the fis-cal deficit of 2.6% of GDP. In July 1998, the additional fis-cal package of 7 billion MYR (2.5% of GDP) was declared. New funds were to be spent on development projects. Some of previously suspended undertakings were resumed. The fiscal expansion continued in 1999 with an expected deficit of about 5% of GDP. Despite this effort, and despite the fact that due to con-secutive budget surpluses in the 1990s Malaysia was pre-pared for a period of fiscal expansion, the federal govern-ment response failed to provide the expected stimulus. Actually, the public sector contribution to domestic demand was negative – public consumption and investment [18] Although commercial banks didn't engage themselves excessively in equity and broad property credit they also incurred heavy losses because CASE Reports No. 39 of their capital interdependence with finance companies and merchant banks.
- 69. Figure 4-20. Malaysia: monetary base and currency in circulation 100 90 80 70 60 50 40 30 20 10 0 1994M1 1994M4 1994M7 1994M10 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 [19] Income tax is based on the income received during the preceding year. The direct tax revenues were 8% larger than planned. [20] Spread emerged because the domestic interest rates were kept artificially low relative to market sentiments. At the same time, ringgit funds were needed for currency speculation, driving offshore interest rate high. The important source of speculative ringgit funds for nonresidents were offer side swaps. 71 The Episodes of Currency Crises in Latin... fell, compared to 1997, by about 3% and 10%, respec-tively. Actual deficit of the federal government of 1.5% of GDP occurred to be smaller than planned. Tax revenues were underestimated [19]. Also expenditures fell 0.5% GDP short of what was expected – a result of "institution-al and operational obstacles". Because the authorities relied on inert infrastructure projects spending as an instrument for reviving the economy, the fiscal stimulus got delayed and spread to 1999–2000. 4.3.3. Monetary Policy Response and Capital Control There were many phases and policy stance changes in the central bank's response to the crisis. In July 1997, BNM engaged in the heavy support of the ringgit, As a conse-quence of its interventions, short-term interest rates remained very high for some time and foreign exchange reserves were sharply falling. After realizing that continuous pressure on the ringgit and capital outflow was not going to ease soon and that high interest rate was likely to damage fragile financial system and after depleting 4.9 billion USD or 20% of its reserves in fruitless market intervention the BNM gave up its direct support to the currency. It started to focus on domestic money aggregates and financial market stabilization, thus opting to sterilize the capital outflow. As a result, domestic interest rates returned to their usual levels, the monetary base remained virtually unchanged (it actually rose slightly), as well as M1 aggregate – but a spread emerged between offshore and domestic interest rate encouraging further capital outflows [20]. In an attempt to interrupt this process, controls were imposed on banks to limit non-commercial- related offer-side swap transactions to 2 mil-lion MYR per foreign customer. Short selling of stocks on KLSE was also prohibited. The BNM turned to the domestic market overlending problem. It established guidelines aimed at reducing annu-al credit growth to 25% by end-1997, to 20% by end- March 1998, and to 15% by end-1998, put restrictions on financial companies consumption lending and on all bank-ing sector property projects. Relatively loose monetary policy continued for some time. The measures taken did not meet their aim – capi-tal outflows were not stopped, reserves continued to fall, and the ringgit was permanently depreciating at a very stable pace. These developments increased the burden of public and private external debt, increased the price of essential import items, and most notably, set off inflation-ary pressures in the economy. The inflation-depreciation spiral became a real threat. From November-December 1997 through February 1998, the BNM changed its policy stance to a contrac- CASE Reports No. 39 Monetary base Currency in circulation (bln MYR) Source: IFS
- 70. 72 Marek D¹browski (ed.) 90 80 70 60 50 40 30 20 10 tionary one, increasing several times its intervention inter-est rate to as high as 11% but at the same time lifting some lending restriction as domestic banking sector con-dition deteriorated. The monetary policy remained tight until August 1998 leading to a sharp contraction in mone-tary base followed by decrease in M1 aggregate money. In the meantime, the current account sharply improved and inflation, which prospects were in the beginning overesti-mated, was successfully subdued [21]. But the tight mon-etary policy contributed to further weakening of the economy while, after a short break, the ringgit was again started its downward descent. BNM analysts could hardly see any stable correlation between interest rate and exchange rate [22]. The offshore-onshore interest rate spread reached 20%, constraining the effectiveness of domestic monetary policy and causing capital outflow to continue. In this environment, the Malaysian authorities decided to impose severe capital account controls on Sep-tember 1, 1998. The main objective of capital controls was to regain monetary independence [23] and to terminate capital out-flows which was to be achieved by the effective elimina-tion of offshore market and insulating domestic interest rates from external developments. Restrictions eliminated practically all legal channels for the transfer of the ringgit assets abroad, required the repatriation of the ringgit held offshore to Malaysia within a month, prohibited the repa-triation of portfolio capital held by nonresidents for 12 months, and imposed tight limits on the transfer of capital abroad by residents. Residents were prohibited to grant ringgit credit to nonresidents. However, payment and transfers for international trade and FDI was not subject-ed to restrictions. All import-export transactions had to be settled in foreign currency. All transactions with Malaysians assets (stock, securities, etc.) could only be concluded and registered through authorized institutions like KLSE. Securities had to be repatriated to Malaysia – some 170.000 investors had their portfolios totaling 10 billion MYR frozen. In such an environment the BNM could easily exercise monetary policy and regain control over foreign exchange market. The ringgit has been pegged to the dollar at 3.8 MYR/USD and remains stable at this level until now (November 2000). Monetary policy swung from contrac-tionary to expansionary and central bank became engaged in providing liquidity for the domestic market. Rapid cuts of interest rates from 10 to 6% followed what was wel-comed by the domestic corporate and financial sector. CASE Reports No. 39 Figure 4-21. Malaysia: money M1 (bln MYR) 0 1994M1 1994M4 1994M7 1994M10 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 Source: IFS [21] Not only was inflation smaller than expected, unemployment did not become a serious problem, increasing from 2.6% in 1997 to 3.9% in 1998. The main burden of labor market adjustment was absorbed by migrant workers. Right after the emergence of crisis the government canceled the law giving migrant workers an automatic prolongation of their working-contract after expiration. [22] Gould and Kamin (2000) remark that monetary tightening may lead to counter-intuitive result of further currency depreciation, because it threatens to further weaken the banking system and the economy as a whole. The unstable relationship between interest rate and exchange rate can reflect market confusion about that issue. [23] As it is well known, the authorities can maintain only two out of three following things: control over exchange rate, freedom of capital move-ment and monetary policy independence at the same time.
- 71. [24] In some countries, e.g. USA, large pension funds are prohibited by law to invest in securities below "investment standard" according to rating 73 The Episodes of Currency Crises in Latin... Figure 4-22. Malaysia: Yearly inflation (%) 7 6 5 4 3 2 1 0 1994M1 1994M4 1994M7 1994M10 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 1999M1 1999M4 1999M7 1999M10 Source: IFS Malaysian authorities argued that the capital control was only a temporary measure in the crisis recovery plan but the international reaction was negative. Malaysian securi-ties were downgraded by rating agencies and the country lost a lot confidence among investors. Malaysia modified and eased capital account restrictions in February 1999 – partly because of the fear of massive capital outflow after the 12-month moratorium and because the domestic and external market situation stabilized, but also to prove that controls were indeed transitory. 12- month moratorium period was lifted and replaced with a system of declining exit tax. The highest levy amounted to 30% and was to be gradually decreased. This step was con-sidered as an improvement, so in April 1999 Malaysia was upgraded by rating agencies. Only a small proportion of invested funds were withdrawn in the period following the relaxation of capital controls [24]. Authorities continued to relax monetary policy – interest rate further decreased and fell below 3% in mid-1999. There is no consensus whether Malaysian capital restric-tions have been necessary, neither have they been effective. The main problem is the difficulty to separate the impact of these restrictions from the usual market developments. There are arguments that the vast majority of investors who wanted to withdraw from Malaysia had already withdrawn before September 1998, so the restrictions might not have actually been necessary. On the other hand, as Edison and Reinhart (2000) point out, the absence of speculative pres-sures on the ringgit, the exchange rate stability, sharp decrease in interest rates, reviving economy, steady inflow of new FDI's, and absence of any parallel or black exchange market provide the arguments that restrictions has been accepted and probably effective. 4.3.4. Financial and Corporate Sector Restructuring The consolidation of the financial system started in March 1998 and is continuing. The main components of Figure 4-23. Malaysia: Unemployment rate (%) 5 4 3 2 1 0 1994 1995 1996 1997 1998 Source: IMF agencies. CASE Reports No. 39
- 72. 74 Marek D¹browski (ed.) this process are: setting up the institutions in charge of cleaning up bank's non-performing loans (Danaharta), injection of capital to undercapitalized banking institutions (Danamodal), corporate debt restructuring, company merger program, and tightening of prudential regulations. Danaharta is a financial institution set up by the authorities with the mandate to deal with the problem of non-performing loans, and was expected to push borrow-ers and banks for fast restructuring. Its staff came mainly from the private sector. At the first stage, Danaharta acquired from banks at market value bad loans [25] that could not be restructured by banks themselves, with a price averaging about 40% of the book value [26]. The capital essential for the bailout operation was provided by the Ministry of Finance (1.5 billion MYR), and came from the issuance of zero-coupon bonds that had explicit gov-ernment guarantee. Through March 1999 Danaharta pur-chased 23 billion MYR of non-performing loans, or a quar-ter of such loans outstanding. In the second stage, Danaharta started to manage, restructure or/and execute the loans. In order to shorten maximally the period of financial clean-up and force banks to give up bad loans at prevailing low market value, the company built up a very strong system of incentives. Banks selling loans to Danaharta have a right to 80% of any profit that this company makes out of the loan, they can also amortize for five years the loss resulting from selling assets under its book value instead of recognizing it immediately. (In case of undercapitalization caused by such a transaction banks are eligible for recapitalization carried out by the Danamodal). Non-interest earning bad loans are exchanged with government guarantees, zero risk interest paying bonds. Danaharta has also special and unusual powers over the borrowers, whose debt has been acquired. Another body set up especially for restruc-turing large corporate loans is called the Corporate Debt Restructuring Agency. Through May 1999, the CDRA was reviewing 57 applications encompassing 31 billion MYR, out of which 2.5 billion was restructured. Danamodal is an institution set up by BNM in order to provide the additional equity to undercapitalized financial companies. Its capital is funded by BNM (3 billion MYR) and by the government zero coupon bonds (8 million MYR). It intervenes to restore adequate capital ratio, however, the effectiveness of each investment has to be assessed by the independent international specialists. After intervention, Danamodal takes a stake in the institu-tion proportional to its involvement, and engages in the management of the company (it appoints at least two, high rank board members). By mid-1999, Danamodal injected 6.2 billion MYR to 11 banking institutions. Another way of restoring adequate capital ratio in the system is provided by a merger program, which is con-ducted under the BNM encouragement and supervision. There is some concern, however, that the activities of Danamodal mean a hidden form of banking system nation-alization. On the other hand, the authorities take into consideration (re-)selling distressed financial institutions to private sector, mainly to foreign banks. What concerns prudential regulations, from 1998 the BNM requires from individual institutions to conduct monthly tests based on parameters given by the BNM. It has also issued "Minimal Standards on Risk Management Practices of Derivatives", "Minimum Audit Standards for Internal Auditors", and other guidelines for bank man-agers. In addition, it requires incorporating off-balance sheet items into loan classifications, and market risks into the capital adequacy norms. The BNM increased the fre-quency of on-site inspections, etc. The regulatory frame-works of capital market have been strengthened as well as disclosure and transparency rules. 4.4. Conclusions Generally speaking, Malaysia has avoided all the eco-nomic and social havoc characteristic of the crises in neighboring Indonesia, Thailand, and Korea. GDP fell tem-porarily (-7.5% in 1998) and then rebounded. The exchange rate depreciated around 50% and remained sta-ble at the level of 3.8 MYR/USD. There was no social unrest or widespread poverty problem. Malaysian eco-nomic fundamentals occurred to be much sounder than those of its neighbors [27]. The external debt remained manageable, exposure to exchange rate risk limited, domestic debt low enough to absorb the costs of econo-my restructuring, and level of moral hazard in financial institutions not so high. Financial institutions had a lower ratio of non-performing loans and higher capital, and there was a stronger banking culture, with better supervi-sion and prudential regulation than in other Asian economies. In 1999, the Malaysian economy grew by 5.8%, while in the first quarter of 2000 real GDP increased by 11.9%, [25] Minimal eligible loan amounted to 5 million of USD. [26] But excluding one large and extremely faulty loan, the average price jumps to 63% of the face value. [27] There are some opinions that Malaysia has been hit by the crisis unjustified, or by accident or by contagion only. Other opinions suggest that even without Asian crisis in 1997 Malaysia would have its own financial crisis in 1998–1999 due to overheating and speculative overinvestment in equi-ty and asset markets and excessive domestic credit expansion. CASE Reports No. 39
- 73. 75 The Episodes of Currency Crises in Latin... and in the second quarter by 8.8%. FDI continued to flow in, inflation moderated to 1.5%, trade surplus remained on the high level of 7.7 billion USD, and BNM's interna-tional reserves – at the level of 30 billion USD. The over-all risk-weighted capital ratio of the banking system amounted to 12.6% with non-performing loan propor-tion equal to 6.4%. The cost of fiscal stimulus amounted to 5% of GDP, also less than projected. The exchange rate remains stable, but on somehow undervalued level [28], which gives good prospects for current account but might put some inflationary pressure on the economy. Asset prices stand at around 60% of their pre-crisis level. The cost of restructuring of the banking sector totaled to about 10% of GDP, which was less than expected Contrary to the macroeconomic "soundness", micro-economic reforms have been very slow. The consolida-tion of the financial system has not been completed. Until August 2000, Danaharta acquired, restructured and dis-posed a total 31.5 billion MYR of non-performing assets and CDRA completed 27 restructuring cases, involving 23.6 billion MYR. However, there are allegations, that some of the "restructuring" that take place is just a reshuf-fling of debt among subsidiaries in order to avoid insol-vency. In other cases a debt solution is obstructed by the political connections of debtors. Although Malaysia is a leader in restructuring among post-crisis Asian countries, the failure to clean up the debt overhang can have serious negative consequences for the country. The stock market has not yet resumed the role of a fund-rising institution. Malaysia is still maintaining a rather restrictive capital account regime and because of that is probably not going to fully regain investors' confidence any time soon. Appendix: Chronology of the Malaysian 1997–1998 Crisis 1997 March 28, Malaysia central bank restricts loans to property and stocks to head off a crisis. early-May, Japanese officials, concerned about the decline of the yen, hinted that they might raise interest rate. Investors start gradual withdrawal from South East Asian markets. mid-May, The BNM defends ringgit with few days of very high interest rates. July 2, Thai baht collapses and turmoil begins. July 8, Malaysian central bank, Bank Negara, has to intervene aggressively to defend the ringgit. The inter-vention works for a while (the currency slightly appreci-ates). July 14, Malaysia central bank abandons the defense of the ringgit and engages in stabilizing domestic money market with relatively loose monetary policy. July 24, Ringgit hits 38-month low of 2.65 MYR/USD. July 26, Prime Minister Mahathir blames George Soros and other "rogue speculators" for the attack on the ring-git. September 4, Ringgit breaks through 3 MYR/USD. September 20, Mahathir tells the public, that specula-tion is immoral and should be stopped. October 1, Mahathir repeats his call for tighter regu-lation, or a total ban on forex trading. The ringgit falls 4% in less than 2 hours to a low of 3.4MYR/USD. October 17, Malaysia tightens budget in effort to stop the economy from sliding into recession. December 5, Malaysia finally and radically changes its policy and imposes tough reforms in order to deal with a crisis. These include an 18% cut in government spending, restriction on large-volume import, on bank credit and in stock market regulations. There were to be "no question of bailout" for financially ailing companies. 1998 January-February, Several increases in BNM's inter-vention interest rate were planned to stop the currency from depreciating and restrict inflationary pressures March, The severity of the crisis is gradually recog-nized and the fiscal policy changes to more expansionary. The fiscal package of 3 billion MYR (1% of GDP) is announced and a drastic revision of federal budged aims the deficit of 2.6% of GDP. July, Another additional fiscal package (7 billion MYR (2.5% of GDP)) announced in order to stimulate the economy. January-August, Despite the austerity fiscal measures and firm monetary policy the crisis and the capital outflow continue. September 1, Malaysia introduces capital controls; financial investment can be repatriated only after a 1-year period. Rental and profits from sales can be repatriated. 1999 February 5, Malaysia replaces one year holding period with exit tax. Repatriation of principal and profits will be subjected to a maximum levy of 30%. [28] The estimates of current undervaluation (exchange rate of 3.8 MYR/USD) reach even 19%. CASE Reports No. 39
- 74. 76 Marek D¹browski (ed.) References Annual Report on Exchange Rate Arrangement, IMF 1996, 1997, 1998, 1999, 2000 Bank Negara of Malaysia (BNM), Annual Report, 1998, 1999 Corsetti G., Pesenti P., Roubini N.(1998). ”What Caused the Asian Currency and Financial Crisis”, Banca d'Italia, Temi di discussione 343. Claessens S., Djankov S., Lang L.(1998). ”East Asian Corporates: Growth, Financing and Risks over the Last Decade”. Mimeo, World Bank. Dato' Shafie Mohd. Salleh (2000). Statement by the Hon. Dato' Shafie Mohd. Salleh, Gov. of the Fund and the Bank for Malaysia at the Joint Annual Discussion, IMF. Delhaize P. (1999). ”Asia in Crisis”. John Wiley&Sons. Edison H.J., Reinhart C.M. (2000). ”Capital Controls During Financial Crises: the Case of Malaysia and Thai-land”. Board of Governors of the FRS IFDP 662. Edison H.J. (2000). ”Do Indicators of Financial Crises Work? An Evaluation of an Early Warning System”. Board of Governors of the Federal Reserve System IFDP 675. Financial Times, 1996: 26-17 X; 1997: 19 V, 21-22 VI, 26-27 VII, 8 X. Gould D.M., Kamin S.B. (2000). ”The Impact of Mo-netary Policy on Exchange Rates During Financial Crises”. Board of Governors of the Federal Reserve System IFDP 675 IMF (1999a). Malaysia, selected issues. IMF (1999b). Malaysia: Recent Economic Develop-ment. IMF (2000). Recovery from the Asian Crisis and the role of the IMF. Milesi-Ferretti G., Razin A. (1996). ”Current Account Sustainability, Selected East Asian and Latin America Experiences”. IMF WP 96/10. Kamin S. (1999). ”The Current International Financial Crisis, How Much is New?”. Journal of International Money and Finance, vol.18, no.4. Krugman P. (1994). ”The Myth of Asia's Miracle”. For-eign Affairs, Nov.-Dec. Krugman P. (1998). ”What Happened to Asia”. Mimeo. Sarel M. (1997). ”Growth and Productivity in ASEAN Countries”. IMF WP 97/97. Sarno L., Taylor M.P. (1999). ”Moral Hazard, Asset Price Bubbles, Capital Flows and East Asian Crisis, a First Test”. Journal of International Money and Finance, vol.18, no.4, 1999. Stone M.R. (1998). ”Corporate Debt Restructuring in East Asia: Some Lessons from International Experience”. IMF PPAA 98/13. World Bank (1998). ”Responding to the East Asian Cri-sis”. CASE Reports No. 39
- 75. 77 The Episodes of Currency Crises in Latin... Part V. The Indonesian Currency Crisis, 1997–1998 by Marcin Sasin 5.1. Overview 5.1.1. Introduction Indonesia, the biggest archipelago in the world, located in Southeast Asia, declared its independence from the Dutch and Japanese in 1945 and has been a republic since then. The President, as the head of the state and chief of the cabinet, is elected every five years. Since 1967 (till 1998) this post was occupied by Soeharto, who ruled in an authoritarian manner. The elctoral system has been constructed to always secure the victory of the (then) ruling party, "Golkar". The power of President Soeharto and the Golkar has been derived from its close relationship with the military and administration – the military officers, under the doctrine of "dual function", have always served in civilian positions, such as cabinet ministers, local governors, heads of state corporations, supreme court judges, etc. In the last elections before the crisis, in May 1997, Golkar increased its share in the Parliament from 68% to 74%. The activities of the Parliament were merely to ratify the government's (President's) directives. 10 5 0 -5 -10 CASE Reports No. 39 The 200-mln population is a complex ethnic and reli-gious mixture (Muslim, Christian, and Buddhists). The Indonesian-Chinese, although constituting only 3–4% of the population and having little political resources in their dis-posal, are disproportionately important in trade and com-merce – they control about 80% of total private capital, ranging from the biggest corporations to small provincial shops and stores. In the Muslim majority, notably among the poor, there is a feeling of exploitation by the Chinese minor-ity. Social and ethnic tensions have been hidden under the authoritarian grip of the system. In addition, some provinces of Indonesia, namely Aceh, Papua and, till recently, East Timor – possess strong independence movements and often take active armed resistance, which is quickly sup-pressed by the authorities. 80% of the Indonesian economy is private, but there are still large state-owned companies – notably in the oil and mineral extraction sectors. The dominance of politics over the economy is the key issue when analyzing the Indonesian economic system. Under Soeharto regime, this system relied on the partnership between Chinese business and the military, as well as politically connected civilians (often the Figure 5-1. Indonesia: GDP growth (% p.a.) -15 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: IMF
- 76. 78 Marek D¹browski (ed.) Figure 5-3. Indonesia: Employment by sector CASE Reports No. 39 Figure 5-2. Indonesia: GDP by sector Agriculture, mining, etc. 25% Manufact. and costruction 34% Services (and other) 41% Source: IMF relatives of senior government officials) in which the latter took a share in profits in exchange for providing protection and preferential access to contracts, concessions, licenses, tax exemptions etc. The same system of concessions and restrictions prevented others from participating in the most profitable economic activities. The children of the President were owners and chiefs of banks, oil firms and other strate-gic companies – corruption and nepotism were widespread. Indonesia was notoriously ranked in the last position in the Transparency International rankings [1]. The economy has been, to large extent, monopolized – from oil extraction to clove cigarettes – there was no Western standard competi-tion law. Imports and distribution of essential food items were monopolized by state agency Bulog; food, electricity and other everyday products were subsidized. Nevertheless this system, which derived its legitimacy from sustained improvement in the standard of living of the mass Indonesians, has been so far successful in achiev-ing its goal of fast development. As a result of political sta-bility and the impressive mobilization of country resources (saving rates around 30%, investment rates around 30–35% of GDP- in the 1990s), Indonesia experienced three decades of permanent growth and transformed itself from an underdeveloped, impoverished, agricultural and mineral-exporting country to a rapidly urbanizing industri-al economy. In the years preceding the crisis, growth aver-aged around 7% per year. Recent reforms reoriented the economy so as to reduce its dependence on the oil sector, encourage the creation of competitive non-oil export-ori-ented industrial infrastructure and expand the role of the private sector. In early 1997, GDP per capita stood at around 1140 USD, while total GDP was about 225 bln USD. The primary sector (agriculture, mining, etc.) had 25% share in total GDP and about 40% of total employ-ment. Manufacturing and construction constituted 35% and employed about 20% of the labor force. Tertiary sec-tor (services etc.) had around a 40% share in both GDP and employment. 5.1.2. Monetary Policy and the Financial Sector The central bank of Indonesia is Bank Indonesia (BI). Its task is to conduct monetary policy to stabilize macroeco-nomic environment, issue the national currency, the rupiah (IDR), handle foreign exchange assets and debt servicing, as well as exercise control over national banking and financial system. To this end, BI monitors broad money, credit aggre-gates and reserve money. In addition, the authorities moni-tor the real value of the rupiah against an undisclosed basket of currencies. The main instrument in BI actions are open market operations involving Bank Indonesia papers (intro-duced in 1984) and commercial bank papers (from 1985), as well as reserve requirements and foreign exchange inter-ventions. In the 1990s, BI has succeeded in limiting inflation only to rather moderate level of 7–11%, however, in a peri-od leading to the crisis inflation was steadily falling and reached about 5% in mid-1997. Foreign reserves, together with reserve money, have been rising largely due to capital inflows, while domestic credit accelerated, triggered by financial liberalization. In the 1990s, BI had permanent problems with rapidly growing [1] In 1995: 41/41 (41 place out of 41 reviewed countries), 1996: 47/54, 1997: 46/51, 1998 80/95. Agriculture, mining, etc. 42% Manufact. and costruction 20% Services (and other) 38% Source: IMF
- 77. 79 The Episodes of Currency Crises in Latin... credit (over 20% a year in the period preceding crisis). Because of that, BI has conducted a relatively restrictive pol-icy of high interest rates. Reserve requirements have been raised in accordance with BI plan to reduce credit growth to 17% in 1997 [2]. Bank lending rates in the 1990s have been usually 10% or more above inflation. The exchange rate policy can be described as real exchange rate targeting – with the rupiah gradually and predictably depreciating with respect to the USD about 4% year, i.e. from around 1900 10 8 6 4 2 1997M1 1997M2 1997M3 1997M4 1998M1 1998M2 1998M3 CASE Reports No. 39 IDR/USD in 1990 to around 2400 IDR/USD in mid-1997. Such a policy combined with capital account liberalization created incentives to borrow abroad to take advantage of interest rate differentials and prepare the ground for large capital inflows. BI tried to counteract excessive inflows by widening the rupiah's trading bands from 2% to 3% around daily mid-rate in 1995, and further to 5% in June 1996, and to 8% in September 1996, adding therefore some risk to foreign exchange market. Before the crisis, the size of the Indonesian financial sec-tor amounted to some 60% of GDP [3]. The leading finan-cial institutions in Indonesia were commercial banks, accounting for 87% of total assets, out of which the biggest seven state-owned banks accounted for around half of that figure, the other half being distributed among around 170 private banks (1995). Other financial institutions like insur-ance companies, pension funds, stock brokerages or other financial intermediaries played only a minor part in the sys-tem and had no impact on the overall picture. The process of liberalization of the banking system began in 1983 with interest rates liberalization and the elim-ination of credit ceilings. But ever since the government opened up the system to new entrants in 1988–89, the sec-tor started to thrive. In 1988, reserve requirements were reduced from 15% to 2%, licenses for new private and joint-venture banks issued, and state-owned firms were allowed to put 50% of their funds with private banks. The following year, the requirement of BI license in long-term loans granting and offshore loans ceiling was removed. The number of banks soared from 112 to around 240 as anyone Figure 5-4. Indonesia: Size of Indonesian banking system, assets of banks in 1996 Joint 6% Foreign 4% Local state 3% State commercial 40% Private national 47% Source: IMF Figure 5-5. Indonesia: Trading volume on forex market (bln USD/day) 0 forward&swap spot Source: IMF [2] The plan failed to large extent and eventually credit growth reached 23%. [3] Data for 1994. Compare Malaysia 100% of GDP and Thailand 110% of GDP.
- 78. 80 Marek D¹browski (ed.) CASE Reports No. 39 800 700 600 500 400 300 200 100 with access to around 3 mln USD minimum capital could set up shop. So did bank credit that rose by 350% from 1988 till 1995. In 1994, new private banks overtook state banks with lending activity. Without a proper supervision frame-work, and in combination with a severe shortage of trained and experienced bankers, this quickly led to a problem with prudent asset management and bad debts. To counteract this problem, BI introduced a minimum capital adequacy ratio of 8% and gradually increased minimum capital need-ed to open a bank to around 30 mln USD. In 1992, state owned banks were converted into limited liabilities compa-nies. However, the Ministry of Finance announced in 1994 that it would not permit a state bank to default on its oblig-ations. This was not the end of banking fragility and problems. Financial scandals and bank failures were quite frequent, with government intervening to bail out bankrupt banks and cover deposits [4]. Noncompliance in lending limits and off-balance sheet operations were widespread. Strong polit-ical dependence of state banks, which were used as a source of cheap capital for government affiliates' enterprises, explains their permanent inferior performance, in terms of both return on assets and bad loans. After liberalization, the 1990s witnessed a rapid devel-opment of the securities market. In 1988, the market was opened to foreign investors, in 1994 Bapepam (the state supervisory agency) implemented new accounting stan-dards, and in 1995 a new computerized automatic trading system was introduced allowing for much higher trading vol-ume. Despite these changes, the securities market never became a major source of commercial finance. In 1994, equity market capitalization was only of 30% of GDP [5]. Moreover, its importance is still overestimated because cap-italization includes shares that have never been sold – 70% of the shares are held by companies' founders [6], while the Figure 5-7. Indonesia: Equity market size in % of GDP 35 30 25 20 15 10 5 0 1989 1990 1991 1992 1993 1994 Source: IMF Figure 5-6. Jakarta Composite Index 0 1992M1 1992M5 1992M9 1993M1 1993M5 1993M9 1994M1 1994M5 1994M9 1995M1 1995M5 1995M9 1996M1 1996M5 1996M9 1997M1 1997M5 1997M9 1998M1 1998M5 1998M9 1999M1 1999M5 1999M9 Source: Bloomberg [4] Examples include looting of Bank Bapindo, or Bank Lippo scandal. [5] Compared to 280% in Malaysia and 95% in Thailand. [6] Indonesia has the highest ownership concentration in corporate sector among Southeast Asian countries. As it has been stated above, main Indonesian companies are characterized by the close connections with the authorities, strong family ties, and are monopolized by the small elite of Chi-nese businessmen, and senior government officials.
- 79. 81 The Episodes of Currency Crises in Latin... government holds large stakes in privatized and listed for-mer- state-owned companies. When adjusted for that the stock market has provided about 15% of total business finance. The small size of the market and open access to it are the reasons why the Indonesian stock exchange was 70% dependent on foreign investors and relatively volatile. International investors complained about small liquidity, poor audit standards, doubtful fairness and quality of com-panies' disclosures as well as notorious insider trading. Capital account liberalization started with a gradual pro-motion of foreign direct investment in designated sectors and with efforts to restructure and modernize the economy from oil-export dependence [7]. Investment procedures were simplified, certain restricted fields gradually opened, equal treatment with domestic investment sequentially granted, etc. In 1989 so called priority list ("in what fields to invest") was exchanged by so called negative list consisting of domains in which foreign investment was restricted. The number of items on the negative list was decreasing. As a result, foreign direct investment has been steadily pouring into Indonesia, especially from 1995 (in 1996/97 accounting year [8] FDI inflow amounted to 6.5 bln USD). In 1988, Indonesia accepted the obligation of Article VIII [9] – the foreign exchange market was developed and the swap transactions liberalized. The forex market remained rela-tively shallow with a 10 bln daily turnover in mid-1997. In 1989 and 1996 the authorities liberalized portfolio capital inflows by eliminating quantitative limits on banks' borrow-ing from nonresidents, foreigners were permitted to freely invest in stocks up to 49% of share capital, and the central bank withdrew from its obligatory intermediation in foreign exchange transactions. High interest rates, a stable rupiah exchange rate, a growth-oriented environment and an expanding stock mar-ket resulted in constant inflows of foreign portfolio and short term capital (7 bln USD in acc. year 1995/96 and 6 bln USD in acc. year 1996/97). As a result, foreign reserves in Bank Indonesia have been steadily increasing. BI attempted to sterilize the resulting increase in base money by sales of central bank papers and through interventions on foreign exchange market. Capital inflows mainly took the form of borrowing of commercial banks, which in turn were con-verted it into local currency and lent to domestic corporate sector. 5.1.3. The External Sector As a result of years of trade protection, Indonesia is not a particularly open economy in Southeast Asian terms – the average trade [10] is about 25–30% of GDP. The most important export items are oil and natural gas – their share in total exports is one-fourth. Indonesia has made an effort to reduce its dependence on oil, and the export composition weights towards manufactured goods, such as textiles, wood products and electrical appliances, and now the share of man- Figure 5-8. Indonesia: Capital account (bln USD) 15 10 5 0 -5 -10 -15 92/93 93/94 94/95 95/96 96/97 97/98 98/99 [7] In the beginning (mid-1980s) firms were encouraged to export all their production and to invest in remote areas. [8] In Indonesia the reporting and accounting year starts in April. [9] Article VIII of the IMF's Articles of Agreement, i.e. general obligations of members. [10] (import+export)/2/GDP. CASE Reports No. 39 total net official (=new loans-repayment) FDI short term Source: IMF
- 80. 82 Marek D¹browski (ed.) Figure 5-9. Indonesia: rupiah exchange rate (IDR/USD) IDR/USD 18000 16000 14000 12000 10000 8000 6000 4000 2000 0 95-11-16 96-01-16 96-03-16 96-05-16 96-07-16 96-09-16 96-11-16 97-01-16 97-03-16 97-05-16 97-07-16 97-09-16 97-11-16 98-01-16 98-03-16 98-05-16 98-07-16 98-09-16 98-11-16 99-01-16 99-03-16 99-05-16 99-07-16 99-09-16 99-11-16 00-01-16 00-03-16 00-05-16 00-07-16 00-09-16 Source: Bloomberg ufactured goods in export is about 30%. Indonesia is also dependent on oil/gas sector-related imports (that accounts for 10% of total import) as well as on intermediate and capi-tal goods imports. There is one major item to be mentioned with respect to imports – Indonesia was importing food (rice, wheat, etc.) for about 2–3 bln USD in years before crisis. Major trade partners are Singapore, US, Japan and the Euro-pean Union. To meet the WTO criteria Indonesia was gradu-ally reducing its import tariffs and they averaged 12% in 1997. Because the authorities were monitoring and targeting the real exchange rate, the rupiah overvaluation was not a major issue in Indonesian external position – in 1990s before crisis it has been fluctuating around 100% (+/- 3.5%). At the end of 1996, due to an extremely weak Japanese yen, the CPI-based real effective exchange rate was somehow overvalued and stood at 105% [11]. But with a tradition of sluggish and steady rupiah adjustment, there was no threat of any drastic exchange rate adjustment. Figure 5-10. Indonesia: Current account (% of GDP) 20 15 10 5 0 -5 -10 -15 92/93 93/94 94/95 95/96 96/97 97/98 98/99 total trade services+net factor income Source: IMF [11] It must be noted that there are well known problems in assessing real exchange rate, so estimates differ depending on the source. There is a CASE Reports No. 39 consensus, however, that the real exchange rate was in the neighborhood of its parity equilibrium.
- 81. 83 The Episodes of Currency Crises in Latin... Figure 5-11. Indonesia: Export by sector Other Manufacturing Electr. appliances 7% In the 1990s, Indonesia was experiencing moderate cur-rent account deficits. From some 4% of GDP, the deficit has been contained to around 1% in mid-1990s and started to increase in the period leading to the crisis to reach 3–4% of GDP – it was, however, very reasonable compared to other Southeast Asian countries and could be interpreted as long term consumption smoothing. In addition, the trade balance was positive. The key items responsible for the current account deficit were the 7,6 bln USD capital goods import induced in large part by foreign investment, permanently negative service balance (3 bln USD in oil/gas service pay-ments in 1996/97) and especially net factor income balance (6 bln USD of investment interest payments). These figures suggest that the side effects of a large long-term capital inflow (i.e. interest payment and capital good import) were the main factors in current account developments. As these payments were also the most inelastic items in the current account it can be concluded that this deficit had a rather structural character. Notwithstanding its moderate size, Indonesian authorities have taken some fiscal actions to contain these imbalances. 5.1.4. The Public Sector The size of the public sector in Indonesia is small. In 1996/97 federal governments raised 16% of GDP and spent 14.6% of GDP. The local finances did not exceed 1–2% of GDP. Direct taxes provided 56% of revenues, out of which the most important is a tax levied on oil production (23% of total budget). 33% came from indirect taxes, remaining 11% were derived from non-tax sources. On the expendi-ture side, the current (operative) expenditures had a 60% share, while developmental (plus net lending) – remaining 40% of total expenditures. The authorities provided many subsidies, notably to food and electricity. The fiscal sector was far from being transparent. The participation in public infrastructure projects has been subject to restrictions and entails unclear, noncompetitive bidding. Well-connected companies could count on numerous tax concessions and exemptions, borrowing on favorable terms, etc. The fiscal balance was threatened by the possibility of a bailout of some insolvent financial institutions, part of budget deficits could be hidden in the balance sheets of state-owned banks. In the late 1980s, Indonesia came through a process of fiscal consolidation, which resulted in the very positive CASE Reports No. 39 11% Copper 3% Other Minerals Rubber 4% 3% Animal products 3% Other Agriculture 3% Oil 14% Gas 10% 12% 30% Textiles Wood products Source: IMF Figure 5-12. Indonesia: government revenues VAT 24% Oil/gas income tax 23% Other direct taxes 33% Nontax revenue 11% Other indirect texes 9% Source: IMF
- 82. 84 Marek D¹browski (ed.) record of budget surpluses in the 1990s. Rather conserva-tive fiscal policy was aimed at reducing domestic demand stemming from rapid credit expansion, as well as containing the current account deficit. In the accounting year 1996/97, the conventionally measured government budget surplus stood at 1.4% of GDP. Prudent fiscal policy, combined with high rates of eco-nomic growth, led to a declining public debt ratio, which in 1997 equaled some 25% of GDP [12]. The government external debt in 1996/97 amounted to 56 bln USD or 23% of GDP – 44% denominated in USD and 37% in Japanese yen. But the main problem was a pri-vate sector external debt officially estimated at 57 bln USD. Given attractive interest rate differentials and a pretty free capital flow regime, the private sector got heavily indebted ( mostly with short term debts). There is no reliable data on the extent of the debt or the precise maturity structure [13] – authorities were obviously interested in underestimating its share. According to the World Bank, 25% of the debt was short term, among Bank of International Settlements members, controlling most of Indonesian debt, this ratio was 61%. Total external debt was officially 113 bln (around 50% of GDP). Other sources estimate this number to be around 130 bln USD, or even more. The debt required 15 bln USD debt-service payments in 1996/97. 5.2. The Crisis 5.2.1. Introduction The direct cause of the Indonesian financial crisis was contagion from its Southeast Asian neighbors. After the Thai baht was floated, the general Asian market risk was reassessed sharply and caused adjustments which Indonesia couldn't endure or counteract. The mechanism of crisis development was more or less similar to other countries [14]. A gradual (yet fairly visible) deteriorating macroeco-nomic situation in the country was making international cap-ital market anxious about the further profitability of capital investments in a given country, and generally about the future prospect of Asian-style capitalism. A long period of constant growth gave domestic firms a false impression of stability. A highly competitive and poorly regulated banking sector engaged itself in more risky projects to take advan-tage of an investment boom. After the economy overheat-ed, a slowdown was expected, some short-term capital was withdrawn, and a number of domestic firms hedged against possible currency depreciation. The downward pressure on the currency forced the authorities to choose one of the alternative solutions. First, if they had sufficient foreign reserves, they could defend the currency by hiking interest rates, but it would lead to a contraction of the economy and could be accept-able only as a temporary measure (longer debt maturities are usually not affected). This problem becomes more seri-ous if a large portion of domestic debt is short-term. A pro-longed period of high short-term interest rates can force some cash-short companies to postpone investments or even default on their obligation. One default can cause another default, and suddenly the economy falls into illiquid-ity with serious contractionary consequences. Things get even worse if banks and firms invest in assets (like stocks, real estate) as asset prices usually decrease substantially in a crisis. Secondly, the authorities could let the exchange rate depreciate. But if the economy has a large stock of external debt, especially in the private sector, or if companies bor-row in foreign currency, having receipts in domestic curren-cy (and unhedged) can become overwhelmed by the increase in their debt and get cut from any financing and eventually fall. There is usually no need for creditors to liq-uidate pending deposits – when the debt is mainly short term it is sufficient only to postpone or refuse to roll over. Illiquidity turns into insolvency – the ratio of bad loans soars and if the banking system is weak, poorly managed and undercapitalized, some banks may go bankrupt. If there is little confidence in the financial system, there could be a run on other banks and a collapse as well. Authorities then have to get engaged in an emergency bail-out and massive capital injection, the monetary base explodes, inflation rises, cur-rency plummets, depreciation fuels inflation (and vice verse), and external debt problems increase. Confidence among foreign investors evaporates. The lack of political will to stick to harsh anti-crisis measures and social unrest only makes the problem more serious. So if there is both a large stock of short-term debt and a large stock of unhedged external debt, combined with other ingredients like a shortage of foreign reserves, a weak finan-cial system or political instability, the authorities might have no degree of freedom to maneuver and no power to reas-sure foreign and domestic investors about the safeness of [12] Compared to 55% in 1987. [13] Bank Indonesia probably lost track of all private sector foreign borrowings. Other estimates suggest 65 bln USD or even 80 bln. There are also well understood problems with maturity classification of different loan arrangements – the average maturity of foreign debt was in the neighbor-hood of 18 months. [14] Japan was the first example, that the times of highly regulated economies whose growth is based on extensive mobilization of national (or for-eign) resources might come to an end, and that their growth prospects has to be revised. CASE Reports No. 39
- 83. 85 The Episodes of Currency Crises in Latin... their investment. Such a situation falls under the term "vul-nerability". When this becomes common knowledge, spec-ulators join capital outflows and sell the domestic currency short. At this point, it is virtually impossible to reverse the crisis without a cost. In this respect, Indonesia typifies the worst-case scenario. 5.2.2. Indonesia's Vulnerability Analysis For decades, economists were making an effort to improve crisis predictability and their work is far from being accomplished. For some time, the occurrence of a curren-cy and financial crises is explained in terms of "vulnerability". When certain conditions are satisfied there is an increased probability of a crisis, but its timing and actual occurrence are determined by market sentiment and shifts in expecta-tions. Usually some macroeconomic indicators (called "leading crisis indicators") endowed with above average predictive power exceed their standard values before the crisis and this is the most frequent way of describing vulnerability. It is possible to construct a so-called "early warning system" (see, for example, Edison (2000)) based on the combination of these indicators that would issue a signal warning against a possible crisis. However, the performance of such systems leaves much to be desired [15]. Many researchers come up with many proposals for leading indicators but the most popular include: real exchange rate, foreign reserves, cur-rent account, the level of short term debt (external and domestic), rapid credit growth, fiscal and monetary expan-sion, short-term capital flows as well as other, hardly mea-surable features like the extent of moral hazard ("the incen-tive structure of financial system"), exposure to contagion etc. From the point of view of a domestic investor borrow-ing externally, or a short-term international investor holding the rupiah, the most important point is IDR/USD exchange rate, at which they exchange their rupiah receipts into hard currency. They thus count on the authorities that the policy of assuring exchange rate stability won't be altered. For the defense of the currency against pressures, there is an essen-tial need of foreign reserves. Indonesian foreign exchange reserves stood at about 20 billion USD in mid-1997. Total external debt service to foreign reserves ratio gives an answer to the question of whether the country can serve its current obligations. This debt service (interest and amorti-zation) was about 15 bln USD in 1997, producing a danger-ously high ratio of around 75%. This means that without further foreign loans, Indonesia would have had problems meeting its obligations. Debt can be serviced also by the export proceeds. But the ratio of external debt service to export stood at around 30% – the highest among Southeast Asian countries. Once all short-term creditors would like to withdraw their funds at one moment, would the reserves be sufficient enough to meet their demand? Again, the ratio of total external short-term debt plus external debt service to foreign reserves, depending on the estimate, ranged from 210% to 320%. It has been therefore absolutely essential for Indonesia to have its short-term debt rolled over, otherwise it would have to default on its obligations. The country's external finances depended wholly on the sentiments of short-term foreign creditors and domestic investors' willingness to hedge foreign debt, as well as their Figure 5-13. Indonesia: Total reserves minus gold (bln USD) 30 25 20 15 10 5 0 1996M1 1996M3 1996M5 1996M7 1996M9 1996M11 1997M1 1997M3 1997M5 1997M7 1997M9 1997M11 1998M1 1998M3 1998M5 1998M7 1998M9 1998M11 1999M1 1999M3 1999M5 1999M7 1999M9 1999M11 Source: IFS [15] Actually, Indonesia did fairly well in the rankings of crisis probability. CASE Reports No. 39
- 84. 86 Marek D¹browski (ed.) beliefs about the future exchange rates and the behavior of other market participants. A second indicator of financial solvency is the money to foreign reserves ratio. In the event of financial panic and irrational domestic asset selling, the central bank should be able to cover its liabilities (reserve money) – theoretically all the rupiah (cash or power-money) can be exchanged into hard currency if the central bank sticks to its exchange rate. A ratio below 100% is a good sign for the soundness of the system. In Indonesia it was 100%. In practice, BI acted as a lender of last resort, so if it was ultimately determined to support the banking system in the event of financial panic, all liquid money assets (M1 or even M2 [16]) could potentially be converted into foreign currency. Therefore the ratio of monetary aggregates to foreign reserves is another leading indicator of possible distress [17]. The ratio of money (M1) to foreign reserves was 135%, but more commonly used M2/foreign reserves ratio, as a result of thriving banking sec-tor activities, was about 700%, which can be regarded as very dangerous [18]. On the other hand, the informative content of this indicator is not very clear, as M2/FX ratio is to large extent country specific and reflects rather the development of domestic banking system. Although the indicator of the real exchange rate and fis-cal stance were well below the warning level, the current account deficit needs to be addressed. Throughout the 1990s, the balance was negative, while standing at 3% of GDP in 1996 it was not a big problem in the period leading to the crisis. Nevertheless, when adjustment had been made for usual oil/gas surplus the deficit would have risen to more than 5% of GDP. The variability of oil prices and ever increasing foreign debt service payments might have cast some doubts on its sustainability. The notion of sustainability can be implemented by introducing non-increasing foreign debt to GDP ratio. The current account is sustainable if it doesn't cause an excessive build-up of foreign debt. By taking arbitrary 1% difference between long-run interest rate and long-run growth rate Corsetti, et.al. (1999) show that a sustainable current account in case of Indonesia equals about 3.3% of GDP, which was more or less equal to its actual record. In accor-dance to that finding, Milesi-Ferretti and Razin (1996) argue that the Indonesian deficit pattern in the 1990s matched the consumption smoothing theory rather closely, so it should be nothing wrong with such a deficit. But the consumption smoothing theory predicts that at some moment in future there would be a switch into a surplus. So, from a political-economic point of view, the deficit is sustainable if it can be reverted into a surplus according to a optimal development path, without a crisis or drastic policy change. But in fact, in 1996–97, the current account imbalance seemed rather structural. FDI inflows, the main source of current account financing, were never sufficient to cover the deficit, and Indonesia had to rely on new foreign loans. This led us to the issue of sustainability of capital inflows, and thus, sustainabil-ity of economic growth. Indonesia grew at average annual rate of about 7% in the first-half of the 1990s. That was possible thanks to sizeable capital inflow, and, vice versa, the capital was attracted by anticipated high rates of growth and invest-ment opportunities. However, the abundance of capital, high investment rates and rapid credit expansion led to deterioration of the quality of investment projects. Politi-cally connected monopolies paid no attention to cost reduction, a bulk of government investments was designed under political or propagandist considerations rather than out of efficiency reasons. The private corporate and bank-ing sector, facing increased supplies of capital and a tradi-tion of poor regulation, turned to more risky projects. Claessens et.al. (1998) remark that the return-on-assets ratio has decreased in Indonesia from 8% between 1988–1994 to 5.5% in 1995–1996 [19]. However, the main vulnerabilities of Indonesia originat-ed from a building up of short-term external debt, a shaky domestic banking system and rising political instability. Aging President Soeharto showed not only no signs of retiring but had also been very reluctant to even discuss the issue of his successor and his idea of how the transition of power might be accomplished without political and social tension what in presence of Indonesian political strife and social/ethnical unrest has been a matter of importance. There was also a lack of confidence that in an event of a cri-sis the corruption-ridden government could deal with it effectively. [16] If banks allow the depositors to break time deposits. [17] Compare Obstfeld and Rogoff (1995). [18] It is worth to notice that in November 1994, just before Mexican crisis M2/foreign reserves were 9.1 in Mexico, and 3.6 in Brazil and Argenti-na. In Malaysia it was 4.8. [19] The issue of capital productivity is closely linked to the ongoing and yet unresolved debate about the causes of the Asian miracle, namely whether the fast growth of Asian countries results just from abundance of capital and labor force or from productivity growth. The first view has been advanced by Young in the beginning of the 1990s and popularized by Krugman (1994, 1998). They argue that the total factor productivity (TFP) in East Asia has been significantly lower (sometimes even close to zero) than the rate of GDP growth. There are also studies that contradict this view. Sarel (1997) finds that TFP growth in Indonesia in 1978–1996 has been quite remarkable and amounted to 1.2% per year. Also the estimates of incremen-tal capital output ratio (ICOR – a measure of capital productivity) show the decrease form 4.0 in 1987–92 to 3.8 in 1993–96 indicating a some improve-ment in investment efficiency. But on the other hand (according to World Bank data), ICOR has been rising (weakening efficiency) till 1987 and again CASE Reports No. 39 from 1994.
- 85. 87 The Episodes of Currency Crises in Latin... Table 5-1. Noncompliance to prudential rules (in number of institutions). Capital adequacy ratio (Car), Legal lending limits (Lll), Loan-deposit ratio (Ldr) What concerns the banking sector, its weakness was structural. The audit standards, transparency and compli-ance to prudential principles record was astonishingly poor. In 1995, half of private banks and 40% of state banks failed with respect to either capital adequacy ratio, legal lending limits or loan deposit ratio. Another issue was that state banks did not necessarily make loans on a commercial basis and were subject to polit-ical pressures. This problem was also present among private banks often maintaining close connections with their bor-rowing customers [20] which created incentives for risky or even fraudulent lending to these customers and did not encourage accurate loan monitoring. Poorly capitalized and monitored banks competing with other similar small banks on a segmented market had even more incentives to make riskier loans, especially if the management expected to become bailed-out if things go wrong (moral hazard). This quickly led to the problem of bad loans – about 10% of total loans were classified as non-performing [21]. State banks had an especially bad record with this share at 17%, while private national banks had, on average, 5% of their loans non-performing. Indonesia also had previous experience with banking scandals, financial sector bankrupt-cies and even bank runs. Instead of closing insolvent and bankrupt banks, the authorities arranged bailouts, encour-aged mergers and provided other forms of support. They also announced that no state bank would be left alone to default on its obligations. Such actions were the reason why bank managers could have an impression of having (implic-it) government guarantees. The moral hazard problem became even more serious in the presence of a poor bank-ruptcy law and inconsistencies in law enforcement. The combination of a relatively open financial market, growing eagerness of global investors to put money into Asia, high Indonesian interest rates and expanding Indone-sian corporate sector resulted in large capital inflows – indeed larger than banks and companies could wisely invest. The dangerous side effect of this inflow and market circum-stance was a mismatch in the balance sheets of banks and corporation. Maturity mismatch resulted from the use of short-term debt to finance long-term projects, while cur-rency mismatch emerged from the use of foreign-currency denominated loans to credit local currency earning projects. In the pre-crisis period this mismatch has not created many problems because of exchange rate stability and a tra-dition of a smooth rollover of short-term debt. The above-mentioned developments (large interest rate differentials, open capital market, (false) impression of exchange rate stability, growth environment, smooth debt rollover) were responsible for the key component of Indonesian vulnerability, i.e. the existence of immense unhedged foreign currency liabilities. In July/August 1997, one of the major global financial consulting companies sur-veyed 34 Indonesian chief financial officers. 2/3 of them had more than 40% of their debt in foreign currencies. Half of this amount was completely unhedged, and most of the rest CASE Reports No. 39 Total in category Car Lll Ldr State 7 0 2 1 Private national 166 18 56 11 Local development 27 2 3 0 Foreign and joint 40 1 9 6 Total 240 21 70 18 Source: Montgomery (1997), BI, data for 1995 Table 5-2. Banks' performance: Return on assets (Roa) and Car 96/97 97/98 98/99 ROA CAR ROA CAR ROA CAR Comm. banks 1.17 12.2 0.38 4.3 -22.6 -24.6 State 0.82 13.9 0.34 2.4 -24.9 -28.4 Private forex 1.13 10.3 -0.47 5.3 -29.2 -18.8 Private nonforex 0.31 9.7 0.97 15.9 -0.35 10.4 Joint 2.49 18 1.54 4.8 -9.88 -7.7 Foreign 4.48 13.8 5.18 12.8 -0.77 12.9 Source: BI [20] Like the ownership of poorly regulated banks by non-financial companies or political connections between bank managers and borrowing firm management. [21] Data for 1995. For end-1996 estimates were about 13%.
- 86. 88 Marek D¹browski (ed.) Figure 5-14. Indonesia: maturity structure of banks 120 100 60 40 20 *<3m 3m<*<6m 6m<*<1y had well under half of their debt hedged. Borrowing US dol-lars was part of life in Indonesia – "it was like going to McDonald's" [22]. Everyone assumed that the money would always be available, and took advantage of this situation. After 30 years of steady economic growth, the corporate sector didn't seem to fear economic downturn. Some of the loans were used to finance speculative investments in such areas as equity purchases and real estate. Property loans grew at an annual rate of more than 60% during 1992–1995 (compared to 20–25% percent rate of growth for total credit) and in April 1997 accounted for 19.6% of outstanding bank credits. To restrain the growing 1y<*<3y 3y<* deposits exposure of the banking system to this sector BI restricted in July 1997 commercial banks from extending new loans for land purchases and property development (except for low-cost housing). Total credit growth also continued, in spite of consecutive statutory reserve requirements increases (from 2% to 3% in January 1996 and an announcement of a rise to 5% in April 1997) and other attempts by BI to contain it. The growing uncertainty about the future development on the exchange rate market was reflected by an increase in the volume of forward and swap rupiah transactions. Their average daily turnover rose from around 4.5 bln USD in early 1997 to about 6.2 bln USD in June/ July 1997. The increase in trading illustrates increased hedging activities among externally indebted domestic companies. However, nobody expected that a sharp downturn and such a severe crisis would erupt. 5.2.3. Crisis Development Indonesia's troubles began on July 2, 1997 when the Thai baht peg to the USD collapsed. Immediately, market confi-dence in Southeast Asian economies was reassessed. Unex-pected devaluation of the baht meant that any country with similar economic and export structure and comparable fun-damentals is likely to give up its exchange rate policy under similar circumstances. Such an event is called a "wake up call", or a focal point for coordinating market expectations - a "sunspot". A prolonged period of downward pressure started. The yield curve inverted dramatically. On July 11, the Philippines gave up supporting its peso, while the Indonesian authorities CASE Reports No. 39 100 80 60 40 20 0 [22] After The Wall Street Journal 31 XII 1997 quoting one Singapore based Indonesia-investing fund manager. credit 0 80 Source: BI, IMF Figure 5-15. Indonesia: bank credit growth (% p.a.) -20 1993 1994 1995 1996 1997 total property sector Source: IMF
- 87. Figure 5-16. Indonesia: rupiah exchange rate (IDR/USD) 18000 16000 14000 12000 10000 8000 6000 4000 2000 0 97-03-17 97-04-17 97-05-17 97-06-17 97-07-17 97-08-17 97-09-17 97-10-17 97-11-17 97-12-17 98-01-17 98-02-17 98-03-17 98-04-17 98-05-17 98-06-17 98-07-17 98-08-17 98-09-17 98-10-17 98-11-17 98-12-17 99-01-17 99-02-17 99-03-17 Source: Bloomberg Figure 5-17. Indonesia: Money market rate 90 80 70 60 50 40 30 20 10 0 1996M1 1996M3 1996M5 1996M7 1996M9 1996M11 1997M1 1997M3 1997M5 1997M7 1997M9 1997M11 1998M1 1998M3 1998M5 1998M7 1998M9 1998M11 1999M1 1999M3 1999M5 1999M7 1999M9 1999M11 Source: IFS [23] To what extent the rupiah collapse has been caused by speculators or by domestic investors suddenly starting to hedge against exchange risk is a subject of great debate. Indonesian authorities (Minister of Justice in August) claimed that the speculators were guilty and their activities could be interpreted as subversive criminal actions (there is a death penalty for subversion), while hedge-fund managers and other participants suggest that this was not the case. Major funds were fully invested in the rupiah and they even supposedly bet on the rupiah rebound at some moment. But it was the case that many unhedged domestic companies decided to insure in forward market against the rupiah decline. Of course, speculators joined the mar-ket 89 The Episodes of Currency Crises in Latin... widened the trading bands from 8% to 12% (and inter-vened when the rate moved outside the band). On July 14, Malaysia gave up its currency peg. It took one more month and a 15% depreciation until Bank Indonesia floated the rupiah and doubled short-term interest rates to over 25% to support its value [23]. On August 29, BI introduced restrictions (up to 5 mln USD per customer) on nonresi-dents' trading in forward currency contract (i.e. supposed speculation). Despite this, capital outflows continued and by October 8, the rupiah/dollar exchange rate had already depreciated cumulatively by 46%. By that day, matters went so badly with domestic financial and corporate sector when depreciation seemed to be inevitable. CASE Reports No. 39
- 88. 90 Marek D¹browski (ed.) Figure 5-18. Indonesia: Inflation annualized 90 80 70 60 50 40 30 20 10 0 1996M1 1996M5 1996M9 1997M1 1997M5 1997M9 that Indonesia's government decided to request IMF assis-tance. The financial and the corporate sector was confronted with an increase in the rupiah value of their foreign indebt-edness. Most private companies were able for some time to cover their foreign exchange losses but they were dramati-cally running out of cash, unable to refinance their short term debt and watching it rapidly expanding (banks and for-eign creditors refused to rollover the existing short term debt). On the other hand, banks were unable to crack down on their debtors because of a weak and inefficient bank-ruptcy law. This, together with tight liquidity and high inter-est rates, gradually pushed many banks and companies into technical bankruptcy. At end-October, rating agencies downgraded the ratings of 10 big Indonesian banks from neutral to negative, further limiting their borrowing abilities. The already poor confidence in the national banking system brought about a gradual build-up of runs on some of the pri-vate banks, reflecting a "flight to quality", as depositor per-ceived state banks to be safe and were moving deposits from presumably troubled private banks. As a result of a cri-sis, the stock market index immediately lost 30%, and then a further 10%, before November 1997 because investors lost confidence. The real estate market collapsed as well. Office, residential and retail property rent and prices (usual-ly quoted in USD) fell from 30% to 80% between June 1997 and June 1998. On October 31, 1997, the IMF unveiled the 23 bln USD aid package for Indonesia and on November 5 approved a 10 bln USD standby loan facility. Apart from a request for 1998M1 structural reforms and tight monetary and prudent fiscal policy, the package included the requirement for the closure of the 16 most insolvent (bankrupt) banks. Authorities, however, failed to extend appropriate deposit guarantees and a panic erupted among depositors running the whole system [24]. A massive and sudden withdrawal of deposits started, a large number of banks failed to meet their obliga-tions and had to resort to central bank liquidity support. One reason for a sudden drop in overall confidence was that people saw the end of the regime approaching quickly. They doubted the political capacity of the government to fulfil its commitments to the IMF. In the beginning of Decem-ber 1997, Soeharto was ordered to retire to bed and disap-peared from public life for about a month. The implementa-tion of IMF packages was already delayed or off-track. The rupiah collapsed badly to almost 6000 IDR/USD. Soeharto reemerged in public on January 6, 1998, only to unveil the 1998/1999-draft budget that had virtually nothing to do with the reforms agreed upon with the IMF. At that moment, confidence in the Indonesian government was lost com-pletely and on black Thursday, January 8th , the rupiah plum-meted to 10000 IDR/USD – and later even to 14000. The market panic across the country in anticipation of food shortages and overall social unrest and violence started. Food prices skyrocketed indeed and through 1998 increased by 100%, compared to 70% of total CPI increase. Financial panic continued, the surge in liquidity provided by BI to tumbling banks (about 7% of GDP before end-Janu-ary) far exceeded the real liquidity needs of the economy and contributed to sharp rise in inflation and put further [24] Authorities guaranteed only deposits up to about 5000 USD. The guarantees covered 90% of depositors but not even 20% of total deposits. However, the lack of confidence in banking system was so great that there were hardly any awareness of any deposit guarantees (or belief in such guar-antees) CASE Reports No. 39 – depositors with deposits less than 5000USD were also running on banks. 1998M5 1998M9 1999M1 1999M5 1999M9 Source: IFS
- 89. 91 The Episodes of Currency Crises in Latin... 250 200 150 100 50 downward pressure on the rupiah – the inflation-devalua-tion spiral began. In order to stop the bank runs, on January 26, 1998, the government announced a blanket guarantee for all deposits as well as the establishment of a banking sec-tor restructuring institution. Such a massive depreciation had a devastating effect on the balance sheets of banks and companies. By December/January, many of them already quietly stopped paying back loans. The overwhelming majority of banks became paralyzed with an average of 50–70% share of non-performing loans. The debt moratorium on corpo-rate debt payments announced on January 27, 1998 was the official confirmation of this and was met with mixed reception, but also with relief that any actions were taken at all. In the meantime, on January 15, the second agree-ment with the IMF was concluded – previous reform claims were reiterated but the policy somehow eased as the seriousness of the crisis has been realized. After these measures, the rupiah stabilized and moved within 8000–10000 IDR/USD band from end-January to the beginning of May. CASE Reports No. 39 Tight financing conditions, heavy burden of debts, cash-shortages, negative wealth shock connected with rapid depreciation of asset prices, political instability and uncer-tainty about the future of the regime, social and ethnic ten-sion, accelerating inflation and general uncertainty about the prospects for the economy were the main reasons for the sharp domestic demand contraction. Individuals, expecting tough times, postponed consumption and switched to sav-ings. The corporate sector halted or delayed investment plans. Consumption fell by 9% and investment by 45%. A decline in demand and damage to production and distribu-tion facilities caused by social unrest contributed to a sharp contraction of economic activity, the most severe in con-struction (-37% from 1997 to 1998) and financial, rental and corporate services (-58%). The total output declined in 1998 by 14%. Both exports and imports fell but import con-traction was much more severe (-49.4%) and was a reason for achieving a current account surplus of 3.8% of GDP in 1998. The surplus (net external demand) however was by no means sufficient to offset the fall in domestic demand. The surplus in the capital account was caused by a large Figure 5-19. Indonesia: Consumer prices 0 1996M1 1996M5 1996M9 1997M1 1997M5 1997M9 1998M1 1998M5 1998M9 1999M1 1999M5 1999M9 Source: IFS Table 5-3. GDP growth decline components in 1998 Component Growth % in GDP decline Domestic demand -17.6 134 Private -2.9 13 Consumption Public -14.4 8 Investment -40.9 96 Stock changes - 17 External demand - -34 Export 10.6 -21 Import -5.5 -13 Total GDP -13.7 100 Source: BI and Cental Bureau of Statistics
- 90. 92 Marek D¹browski (ed.) inflow of official aid, while the outflow of private capital was not reversed. Inflation escalated to the level of 80% in 1998 in response to panic food buying, interrupted production, social violence and increased prices of import commodities. Unemployment rose from 5% to 28% at end-1998. On March 10 1998, Soeharto was reelected to a 7th term in office. On May 4, the government announced sharp price increase of gasoline and other utilities. Widespread protests erupted, among them most importantly student-led anti-regime demonstrations calling for the President's resignation. The army cracked down on protesters. Embassies and for-eign companies evacuated non-essential staff. On May 19, stu-dents started parliamentary compound occupation. The stu-dent demonstrations seeking political reforms were accom-panied by rioting, widespread looting, destruction, crime, as well as religious and ethnic conflicts. Anti-Chinese rioting directed mainly at shopkeepers in small town resulted in complete disruption of supply distribution channels and short-age of basic products. The general erosion of social order went out of control. Over 1000 dead were reported in the May riots. The hard-won relative stability of the rupiah was immediately lost, runs on banks and massive deposit with-drawals started again, the currency crisis renewed and the rupiah plunged to over 16000 IDR/USD. It took five months to bring it back under 10000 IDR/USD. On May 21, urged by his affiliates, Soeharto resigned. Changes in key positions in the administration con-tributed to delays in the implementation of economic reforms. With reference to that and to harsh economic cir-cumstances, the IMF rearranged its agreements with Indonesia towards easier conditionality. The situation start-ed to stabilize. Food security has been gradually restored through emergency import and increased food subsidies. Monetary stability gradually returned around October 1998, inflationary pressure eased and the rupiah stabilized around 9000 IDR/USD. Price levels also finally stabilized and the beginning of 1999 saw some deflation. The sluggish process of financial system and corporate debt restructuring started. 5.3. Response to the Crisis 5.3.1. Introduction When the crisis unfolded, market confidence in Indonesia disappeared and domestic conditions started to deteriorate quickly as economists and market participants came to the conclusion that the rescheduling of Indonesia's debt and the rescue of the financial system would require the backing of western governments and international institutions: "No one would want to buy Indonesian debt if it was just Indonesian debt" [25]. The Indonesian authorities also noticed that with-out additional backing and emergency loans the country would soon default on its debt. So, the IMF assistance was requested in October 1997 and the first Letter of Intent, aid package of 23 bln USD and a 10 bln Standby arrangement was announced in early November. The IMF, not expecting the seriousness of the crisis, insisted on tough monetary and fiscal policies as well as economic reforms, including the closure of some bankrupt banks. These measures, however, only aggravated the bank-run problem. On the other hand, there was no political will to implement the reforms. As result, confidence further col-lapsed. Indonesia and the IMF signed the second agreement in mid-January 1998. The hardships with implementation of agreed reforms and civil unrest again threw the program off-track. The IMF has been several times threatening to postpone or suspend the aid. From the third letter of intent in April, and an agreement with the new government after the fall of Soe-harto, the IMF acknowledged the severe economic conditions and approved the implementation of less stringent measures. In July 1998, the Standby agreement was replaced with an Extended Fund Facility program. As of June 2000, the international aid commitment to Indonesia amounted to 50 bln USD: 12 bln from the IMF, 10 bln from multilateral financial institutions (like World Bank or Asian Development Bank) and 15 bln from bilateral agreements and programs (like Japanese Miyazawa plan). About half of this package, i.e. 22 bln USD, has already been disbursed. 5.3.2. Monetary Policy Response The government responded to the first run on the rupi-ah in July/August 1997 with a drastic credit contraction. BI stopped repurchasing central banks certificates, decreasing the supply of local currency in an effort to discourage mar-ket participants from exchanging the rupiah into hard cur-rency, interest rates rose from around 15% to 30%. Although official statistics do not show a sharp decline in for-eign reserves, part of the stock was probably tied down in forward contracts, so the usable reserves were actually lower. From September 1997, the reserves started to shrink much faster. Facing a deteriorating domestic situation, from September BI gradually cut yields on its commercial papers. Monetary policy was being carried out in an environ-ment of high debt-equity ratios and overall financial system distress, which made a prudent policy of high interest rates almost impossible to implement. Eventually, the Indone-sian authorities had to resort to IMF help. One of the main goals that the IMF program pursued was to restore market confidence. Based on the presumption that the Indonesian economy was suffering from structural weaknesses than CASE Reports No. 39 [25] After The Financial Times 30 I 1998 quoting a regional economist of one major western banks branch in Jakarta.
- 91. 93 The Episodes of Currency Crises in Latin... Figure 5-20. Indonesia: money and quasi-money (trin IDR) 700 600 500 400 300 200 100 0 1996M1 1996M3 1996M5 1996M7 1996M9 1996M11 1997M1 1997M3 1997M5 1997M7 1997M9 1997M11 1998M1 1998M3 1998M5 1998M7 1998M9 1998M11 1999M1 1999M3 1999M5 1999M7 1999M9 1999M11 traditional, temporary macroeconomic imbalances and might require some real adjustment, there was no specif-ic exchange rate or interest rate target. Instead, the authorities decided to stick to nominal base money targets as a nominal anchor consistent with the free flow exchange rate regime. During the first week of Indonesia's program (November 1997), the authorities engaged themselves in unsterilized intervention and allowed for short term interest rates hike again – the rupiah appreci-ated and regained some losses. However, within less than a week – and contrary to the agreement with the IMF – BI cut the interest rates to their initial level and started to increase liquidity. The result was a near collapse of the banking system during November 1997 - January 1998. After runs on banks started, the authorities completely abandoned tight policies agreed with the IMF and injected massive liquidity into the banking sector as people were withdrawing their deposits. There were only limited efforts to sterilize this increase in net domestic assets by open market operations and foreign exchange interven-tions – base money grew by 126% in six months instead of 10% as was intended. Cash-in-circulation also increased as a result of panic withdrawals. The authorities and the IMF grossly underestimated the negative sentiment and a drop in confidence of market participants. The BI lost con-trol over monetary aggregates. 200 150 100 50 CASE Reports No. 39 Money Quasi-Money, (trin IRD) Source: IFS Figure 5-21. Indonesia: liquidity support (trln IRD) 0 1996M12 1997M2 1997M4 1997M6 1997M8 1997M10 1997M12 1998M2 1998M4 1998M6 1998M8 1998M10 1998M12 1999M2 1999M4 1999M6 1999M8 1999M10 1999M12 2000M2 Source: BI, IMF
- 92. 94 Marek D¹browski (ed.) 120 100 80 60 40 20 There were two major waves of bank runs – in Novem-ber 1997/January1998 and in May 1998. In both cases, liq-uidity support was extended, base money rapidly increased, as did the currency in circulation and broad money, thus fueling inflation. The total liquidity support surged from 9 trillion rupiah at end-1996 to 62 trillion rupiah at end- December 1997 (equivalence of about 7% of GDP). By June 1998, the figure stood at 168 trillion rupiah. The open mar-ket operation and selling of hard currency absorbed only 30 trillion rupiah. In the meantime (February 10, 1998), reflecting a des-perate attempt to restore market confidence, at the initia-tive of President Suharto, the Finance Minister announced that Indonesia was considering establishing a currency board by fixing the rupiah at around 5,500 IDR/USD. The idea was that the currency board would discipline the central bank with respect to reckless money supply, immediately restor-ing its credibility, quickly breaking the vicious circle of infla-tion and depreciation. It was technically feasible to imple-ment, as international reserves far exceeded reserve money. Indonesian authorities strongly insisted at this idea, but the overall reception was negative. The IMF ultimately rejected it as too dangerous for Indonesia for the following reasons: First, if the currency board is even slightly less than fully credible (what seemed to be the case judging from unstable political regime and violent social tensions), it automatically leads to a contraction of the economy and excessively high interest rates. Second, an unsustainable currency board at the appreciated exchange rate (5500 IDR/USD, while on the day of the proposal the rate stood at 7287 – just down from 14000 and soon up to 10000) would prompt massive capi-tal outflow and eventual system breakdown. Third, a cur-rency board prevents the central bank from acting as a lender of last resort – BI would have to revoke its deposit guarantees, which would trigger another panic (honoring these deposits was technically unfeasible). Despite the efforts to implement tight monetary policy and prevailing high nominal interest rates, the actual stance of monetary policy has been loose with (ex post) real inter-est rates distinctly negative which probably reflected the expectations of a severe economic downturn. With a change of the political regime in May 1998, the appointment of a new government and the EFF agreement with the IMF in July 1998, the authorities made an effort to strengthen the credibility of monetary policy. This time, base money was to be monitored closely. The monetary policy through base money restraint was directed toward maintaining price stability, while the exchange rate was left to market mechanisms. BI also made an effort to strengthen the credibility and transparency of policy by making period-ic announcements of its targets. In achieving the quantitative target, BI resorted to open market interventions – on July 29, 1998 the central bank certificates auctions system was improved and changed: emphasis was shifted from interest rate to quantity target. To prevent a further expansion of liq-uidity, a high penalty on the discount window facility and commercial bank's negative balance with Bank Indonesia has been imposed, together with ceiling on deposit rates and interbank rate for banks guaranteed by the government. The expansion of liquidity ceased. The government took over from central bank most of the outstanding banks' liq-uidity support liabilities in exchange for promissory notes worth 144 trillion rupiah. To control the monetary expan-sion originating from increased government expenditures, BI conducted sterilization in the foreign exchange market, helping the same the rupiah to strengthen. After its July agreement with the IMF and the introduction of new auc-tion system, BI tried to stick firmly to its policy. Base money CASE Reports No. 39 Figure 5-22. Indonesia: reserve money and currency in circulation 0 1996M1 1996M3 1996M5 1996M7 1996M9 1996M11 1997M1 1997M3 1997M5 1997M7 1997M9 1997M11 1998M1 1998M3 1998M5 1998M7 1998M9 1998M11 1999M1 1999M3 1999M5 1999M7 1999M9 1999M11 Reserve Money of which: Currency Outside DMBs, (trin IRD) Source: IFS
- 93. Figure 5-23. Indonesia: rice prices (IDR/kg) 3500 3000 2500 2000 1500 1000 500 0 1995Q1 1995Q2 [26] For example: if there is no budget deficit there is also no temptation to monetize it in a crisis period; on the other hand, increase in public sav-ing contributes to current account improvement (in 1997 current account deficit has been regarded as a problem). 95 The Episodes of Currency Crises in Latin... started to move within designated bands, monetary condi-tions stabilized and BI regained much control over the finan-cial market. The interest rate decline started since Septem-ber/ October 1998 together with monetary stabilization and inflation decrease (CPI rise halted and then turned into slight deflation in 1999). 5.3.3. Fiscal Policy Response The initial November 1997 IMF plan of fiscal tightening was expected to restore confidence [26], demonstrate the authorities' eagerness for reforms and make room for pos-sible bank restructuring costs. Already in September 1997, about 80 infrastructure projects (including 13 power plants and 36 toll roads) were suspended. According to the plan, wide ranging cuts in public spending and the postponement of about 35 bln USD in infrastructure projects were to be implemented in order to reduce the current account deficit and generally improve the soundness of the economy. The government budget surplus was planned to amount to 1% of GDP. In the sphere of structural reforms, the dismantling of state monopolies, trade liberalization and other similar measures were envisaged. However, the authorities ignored their most important commitments which was revealed in the 1998/1999-budget proposal in January 1998. The constitutional validity of the IMF-supported stabilization program was also questioned on the grounds that it goes against "family values". The dete-riorating situation forced the government to seek another agreement with the IMF on January 15, 1998. The macro-economic assumption of the second program was revised downward: 0% GDP growth in 1998, 20% inflation and 5000 IDR/USD rate. the fiscal stance was eased to meet a 1% deficit. Calls for structural reforms were reiterated. Budgetary support, tax and credit privileges to the new air-plane and national car projects (owed by Soeharto family) were to be canceled, cartels in cement, paper and plywood dissolved, domestic agriculture deregulated, import and dis-tribution restrictions lifted, fuel subsidies gradually removed, fiscal transparency improved, autonomy for mon-etary policy granted. Nevertheless, the authorities were still very reluctant to fulfill these demands, as many of them were directly targeted at businesses from which govern-ment officials' relatives and friends profited. Later in 1998, in accordance with the changing political and (worsening) economic situation, the program was revised and included not only accommodation of the shock, but also some additional fiscal stimulus. Current account (that actually quickly turned into surplus by itself) and confi-dence issues ceased to be main problem. The severity of the recession was not taken adequately into consideration while designing previous programs. In accordance with the agreement with the IMF the government raised expendi-tures that were associated with the social safety net as well as subsidies of oil-based fuel, electricity, medicine and food-stuff. Subsidies increased dramatically from 0.3% of GDP in 1996/97 to 3.1% of GDP in 1997/98 and 4.4% of GDP in 1998/99 budget (however less than planned 6.6%). The state budget was planned and estimated to run into deficit CASE Reports No. 39 1995Q3 1995Q4 1996Q1 1996Q2 1996Q3 1996Q4 1997Q1 1997Q2 1997Q3 1997Q4 1998Q1 1998Q2 1998Q3 1998Q4 Source: IMF
- 94. 96 Marek D¹browski (ed.) Figure 5-24. Indonesia: government subsidies (%GDP) 5 4 3 2 1 0 92/93 93/94 94/95 95/96 96/97 97/98 98/99 Source: IMF of about 8.5% of GDP in 1998/99. In the end, however, the government failed to provide a sufficient boost to the econ-omy, the realized deficit reached only 2.2% due to lower government expenditures attributed to exchange rate appreciation and technical constraints in general [27]. As for deficit financing, foreign borrowing financed 99% of it, while remaining 1% was financed domestically. 5.3.4. Banking System and Debt Restructuring The authorities have begun to restructure the banking system through a mixture of bank closures, mergers and takeovers. After the closure of 16 banks in November 1997 and the following bank runs, BI guaranteed eventually in Jan-uary 1998 all deposits at domestic banks. At the same time the establishment of Indonesia Bank Restructuring Agency (IBRA) was announced. The task of IBRA was to assume control over troubled private banks, review them for liqui-dation or recapitalization and manage non-performing loans. There were initial problems with the operation of that body, as bank managers failed to change their behavior in accor-dance with IBRA's recommendation. By April 1998, it became apparent that forceful ("hard") intervention [28] was necessary. Efforts were made to recover the liquidity credit extended to banks by the central bank, so IBRA focused and finally took over (effectively nationalized) seven private banks responsible for 75% of all liquidity support and accounting for 16% of total banking system liabilities. Another seven very small and completely insolvent banks were closed. The operations of IBRA were subject to rising uncertainty among a public unaccustomed to the implica-tions of bank takeovers. Several "IBRA banks" were being run on for some time. In September and December 1998, the authorities announced a comprehensive plan of restruc-turing of the banking system. Banks were categorized depending on their capital adequacy ratios (CAR). Banks with CAR above 4% would be allowed to continue opera-tion. Banks with CAR below –25% were given one month to recapitalize, failing which they would be merged or closed. The rest of the banks were to submit reliable busi-ness plan, which would be assessed by independent experts. Banks were required to meet capital adequacy ratio of 4%, 8% and 10% by the end of 1998, 1999 and 2000 respec-tively. These requirements were strictly executed and dur-ing the financial year 1998/99 the government closed 48 banks. State banks were jointly recapitalized and four of them merged into new state bank (Bank Mandiri) which became the largest bank in the system with about 30% of all deposits. After mergers and closings, the number of banks dropped from 238 to 157. As a result there was a dramatic change in the ownership structure of the banking system – the government's stake rose from 40% to 70%. Despite these efforts, the state of the banking system, as of March 2000, still leaves much to be desired with non-performing loans ratio of 32%. The recapitalization program has been financed by the issuance of bonds worth over 50% of GDP [29]. The cost of the bailout will be a substantial burden for public finances – the public debt amounts already to over 90% of GDP. IBRA liquidity credits were to be converted into equity or subordinate debt. Some of the capital is planned to be regained by consequent privatization. Restrictions on for-eign investors to own banks in Indonesia have been accord-ingly removed. However, the prospects of asset recovery from bankrupt and restructured banks are very dim – the market estimate of the IBRA portfolio is 20% of its book value. The second urgent problem was corporate and inter-bank external debt restructuring. Foreign banks were very reluctant to rollover the debt of falling companies. On Janu-ary 27 1998, the government had to announce a corporate debt payment moratorium. Talks with a steering committee of private bank creditors concerning the restructuring of interbank and corporate debt began in February 1998 and were concluded on June 4, 1998, in Frankfurt. Agreement on interbank debt involved an offer to exchange the debt maturing by end-March 1999 with the new loans. They [27] Similarly the 1999/2000 fiscal deficit was only 1.5% of GDP and fail to reach 5% planned in the budget. [28] Suspension of shareholders rights, assumption of ownership by IBRA and management replacement. [29] This cost is significantly higher than the costs other crisis countries had to incur. The cost of the banking sector restructuring in percent of GDP CASE Reports No. 39 were: 17% in Korea (1997- ), 29% in Thailand (1997-.), 29% in Chile (1981-87), 19% in Mexico (1994–99).
- 95. [30] This was similar to Mexican corporate debt restructuring framework called "Ficorca". [31] The London approach to debt resolution is a voluntary, non-binding framework in which creditors agree to keep credit facilities in place, seek [32] In USD terms Indonesian shares were 12 times cheaper (!) in September 1998 than in June 1997 – this is what people call "the fire sale FDI". 97 The Episodes of Currency Crises in Latin... were backed by a full dollar guarantee of Bank Indonesia and of maturities from one year (not more than 15% of the new loans) to four years (at least 10% of the new loans) at an interest around 300 basis points above Libor. Foreign banks committed to maintain trade financing as far as possi-ble. To eliminate a crunch in international trade payments and kick-start import/export activities, Bank Indonesia set-tled the trade arrears of commercial banks amounting to more than 1 bln USD. Unlike with the banking system, the Indonesian government was reluctant to extend direct sup-port to the private sector, but preferred instead to provide a government-supported umbrella for restructuring private sector debt with some tax concessions and preferential financing rates, but without any formal guarantees [30]. This was reflected in the scheme of corporate debt restructuring agreed in Frankfurt. It provided a framework for voluntary restructuring of external debt through direct negotiations between debtors and creditors with a support and mediation of a new governmental body called the Indonesian Debt Restructuring Agency (INDRA) established in August 1998. Its task was to provide exchange rate guar-antees under condition that the agreement met certain con-ditions (a minimum eight years maturity and three years grace period). INDRA would not guarantee payment, but only the exchange rate and would supply foreign exchange using the best 20-day average rate before June 1999, with a reset option if the rupiah appreciate more. The INDRA scheme was complemented by so called "Jakarta Initiative", i.e. the set of guidelines for debt restructuring workout based on a London approach [31]. In the meantime the new, tough bankruptcy law took effect in August 1998, and the reluctant companies had an additional incentive to join INDRA-scheme. The market reactions and the experience with implementation are not too satisfactory. By February 1999, some 120 companies with total debt of 18 bln USD were registered to Jakarta Initiative. By July 2000, only 5 bln USD has been rescheduled, i.e. not much more than 1% of total eligible debt. Numerous institutional and political obstacles and the failure of the legal system to pose a cred-ible threat to the debtors obstruct the process. The corpo-rate debt resolving continues, but its slow pace undermines the economic recovery and market confidence in Indonesia. In December 2000, the IMF warned Indonesia of the possi-ble consequences and urged the authorities to deal with the problem. The stock market index hit an all-time low in September 1998 but as soon as monetary and political conditions stabi-lized the market rebounded quickly as foreign investors took advantage of unbelievably cheap equity prices [32]. The bourse reached pre-crisis levels in early 2000 but soon after that, in the first month of 2000, lost 40% due to a pro-longed crisis and higher US interest rates. The property market remains weak with a 35% vacancy rate in office real estate. 5.3.5. Prospects for the Future In early 1999, new electoral laws were adopted and in June the country's first free and honest elections were held. In contrast with the past, there was more than one candidate for presidential post, which finally was won by Abdurrahman Wahid, an open-minded and relatively liber-al leader of the major Muslim organization. GDP grew a slight 0.1% in 1999 and is estimated to grow between 3% and 4% in 2000 and 4–5% in 2001. The rupiah gradually depreciated from 6900 IDR in October 1999, right after the new elections, to around 9500–10000 IDR/USD in December 2000. Indonesia is still heavily dependent on international support. Failure to meet the IMF's and for-eign creditors' expectations can still have serious conse-quences but pressure from outside strengthens the pro-reform faction in the government. So, ironically, the authorities (can) take advantage of the crisis to push through some important reforms. 5.4. Conclusions Indonesia is the most hard hit country among the East Asian crisis' victims. GDP fell 14% in 1998 and only after two years does it show any sign of recovery – it is going to take a long time until GDP growth returns to pre-crisis lev-els of 7%. As a result of the crisis, the political regime col-lapsed, social and ethnic tensions erupted, the country's integrity has been threatened and a poverty problem emerged. This paper tried to answer why this was the case. The answer is that Indonesia had probably the worst economic fundamentals of all the East Asian countries. The economy was ridden by corruption and monopolized. The weak financial system engaged itself in reckless credit expansion, dangerously risky investments or even quasi-criminal activi-ties. Overoptimistic corporate sector accustomed to the abundance of capital ceased completely to insure against economic risk. Short-term external debt mounted, with a presumption that it would never have to be paid back and out-of-court solution and work together in good will. CASE Reports No. 39
- 96. 98 Marek D¹browski (ed.) that the exchange rate would be pegged forever. The quali-ty and efficiency of investment decreased. The case of Indonesia speaks in favor of a view that there is a relationship between fundamentals and (the severity of) a crisis. The course of Indonesia's development was defi-nitely unsustainable – at some point such a policy would have to fail, so investors withdrew before the moment came, as in the first generation theoretical crisis models. On the other hand, had the financial panic not erupted, first in some other country (Thailand), Indonesia could fur-ther develop uninterruptedly – would need some econom-ic reform but, still, there was nothing about the economy that called for immediate collapse. What eventually brought Indonesia down was self-fulfilling panic among international creditors that drove them to cancel loans just because other investors were doing the same. Such a situation is well described by the second-generation-self-fulfilling-crisis the-oretical models. The case of Indonesia (and more generally of the Asian financial crises) is neither unique, nor can it be fully explained by any of the two main theoretical views. Rather it can be said that for some range of fundamentals, i.e. when the state of the economy worsens but not up to the point when the collapse is inevitable – the country becomes vul-nerable to a crisis caused by self-fulfilling panic. The Asian crisis calls for the reassessment of the notion of a "fundamental". It cannot be limited to easily measurable, "classic" macroeconomic variables such as fiscal deficit, mon-etary expansion, inadequate reserves or "political" factors like unemployment or recession. This notion should be broadened by such vogue ideas as 'credibility', "moral haz-ard", or strictly microeconomic factors, i.e. structure of the corporate sector, strength of a financial system, etc. In the case of Indonesia, these fundamentals were in a very bad state. CASE Reports No. 39
- 97. 99 The Episodes of Currency Crises in Latin... Appendix 1: The chronology of the Indonesian crisis 1997 May: Thai currency comes under speculative pressure. July: Thai, Malaysian, Philippine, and Indonesian curren-cies all depreciate. August 14: Indonesia abolishes its system of a man-aged exchange rate. The rupiah starts to depreciate September 16: 15 government "mega-projects" are postponed. October 8: Indonesia says it will ask the IMF for financial assistance. October 31: Indonesia's IMF package is unveiled. It pro-vides for more than 23 USD bln in aid. November 1: Sixteen banks are closed as first step in IMF package, what causes panic and bank runs among depositors. November 5: IMF approves a US$10 billion loan for Indonesia as part of the massive international package. December 5: Soeharto takes 10 days rest after a 12-day world tour and misses ASEAN summit. December 9-12: Finance Minister fails to negotiate the debt rollover in Washington. 1998 January 6: Indonesia unveils an expansionary 1998/99 budget, contrary to IMF demands of a budget surplus. The rupiah loses half its value over a five-day period. January 9: Ratings agency Standard & Poor downgrades Indonesia's currency to "junk bond" status. mid-January: More or less direct calls start to be made for a change of the regime. January 15: Soeharto signs a new IMF agreements. January till mid-February: Anti-Chinese food riots take place in at least a dozen places throughout Indonesia. mid-January: All but 22 of the 286 companies listed on the Jakarta stock exchange are technically bankrupt. Prop-erty companies are the worst. January 27: Government announces a moratorium on repaying debts and interest, and promises to guarantee all deposits of commercial banks. February: Soeharto proposal of a currency board is announced, criticized and finally turned down. February : Talks with a steering committee of private bank creditors concerning the restructuring of interbank and corporate debt begin March 10: Soeharto is re-elected to a seventh five-year term with Habibie as vice president. May 4: Fuel prices are increased by up to 71 percent. Three days of riots follow. May 9: Soeharto leaves for a week-long visit to Egypt. May 12: The army troops shoot four students at Jakarta protest. May 13-14: Rioting spreads throughout Jakarta. Estimat-ed 1,200 people die in two days. When Soeharto returns from Egypt, he faces a flood of calls to resign. May 21: Soeharto resigns and hands power to Habibie. June 4: Agreement concerning debt restructuring is reached in Frankfurt. June 17: The rupiah again hits 17,000 against the dollar. July 29: The central bank certificates auctions system was improved and changed in attempt to regain control over monetary aggregates. September 24: Paris Club reschedules $4.2 billion of sovereign debt. Annual inflation rises to 82.4 percent in Sep-tember. September 29: Indonesia strengthens bank recapitaliza-tion scheme. October: Monetary stability gradually returns, inflation-ary pressure eases and the rupiah stabilizes around 9000 IDR/USD. November 10: Special session of the Parliament begins to discuss election and political reforms. 1999 March 13: Government closes 38 insolvent banks. June 7: Indonesia holds first democratic election since 1955. August 6: Finance Minister admits there were "irregular-ities" in loan-recovery process. The scandal prompts IMF and World Bank to threaten loan suspension. October 1: Indonesia announces seventh month of deflation, with annual inflation of 1.25 percent. October 20: Wahid has been elected President. [33] This chronology wes much to "CNN Asia Now", October 1999 and "Inside Indonesia", report No. 54, April-June. CASE Reports No. 39
- 98. 100 Marek D¹browski (ed.) References Annual report on exchange rate arrangement, IMF 1996, 1997, 1998, 1999, 2000. Bank Indonesia (BI), Annual report, 1997/98, 1999/99. Corsetti G., Pesenti P., Roubini N. (1998). "What Caused the Asian Currency and Financial Crisis". Banca d'Italia, Temi di discussione 343. Claessens S., Djankov S., Lang L. (1998). "East Asian Corporates: Growth, Financing and Risks over the Last Decade". Mimeo, World Bank. CNN Asia Now, October 1999. Delhaize P. (1999). "Asia in Crisis". John Wiley&Sons. Edison H.J. (2000). "Do Indicators of Financial Crises Work? An Evaluation of an Early Warning System". Board of Governors of the Federal Reserve System IFDP 675. Enoch Ch. (2000). "Intervention in Banks During Banking Crises: the Experience of Indonesia". IMF PDP/00/2 Financial Times, 1994: 24 VI; 1995: 9 VI; 1997: 1-2 XI, 24 XI; 1998: 17-18 I, 30 I, 1 VII. Gould D.M., Kamin S.B. (2000). "The Impact of Mone-tary Policy on Exchange Rates During Financial Crises". Board of Governors of the Federal Reserve System IFDP 675. Inside Indonesia, report no 54, April-June 1998. IMF (1999). "Indonesia – statistical appendix". IMF (1999). "IMF-Supported Programs in Indonesia, Korea and Thailand, a Preliminary Assessment". IMF Occa-sional Paper 178. IMF (2000). "Recovery From the Asian Crisis and the Role of the IMF". International Crisis Group (2000). "Crisis in Indonesia". Mimeo. Milesi-Ferretti G., Razin A. (1996). "Current Account Sustainability, Selected East Asian and Latin America Experi-ences, IMF WP 96/10. Montgomery J. (1997). "The Indonesian Financial Sys-tem: its Contribution to Economic Performance and Key Policy Issues". IMF WP/97/45. Kamin S. (1999). "The Current International Financial Crisis, How Much is New?". Journal of International Money and Finance, vol.18, no.4, 1999. Krugman P. (1994). "The Myth of Asia's Miracle". Foreign Affairs, Nov.-Dec. Krugman P. (1998). "What Happened to Asia". Mimeo. Obstfeld M., K. Rogoff (1995). "The Mirage of Fixed Exchange Rates". Journal of Economic Perspectives 9, 73–96. Sarel M. (1997). "Growth and Productivity in ASEAN Countries". IMF WP 97/97. Sarno L., Taylor M.P. (1999). "Moral Hazard, Asset Price Bubbles, Capital Flows and East Asian Crisis, a First Test". Journal of International Money and Finance, vol.18, no.4, 1999. Stone M.R. (1998). "Corporate Debt Restructuring in East Asia: Some Lessons from International Experience, IMF PPAA 98/13. Transparency International, Annual report 1995, 1996, 1997, 1998, mimeo. Wall Street Journal, 1997: 31 XII, 1998: 9–10 I. World Bank (1998), Responding to the East Asian Crisis. CASE Reports No. 39
- 99. 101 The Episodes of Currency Crises in Latin... Part VI. The South Korean Currency Crisis, 1997–1998 by Monika B³aszkiewicz 6.1. Was Korea Different? 6.1.1. Introduction Financial crises' episodes of the 1990's differ from those of the 1980's in that recently they have occurred in coun-tries on their way to social and economic development. The initial success in implementing reforms, and the evident prosperity for the future, encouraged foreign investors to diversify their portfolios towards emerging markets. Addi-tional capital inflows allows financing economic growth and speeds up the process of integration with the global market. On the other hand, the same capital can cause troubles when it becomes an abrupt and sharp outflow. Why and when does it happen? The group of countries that came under speculative attack in 1997 in Southeast Asia can be roughly summa-rized within the above scenario. Yet the roots underlying the crisis in each country within the same group should be treated as country-specific. Of course, there were some common features among them, but in order to identify main characteristics for an individual economy, a separate analysis is required. This is even truer for South Korea (hereafter referred to as Korea) where strong macroeco-nomic performance until late 1997 did not let most ana-lysts foresee the crisis. Even the eruption of crises in coun-tries from the neighboring region, marked by the July Thai bath devaluation, did not downgrade the assessment given by the international rating agencies to Korea. This paper aims to explore the major factors lying behind the Korean financial crisis. It further looks at the sources of the crisis that took its roots in highly leveraged companies with a weak balance sheet, and a poor func-tioning banking system. It shows that while macroeco-nomic imbalances played a minor role, the close relation-ship among banks, corporations and the government cre-ated CASE Reports No. 39 problems, which resulted in numbers of bankruptcies and in the end led to the sharp and unexpected economic downturn. 6.1.2. Background to the Crisis The currency crisis, which erupted in Korea in the end of 1997 hardly fits the first or second-generation conceptu-al frameworks, where irresponsible government policies and investors' panic play a crucial role. The sudden collapse also cannot be solely attributed to the possible contagion effect across the Asian countries facing similar problems at that time. The 1997 Korean experience is an example, which confirms that financial crises occur not only when macroeconomic but also microeconomic indicators identify vulnerabilities. Although indicators like GDP growth, infla-tion or fiscal balances are important measures of economic soundness, healthy financial and corporate sectors are an essential prerequisite to a successful financial system dereg-ulation as well as liberalization of capital account. In an oversimplification of the classification, the eco-nomic fundamentals can be divided into two broad cate-gories: macro and micro-economic. Looking solely at the former, many failed to predict the 1997 Korean crisis. This is because at the onset of the crisis, macroeconomic funda-mentals in Korea remained relatively sound and did not show many signs of vulnerability. Real GDP growth rate oscillated around 8 percent between 1994 and 1996. At the same time, inflation measured by CPI was under control and averaged at 5.1 percent per annum. The price stabi-lization led to a gradual decline in nominal interest rates. The three-year corporate bond yields, declined from 16.2 percent in 1992 to 11.9 percent in 1996. The consolidated central government position was balanced or even in sur-plus and public debt was less than 10 percent of GDP in the end of 1996 [1]. [1] This number, however, can be misleading due to substantial quasi-fiscal burden in Korea arising from the governmental support to privately own banks. In this case large public expenditures did not appear on the general government balance sheet. Additionally, such a practice posed a moral haz-ard problem.
- 100. 102 Marek D¹browski (ed.) 1994 1995 1996 1997 1998 Real GDP (percent change) 8.3 8.9 6.8 5.0 -5.8 Inflation (CPI) 6.2 4.5 4.9 4.5 7.5 Fiscal balance* 0.1 0.3 0.0 -1.7 -4.2 Gross national savings* 35.5 35.4 33.5 32.5 33.4 Gross domestic investments* 36.5 37.2 37.9 34.2 20.9 Yield on 3-year corporate bonds 12.9 13.8 11.9 13.4 15.1 Current Account balance* -1.0 -1.7 -4.5 -1.7 10.9 *Percent of GDP Source: IMF, IFS; own calculation CASE Reports No. 39 Before financial liberalization, which started in the early 1990's, Korea was targeting monetary aggregates like M2 or MCT [2]. This helped the authorities to maintain financial stability since all other instruments were set to achieve the growth rates of money, inflation and interest as well as exchange rate. Despite the progress made towards the financial sector opening in the 1990's, this policy was con-tinued as the government was convinced about its advan-tages. Two years before the crisis, the annual growth rate of M2 oscillated around 15 and 16 percent. Although this was 3 percent lower than the early 1990's average, domestic credit rose at the very rapid pace during 1994–97 achieving 18.5, 14.1 and 20.1 percent, respectively. Additionally, the broad money aggregate expressed in ratio to foreign reserves was around 6, which was high even in comparison to other crisis economies. In Malaysia this liquidity indicator was equal 4 and in Thailand 4.9 at that time. Nevertheless, other macro-indicators were accept-able. Unemployment rate did not exceed 2.3 percent throughout three pre-crisis years, 1994–1996. Until mid- 1995 investment and saving rates were soaring, averaging approximately at 37 and 35 percent of GDP, respectively. The only exception was the current account deficit, which deteriorated to 4.5 percent of GDP at the end of 1996 and was mostly covered by short-term portfolio invest-ments. There were two factors responsible for the perfor-mance of this indicator. The first related to strong capital inflows during the whole year (especially in the second quarter); the second was a terms of trade shock, repre-senting a 12 percent drop from the previous year (accord-ing to many empirical research shocks to this variable increase the probability of financial crisis). In particular, the unit export price of semi-conductors during 1996 fell by more than 70 percent in the semi-conductor manufac-turing industry. The magnitude of the current account deficit even though large was not tremendous. Many countries suffering from the episodes of financial crises Table 6-1. Macroeconomics fundamentals Figure 6-1. Real effective exchange rate 95 90 85 80 75 70 65 60 55 50 Jan-93 Jul-93 Jan-94 Jul-94 Jan-95 Jul-95 Jan-96 Jul-96 Jan-97 Jul-97 Jan-98 Jul-98 Jan-99 Jul-99 Jan-00 Index Value, 1990=100 Source: Moodys database [2] MCT is composed of M2, certificates of deposits and trust accounts.
- 101. 103 The Episodes of Currency Crises in Latin... experience more severe imbalances. The same is true for the behavior of the won/ dollar exchange rate [3]. Considering tight Korean linkages with Japan (widely fluctuating yen/dollar exchange rate was a key determi-nant of the Korean competitiveness) and appreciation of US dollar vis-a-vis Japanese yen at the beginning of 1995, the won/ dollar exchange rate would be a poor approxi-mation of the total overvaluation of won (annual percent-age changes in 1994 and 1995 were 7.5 and 6.1, respec-tively; in 1996 the real won/ dollar exchange rate was actually depreciating, comparing with its 1995 value). Thus, the real effective exchange rate would be a more adequate way of measurement. But even then, the over-all magnitude of appreciation was not excessively unsus-tainable. From mid-1995, the real effective exchange rate was appreciating comparing with its 1990 value; from May 1996 on it was steadily depreciating. 6.1.3. Signs of Vulnerability Korea's rapid growth during the past decades and its ability to maintain prudent macroeconomic fundamentals masked important structural weaknesses. The successful industrialization process, which transformed the country from one of the poorest nations of the world into one of the most promising, was achieved mainly at the expense Figure 6-2. Net capital inflow 15 10 5 0 -5 -10 -15 -20 -25 -30 1990Q1 1990Q3 1991Q1 1991Q3 1992Q1 1992Q3 1993Q1 1993Q3 1994Q1 1994Q3 1995Q1 1995Q3 1996Q1 1996Q3 1997Q1 1997Q3 1998Q1 1998Q3 1999Q1 bill. USD Source: IMF IFS, own calculation CASE Reports No. 39 of an excessively indebted corporate sector. The 1970's and 1980's strategy, when Korean companies were small and were taking advantage of economies of scale, proved to be wrong in the 1990's. This was mainly due to the overall international environmental change towards tighter linkages among countries that increased an internal and external competition. The external pressure to deregulate and open the financial system in Korea was considerable. The fact that borrowing from abroad was half the price of borrowing domestically (the three-month interest rate on corporate bonds in Korea was as high as 11–12 percent, whereas in the United States it was averaging around 5.5 percent in the second half of the 1990's) com-pounded the foreign exchange exposure from the domes-tic side. Between 1990 and 1996, net capital inflows were equivalent to $69 billion of which $51.8 billion took the form of portfolio investments. Net foreign direct invest-ments were actually negative and equal to $7 billion. Yet, a continuing inflow of short-term foreign capital kept the overall balance of payment in surplus helping to fuel investments and growth. The official foreign reserves accumulated. Neverthe-less, in terms of monthly import (in 1995 and 1996 reserves were enough to cover just three-month imports' obligations) and considering the growing stock of short-term external debt, they were not sufficient to protect [3] Between 1980 and 1989 Korea followed a policy of the 'Managed Basket Peg' to adjust current account. Due to the large investment boom in 1990–91which shifted current account from surplus into deficit, the government decided to adopt the policy called "Market Average Rate System". Within this policy the exchange rate was allowed to fluctuate within a band up to 2.25 percent a day. However, because of still existing capital account restrictions, the system was more similar to fixed but adjustable peg (see OECD, 1998; Black, 1996).
- 102. 104 Marek D¹browski (ed.) CASE Reports No. 39 60 50 40 30 20 10 against the liquidity problems. These numbers were even smaller for gross usable foreign reserves, which are calcu-lated as the difference between official stocks of foreign reserves (minus gold) less overseas branches' deposits of domestic financial institutions. In practice they are hardly available, when the need arises. In Korea the ratio of usable reserves to short-term debt in 1996 and 1997 was equivalent to 0.3 and 0.1, respectively. Due to the above there is no doubt, the Korean crisis had its roots in near depletion of foreign currency reserves. Rushing for additional funds from Korean firms and finan-cial institutions failed to put priority on cash management and other forms of provision against the risk. At the same time, banks and other financial institutions failed to detect the full picture of individual enterprises before offering loans to these companies. Implementation of market principles in an socio-economic environment characterized by over-reg-ulation, concentration of resources around business groups and implicit government intervention in the banking system resulted in the high exposure of Korea to systemic risk. In the end, it eroded the asset side of financial intermediaries when the highly leveraged firms became unable to meet their obligations. A key problem associated with the accelerating stock of external liabilities in Korea was the high proportion of short-term debt in total borrowing. Together with low productivity of investments (the discussion on the invest-ments' efficiency is carried out in the next section) and the low stock of foreign reserves it affected the sustainability of current account deficit, since it mainly depends whether the level of external liabilities is consistent with the country's debt servicing capacity. 6.2. The Role of Cheabols in the Future Development of the Crisis The role of the state in Asia's development in the 1970's and 1980's was substantial. It is often believed that it was exactly the government's intervention and planifica-tion that made the "miracle" possible. However, as the 1997 meltdown showed, this common view is question-able. For many years, investments in Korea were concen-trated around cheabols, the multi-company business groups operating in a range of markets under common supervision and financial control. Although, each compa-ny within a group was legally independent, in reality cheabols were fostered by government policies. It was the Presidential Declaration on Heavy and Chemical Industrialization Policy of January 1973, which encouraged large companies to invest in strategic industries such as semiconductors, shipbuilding, steel etc. [IMF, 1999; OECD, 1998]. The government support, apart from the implicit risk share for preferential industries, was massive. For example, the state-owned banks were pressured to allocate more than half of their loan portfolios to particu-lar sectors. By the same token, strategic industries were provided loans that carried out low interest rates [Nam et al., 1999]. Even after the financial sector's liberalization and privatization in the 1980's, the governmental support for the large firms, affiliates of cheabos, did not vanish. The large economic concentration around business groups that were subject to special regulations in Korea is clearly evident in terms of capital stock invested and the Figure 6-3. Reserves 0 1993 1994 1995 1996 1997 1998 bill. USD Usable Gross Reserves (left scale) Usable Gross Reserves/Import (right scale) Total Reserves - minus Gold (left scale) Total Reserves - minus Gold/Import (right scale) 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 months of import Source: Bank of Korea
- 103. 105 The Episodes of Currency Crises in Latin... Table 6-2. Share of cheabols in mining and manufacturing number of people employed. According to the OECD 1998 Survey on Korea, the thirty largest business groups subject to special regulations accounted for about two-fifths of the capital stock in mining and manufacturing sectors and almost a fifth of employment in 1996. In terms of shipments, their 1994 market share was 39.6 percent (data for 1995 and 1996 was not available). In 1995 the top thirty cheabols' value added accounted for 16 percent of GNP and 41 per-cent of value added in the manufacturing sector [Borenstein and Lee, 1999]. Capital concentration around business groups was definitely important, but not the only problem of weak corporate governance in Korea. The other relates to the corporate concentration of ownership around founding family, relatives and affiliate firms that created corruption. Most of the time, cross-guaranteed debt financing led to the chain reaction, resulting in a collapse of the whole business group. As the OECD report states, in 1996 there were three cheabols with a share of founding family high-er than 20 percent. In 1997 all three were insolvent. Even considering the declining trend in internal ownership (between 1983 and 1997 it fell from 57 to 43 percent) due to capital market development, only about a quarter of the 669 firms affiliated with the top thirty cheabols were listed on the stock market in 1995. Another feature practiced by cheabols was so called "empire-building", the term that relates to diversification of business groups into the broad range of industries. From 1970 to 1996 the number of companies affiliated with thirty largest cheaboles' increased by 18 from aver-age 4 to 22 companies investing in almost 19 industries [OECD, 1998]. Diversification of business into a wide range of different economic activities itself is not a nega-tive practice since it protects against the possible loss at one market by gaining profits at another. Nevertheless, the fact still remains that in Korea, companies felt pro-tected not only by the diversification of their business portfolio, but also because of the governmental interven-tion ensuring takeover rather than bankruptcy. Feeling free of risk, cheabols were engaging themselves in invest-ments based heavily on debt financing. 6.2.1. Debt Financing The total corporate debt measured as the ratio between company's liabilities and its capital employed 900 800 700 600 500 400 300 200 CASE Reports No. 39 1984-89 1991 1992 1993 1994 Top five Shipments 22.5 23,4 23,8 23,0 24,6 Employment 10.02 10,8 10,8 10,4 11,1 Top thirty Shipments 38 38,8 39,7 38,1 39,6 Employment 17.95 17,7 17,5 16,6 17,7 Source: Yoo and Lim (1997) and Fair Trade Commission cited in 1998 OECD Survey on Korea Figure 6-4. Total corporate debt 100 1990 1991 1992 1993 1994 1995 199 6 199 7 1 998 %%, Trillion of Won Source: Bank of Korea
- 104. 106 Marek D¹browski (ed.) was increasing steadily throughout the 1990's, achieving its peak in 1997. There were several remarkable features of this debt, which explicitly contributed to the collapse of many cor-porations and implicitly, through the growing number of non-performing loans, to the bankruptcy of banks and non-banking financial institutions (NBFIs). First, the increasing trend towards indirect financing of the corporate sector in Korea mirrored the relatively undeveloped bond and equity markets (between the first half of 1996 and 1997 the exposure of banks and NBFIs to cheabols almost doubled while direct financing declined by 20 percent (OECD, 1998)). In 1996, in terms of capi-talization, the equity market in Korea was equal to 25.4 percent of GDP. This fell far below that of the developed world (108.7, 67.6, 47.8 for the United States, Japan and G-10 Europe, respectively) and represented a sharp fall from 1990, when the equity market capitalization was equal to 43.6 percent (BIS, 1997 Annual Report). Highly leveraged cheabols became prone to shocks that cause a fall in cash flow (i.e. a terms of trade drop) or an increase in payment obligations (due to an interest rate increase). The situation became worse at the beginning of 1997 when an almost 50 percent slide of market equity value was observed compared to its 1995 high. This affected not only cheabols, but also banks as cheabols were pur-chasing equity for loans granted. The similar trend was observed in terms of the market capitalization of shares of domestic companies (main and parallel markets, excluding investment funds). In 1994, the capitalization was equiva-lent to around 118 percent of GDP. However, by the end of 1997 it dropped together with the stock market decline to only 23 percent of GDP. Secondly, the corporate sector debt in Korea was, to a high degree, concentrated in the thirty largest cheabols. What is more, as the Table 6-3 shows, the exposure of the thirty largest chaebols to non-bank financial institutions in Korea was increasing. Between 1988 and 1992 the share of banks in corpo-rate debt financing dropped by 6.3 percent, but that of NBFIs increased by 8.1 percent. Taking into consideration the minimum supervision imposed on non-banking finan-cial intermediaries, there was no doubt they were eager to make loans to the business sector and individuals. Overall, the high dependence of the Korean corporate sector on debt as opposed to equity finance was clearly evident and extremely high even by international stan-dards. Throughout all the 1980's and 1990's the debt/ equity ratio averaged from around 400 percent to 500 percent plus [4]. On the other hand, the debt/ equity ratio for the United States oscillated around 50–100 percent, for the United Kingdom slightly less at that time. Even Japan, which expanded beyond 350 percent in 1980, in 1994 was down to around 150 percent. Enormous debt/ equity imbalances in Korea had their roots in the system of debt guarantees within cheabols, lax capitalization rules and low effective tax rates on interest income implement-ed to pursue the rapid growth of the economy [IMF, 1999]. 6.2.2. Investments Although it is true that investment rates in Korea were high, the central question remains if they were profitable CASE Reports No. 39 Table 6-3. Share of loans to the 30 largest chaebols in total loans by financial institutions, percent 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 Banks* 28.6 26.3 24.2 20.7 19.0 18.9 17.9 15.6 15.0 13.9 NBFIs - - 32.4 36.6 37.8 38.5 40.5 - - - *Deposit money banks only. Source: The Bank Supervisory Board, The Korea Investors Service, Inc. Quoted from Nam, et al. (1999) Table 6-4. Top five largest cheabols (unit: trillion won, %) Debt/ Equity ratio 1996* 1997 1998 1.Hyundai 440 572.3 449.3 2.Samsung 279 365.5 275.7 3.LG 345 507.8 341.0 4.Daewoo 391 473.6 526.5 5.SK 352 466.2 354.9 *Data for 1996 is from April, otherwise end of the year. Source: Dongchul Cho and Kiseok Hong (1999), Ministry of Finance and Economy [4] The debt/ equity ratios for the top thirty cheabols can be found in Appendix 1 of this paper.
- 105. Figure 6-6. Profitability of the corporate sector 4. 0 3.5 3 . 0 2.5 2 . 0 1.5 1.0 0.5 0 98 96 94 92 90 88 [5] The incremental capital-output ratio (ICOR) is subject to possible faults as its value can be influenced by factors not solely dependent on the [6] Manufacturing ordinary income is calculated as total sales less cost of sales plus selling and general administrative expenses, less net non-oper-ating cost. 107 The Episodes of Currency Crises in Latin... Figure 6-5. ICOR* 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 1988 1989 1990 1991 1992 1993 1994 199 5 1996 * The calculation was done as a five-year moving average to avoid cyclical effects. Source: IMF IFS enough to contract such a high debt? By comparing the cumulative investments with changes in GDP, it is almost certain that some of them were misdirected [5]. Since the late eighties, the value of the capital-output ratio (ICOR) was increasing, indicating the falling quality of fixed capital formation. A slight drop was observed between 1991 and 1993. But after then, the ICOR index was systematically growing. Considering the slow down in GDP growth rates since 1995, there is no doubt there was not enough capacity in the corporate sector in Korea to cope with such high rates of credit. The overheating pressures amplified. Apart from ICOR, other indices like the rate of return on assets, the growth rate of ordinary income as well as the break-even point, showed the low productivity of investment and the growing vulnerability of the Korean corporate sector [6]. The last variable indicates the level efficiency of invested capital. Other factors include structural weaknesses and capital deepening. CASE Reports No. 39 1991 1992 1 993 1 994 1995 1996 % % 86 %% Ordinary income (left scale) Break-even-point (right scale) Return on Assets (left scale) Source: Bank of Korea
- 106. 108 Marek D¹browski (ed.) of cost of production just covered by income (the higher the value of the break-even points, the higher the cost). As the chart shows, from 1995 onwards, all three indi-cators – the break-even point, ordinary income and return on assets – demonstrated the falling profitability of the corporate sector. The high leverage of the corporate sector and its inad-equate governance was an important, if not the major, fac-tor of Korea's ensuing collapse. The question to be answered is why it was sustainable in the 1980's but turned into a collapse in the second half of the 1990's? The simplest answer lies probably in the unfavorable behavior of the terms of trade, which significantly con-strained the cash flow of chaebols, the major exporters in Korea. But there were other factors like the appreciation of the US dollar vis-a-vis the Japanese yen at the beginning of 1995 that weakened the competitive position of firms. Furthermore, the collapse of Hanbo Steel Co. in January 1997, the first big bankruptcy in decades, undermined investors' confidence that Korean firms were 'too big to fail'. This was followed by Moody's decision to lower the long-term rating on three Korean banks, which had a sig-nificant exposure to Hanbo [Park, et. al, 1998]. The com-mon belief in the government willingness to bail out falling companies disappeared together with the declining stock of foreign reserves. Low-productivity of investments was mirrored in the burden of non-performing loans (NPLs) in the banking system. In 1995 NPLs (included loans classified as sub-standard and doubtful) for commercial banks were equal to 5.1 percent of total loans raised. In 1997 it was already 6 percent, according to the Financial Supervisory Service (cited by J. Fleming). Soon after the eruption of the crisis the problem of NPLs in total loans magnified. In 1998 the percentage number increased to 7.4. Although, these fig-ures are high, they maybe even higher since regular reports on NPLs have been available only recently. 6.3. Korean Financial System and its Liberalization Before the liberalization of the Korean financial sec-tor the early 1980's, the government had intervened heavily to pursue its industrial objectives. As the reforms progressed, several commercial banks and non-bank financial institutions were added to the system. Never-theless, the attempts to liberalize were only partially successful, still leaving many regulations in force (i.e. low interest rate ceiling to increase profits and retain earn-ings for selected firms, commercial banks' lending to preferential sectors) [7]. The 1988 plan to deregulate the majority of bank and non-bank's lending rates as well as interest rates on money market instruments was mostly reversed, because of pressures arising from earlier beneficiaries of the pref-erential access to low interest rates credit. On the other hand, the second attempt to implement the plan in 1992 was suppressed by the stock market slump [IMF, 1999]. The next phase of financial system reform took place in 1993 when the first democratically elected civilian gov-ernment came to power. The new government under the President Kim Young Sam was highly committed to finan-cial liberalization. There were two reasons for speeding up the process. One of them was the perspective of join-ing the OECD, the other was the growing ability of pri-vate and already credible firms to borrow funds from abroad. However, most capital flows attracted by firms through the stock market were not free from explicit or implicit quantitative controls, with the exception of trade related short-term financing [Dooley et al. 1999]. During that time, large structural changes led to further rapid growth of non-banking financial institutions. As the Bank of Korea states, the market share of non-banking financial institutions in terms of Korean won deposits between 1980 and 1998 increased from about 29 to 72 percent. The numbers for banking institutions (commercial and specialized banks) were 71 and 28 percent, respectively [OECD, 1998]. The trend in loans and discounts was sim-ilar, increasing for NBFIs and falling for banks. Between 1996 and 1997, the share of funds raised by the business sector from non-banks to the total funds raised increased by 10 percentage points, from 13.9 to 23.9 percent. At the same time borrowing from banks dropped from 14 to 12,9 percent (Bank of Korea). 6.3.1. Non-banking Financial Institutions In Korea, non-banking financial institutions can be roughly classified into five categories according to their business activities. These are: development, savings, investment, insurance, and other institutions (Bank of Korea). The role they played in causing future deteriora-tion in the financial sector balance sheet was significant since they were allowed greater freedom in their man-agement of assets and liabilities. What is more, they were able to charge higher interest rates on their loans as well as apply higher interest rates on their deposits. Regarding [7] Commercial banks in Korea were nationalized in the 1960's and since then the government was influencing the sectoral allocation of credit with CASE Reports No. 39 smaller or greater intensity [IMF; 1999].
- 107. 109 The Episodes of Currency Crises in Latin... Table 6-5. Fund Raising by the Corporate Sector the troubles Korea faced in 1997, the number of new licenses issued to merchant banking corporations was an important factor. In 1993 there were just 6 merchant banks. By 1996 this number increased to 30 as a result of deregulation on financial transactions. In principle, mer-chant banks were supervised by the Ministry of Finance and Economy, but this was minimal as there was no asset classification, capital, or provisioning rules [IMF, 1999]. Besides, most of them were owned by cheabols and were used to finance activities within a group. To attract funds, merchant banks, for example, were offering cash man-agement accounts to their customers (within these accounts, apart from getting checkbooks and credit cards, banks' clients were able to raise loans). Banks were main-ly investing in short-term commercial papers and notes. The contribution of non-banking financial institutions, and merchant banks in particular, in financing investments of corporations was significant. Lax regulations let banks provide loans and guarantees of up to 50 percent of their capital. Additionally, the practice of cross-guarantees was common, where affiliates merchant banks were financing activities of other firms from the same business group. Conflict of interest between these two resulted in banks' failure to monitor the performance of their debtors [OECD, 1998]. This problem in economic literature is known as an adverse selection, the situation where lenders have an incomplete knowledge of the creditwor-thiness/ quality of borrowers. 6.3.2. Capital Account Liberalization Alongside financial liberalization, the capital account was also progressively liberalized. Deregulation of foreign exchange capital account transactions led to the expan-sion of foreign branches of Korean banks. Between 1994 and 1996, Korean banks opened 28 foreign branches in addition to 24 financial companies which were allowed to engage in foreign exchange business upon the conversion into merchant banks described above. In accordance with the design program, the short-term capital movements were liberalized in advance of long-term ones. Regard-less of the increasing list of industries open to foreign direct investments (FDIs), they were still subject to tight restrictions of unclear form. Very often they were denied because they could "disrupt the market" in the situation of surge short-term flows. In 1993, limits on the long-term foreign-currency denominated loans were relaxed, but the long-term borrowing remained restricted. This deci-sion led Korean banks to borrow funds from abroad for the short-term and lend these funds to domestic compa-nies for the long-term. It created a serious maturity mis-match where banks became prone to shocks such as an increase in interest rates. In this case, the burden imposed in the end of 1997 was a natural consequence of their net worth reduction. This is because, by definition, higher interest rates increase value of banks' long-term assets more than lowering short-term liabilities. According to the BIS-IMF-World Bank's statistics, lia-bilities to banks – due within the year – oscillated between 63 and 70 percent of total external liabilities in 1994–96. The situation looked even more precarious in terms of international reserves accumulated. By the end of 1997, total reserve assets only covered 38% of short-term external liabilities. The external financial liberalization, which led to the accumulation of short-term liabilities, exposed the Kore-an banking sector to problems like liquidity tightening, when some adverse news about the market caused sharp investor reactions. Firstly, when in 1997 investors CASE Reports No. 39 1996 % 1997 % 1998 % Fund Raising 118,769 100.0 118,022 100.0 28,360 100.0 Indirect finance Borrowings from DMBs Borrowings from non-banks 33,231 16,676 16,555 28.0 14.0 13.9 43,375 15,184 28,191 36.8 12.9 23.9 -15,003 54 -15,487 -52.9 0.2 -54.6 Direct finance (Commercial paper) (Stocks) (Corporate bonds) 56,097 20,737 12,981 21,213 47.2 17.5 10.9 17.9 44,087 4,421 8,974 27,460 37.4 3.7 7.6 23.3 49,749 -11,678 13,515 45,907 175.4 -41.2 47.7 161.9 Borrowings from abroad Others (trade credits, borrowing from governments, etc.) 12,383 17,058 10.4 14.4 6,563 23,997 5.6 20.3 -10,196 3,810 -36.0 13.4 Source: Bank of Korea, Flows of Funds, 1999
- 108. 110 Marek D¹browski (ed.) 200 180 160 140 120 100 80 60 40 20 stopped believing in the capacity of the government to bail out falling companies, banks and NBFIs, they rushed to pull their money out of the country. Secondly, the exist-ing currency mismatch (Korean banks' foreign liabilities were excessive to domestic assets) limited the ability to convert domestic currency into foreign currency. Thirdly, when the won/ dollar exchange rate started to depreciate, short-term foreign currency obligations of banks and non-banking financial institutions as a share of domestic assets increased significantly. The lack of liquidity of Korean banks was also clear by international standards. On aver-age, between 1995 and 1997, the ratio of liquid assets to liquid liabilities (a three months period is considered to be "liquid") was 60 percent in comparison to 100 – an inter-national standard [Nam et al.1998]. It is also important to note that even though the total external debt as a ratio of GNP increased from 13 to 22 percent between 1990 and 1996, it was not as large as in CASE Reports No. 39 Figure 6-7. External debt 0 1994 1995 1996 1997 1998 %%, USD Liabilities to banks - d ue within a year/ /Total liabilities International reserve assets (excluding gold)/Short-term liabilities Source: BIS-IMF-World Bank joint statistics, BIS web side Figure 6-8. Net foreign assets 20000 15000 10000 5000 0 -5000 -10000 -15000 Jan-94 May- Sep-94 Jan-95 May- Sep-95 Jan-96 May- Sep-96 Jan-97 May- Sep-97 Jan-98 May- Sep-98 Billion of Won Source: IMF IFS
- 109. 111 The Episodes of Currency Crises in Latin... Table 6-5. Liquidity ratio of the 10 largest korean banks 18 16 14 12 10 8 other Asia and Latin America countries [8]. In 1996 the ratios for the Philippines and Thailand were 54 and 46 percent, respectively. For Mexico, prior to the 1994 cri-sis, it was 35 percent [Park et al, 1998]. The same con-clusion is drawn from other indicators. The percentage share of foreign liabilities in total liabilities of the banking system was about 13 percent in 1996–97 for Korea, whereas in Indonesia averaged around 25 percent at that time. In Argentina on the other hand, it surged up to 20 percent in 1997 (IMF, IFS). There were two facts that seemed to be more important than the overall magnitude of external debt. One was its increasing trend since the early 1990's; another was associated with the high expo-sure of the banking system to short-term foreign bor-rowing discussed above. Progressive capital account liberalization also covered a higher ceiling on stock investments for non-residents, with the aggregate ceiling of 26 percent and individual ceiling of 7 percent by November 1997. Others included borrowing from international bond markets by Korean companies with prior notification, foreign purchasing of certain types of bonds or non-guaranteed corporate and SME bonds [IMF, 1998]. But despite of liberalization of capital account trans-actions some restrictions remained (see Appendix 2). 6.3.3. Credit Expansion The industrialization strategy implemented in Korea fuelled by the financial system liberalization resulted in domestic credit expansion. Furthermore, the surge in banking borrowing was typified by five important charac-teristics (all key for the future of the banking system): – Credit, in the most part, was extended to the pri-vate sector to finance new investments; public sector borrowing played a minor role, CASE Reports No. 39 1995 1996 March 1997 September 1997 80-90% 1 3 2 2 70-80% 2 2 1 1 60-70% 4 2 4 5 Below 60% 3 3 3 2 Source: Shin and Hahm (1998) cited in Nam et al. (1998) Figure 6-9. Currency mismatch* 6 1994Q1 1994Q3 1995Q1 1995Q3 1996Q1 1996Q3 1997Q1 1997Q3 1998Q1 1998Q3 1999Q1 1999Q3 Source: *Foreign liabilities over domestic assets, deposit money banks Source: IMF, IFS; own calculation [8] This data does not include offshore borrowing of domestic financial institutions, overseas borrowing of foreign branches of domestic financial institutions and borrowing of overseas branches of domestic enterprises. When these are incorporated, the total external debt jumps from 121 to 170 billion of dollars at the end of 1997 [Park et al, 1998].
- 110. 112 Marek D¹browski (ed.) 40 30 20 10 0 -10 -20 -30 -40 -50 -60 – Explicitly, it indicated a falling quality of loans and ampli-fied the probability of accelerating non-performing loans, – Borrowing, for the most part, was short-term and foreign currency dominated, – Non-banking financial institutions were the major intermediaries of funds, – Expansion of overseas branches of domestic financial institution resulted in the situation where 70 percent of total debt accounted for the banking sector, direct finance played a minor role [Dooley et al. 1999]. Bank credit grew more than 20 percent per annum in 1996 and 1997. Moreover, the ratio of bank credit to GDP was also increasing at a very high pace. The fact that loans were invested in the risky business of low profitability and declining rates of return (the discussion on efficiency of investments and profitability of the corporate sector was carried out in the previous section) led many of them to become non-performing, putting an extraordinary burden on the banking sector. Alongside the domestic credit growth, total claims on the private sector were also increasing. Between 1994 and the first quarter of 1996, total claims of deposits in banks as a percentage of GDP was averaging around 55 percent of GDP. From the second quarter on, it was sys- CASE Reports No. 39 Figure 6-10. Net domestic credit (annual percentage changes) -70 1995M1 1995M4 1995M7 1995M10 1996M1 1996M4 1996M7 1996M10 1997M1 1997M4 1997M7 1997M10 1998M1 1998M4 1998M7 1998M10 %% Source: IMF, IFS Figure 6-11. Claims on the private sector 80 75 70 65 60 55 50 45 40 1994Q1 1994Q3 1995Q1 1995Q3 1996Q1 1996Q3 1997Q1 1997Q3 1998Q1 1998Q3 Source: IMF, IFS
- 111. 113 The Episodes of Currency Crises in Latin... tematically growing, reaching 65 percent at the onset of the crisis. 6.3.4. Risk Assessment in the Banking System The sharp increase in bank lending – fuelled by very lax provisioning rules and insufficient risk assessment – was mirrored by the growing number of non-performing loans (NPLs) in Korea. The fact that there were no regular reports did not permit an adequate assessment of the health of the banking sector. The depth of the problem is clear when the data on non-performing loans as a percentage of total loans from 1996 is compared with the revised data for the same time period. The 1996 number for NPLs as a per-centage of total loans states for 0.8 percent, whereas the revised one for 4.1 percent. Partially, this discrepancy is connected with the classification of substandard loans in Korea, partially with the lack of regular reports already pointed. Usually, loans being three months plus in arrears are considered as substandard. But in the Korean Republic, this rule was extended to six months plus. Compulsory pro-visioning imposed on these loans varied from 20 to 75 per-cent, although this depended on the types of collateral and guarantees. In many cases, only bad loans (NPLs not cov-ered by collateral) were reported as non-performing. The transmission mechanism between the banks, non-banking financial institutions and Korean corporations led to the pre-sumption that the real number of compulsory provisioning was closer to the lower bounder – or was even below it. Adding to this story the number of corporate and banking bankruptcies in the end of 1997 and the beginning of 1998, it is obvious that Korean banks failed to adequately assess credit risk. The economy experienced troubles also in terms of sol-vency indicators of the banking sector. The capital adequa-cy ratio based on the Basle Core Principle requires a mini-mum ratio of eight. Yet, the domestic regulation was looser and required only four percent. In 1996 the actual capital adequacy ratio in Korea was just 9.1 percent representing a 2-percentage point drop from the 1993 value [9]. In Thai-land and Hong Kong, meanwhile, it was equal to 11.3 and 17.5 percent respectively. There were other factors like soft accounting rules, which gave the authorities the room to manipulate the capital adequacy ratio. The growing mer-chant banks' off-balance sheet credit guarantees were alarming prior to the crisis. In 1996, off-balance sheet cred-it guarantees were 49.3 percent expressed in ratio to total assets. It was higher by 12.5 percentage points than the 1993 average. For commercial banks the number was small-er and equal to 8 percent (the 1.2 percentage points drop from 1995). This piece of evidence points on the immense role the merchant banks played in the debt-financed growth strategy in Korea. Of course Koran banks had some prudential regula-tions to limit the probability of financial difficulties, but they were not successful in preventing excessive risk tak-ing. For example, in 1996, in order to prevent excessive risk-taking, Korean banks had constraints on foreign cur-rency exposure. The sum of long positions as a percent-age of total capital was limited to 15; the sum of short positions to 10 percent. The spot short positions were limited to 3 percent of bank capital or to 5 million of USD, whichever was greater. Maximum lending to a single bor-rower was 15 percent on the total bank capital, the same number as in United States. Nevertheless, the growing currency mismatch in Korea suggests that the imposed limits were relatively flexible and that foreign investors were mostly responsible for the growth of foreign assets of Korean banks. 6.4. The Onset of the Crisis The overview of the financial and corporate sectors' situation preceding the crisis showed that microeconom-ic distortions in Korean economy were accumulating for quite some time. Even if the crisis was initiated by the subsequent massive and abrupt capital outflow, the fact remains that many other factors contributed to the 1997 crash more than a simple herding. Definitely financial lib-eralization without adequate regulations imposed, long-lasting governmental control of the financial sector as well as the weak corporate governance were fundamental to the crisis. But apart from structural weaknesses sketched in the previous sections, there were other signs of vul-nerabilities building up in the economy and eventually causing the external liquidity crisis: – the slow down in the GDP growth rate, – the steady decline in stock price levels, lowering the value of banks and corporate equities and further reduc-ing the value of their net worth, – the sharp drop in the terms of trade affecting not only Korean export, but also disturbing banks and cheabols' balance sheets. According to Cho (1999) cited in Mishkin, et al (2000), the 20 percent drop in terms of trade accounted for more than 70 percent loss in aggre-gate corporate profits, – depreciation of the Japanese yen against the US dol-lar additionally weakening the country external position, – financial crashes in neighboring countries. [9] The capital adequacy ratio for Korea includes commercial banks only and is calculated under the Korean provisioning standards. Soft rules and many exceptions given to non-banking financial institutions would probably further decrease this number. CASE Reports No. 39
- 112. 114 Marek D¹browski (ed.) 1200 1100 1000 900 800 700 600 500 400 300 Taking into account risks like currency and maturity as well as the risk of default on external debt, it appears that all of these factors, individually and taken together, were responsible for building up pressures in Korea. These pressures finally turned out to be excessive leading the economy into the deep recession. But even with the clear financial difficulties Korean cor-porations were facing at the beginning of 1997, foreign investors did not downgrade Korea until the second quarter of the year when capital began to leave the country. The behavior of the exchange rate was, however, different since the won was loosing its value relative to the dollar through-out the whole 1996 (see section one). The stock market index was showing a similar trend to that of the exchange rate. In December 1997 it went down by more than 42 per-cent on a year-on-year basis. As in Park et al. (1999), foreign investors were evidently distinguishing between sovereign and domestic risks as late as October 1997. The same was true for rating agencies. Although Standard and Poors low-ered the credit rating of Korea First Bank on April 18, it was solely based on the fact that this bank was highly exposed to the two collapsing cheabols, Hanbo Group on January and CASE Reports No. 39 Figure 6-12. Stock market price index level, year end 200 31-Jan-94 31-May-94 30-Sep-94 31-Jan-95 31-May-95 30-Sep-95 31-Jan-96 31-May-96 30-Sep-96 31-Jan-97 31-May-97 30-Sep-97 31-Jan-98 31-May-98 30-Sep-98 31-Jan-99 31-May-99 Source: Reuters Figure 6-13. Capital flight 8000 6000 4000 2000 0 - 2000 - 4000 - 6000 - 8000 - 10000 - 12000 - 14000 1994Q1 1994Q3 1995Q1 1995Q3 1996Q1 1996Q3 1997Q1 1997Q3 1998Q1 1998Q3 1999Q1 mill. of USD Source: IMF IFS *Capital flight was calculated as follows: Capital Flight = (ΔExternal Debt + net FDI) - (CA (surplus) +ΔReserves)
- 113. 115 The Episodes of Currency Crises in Latin... Sammi Steel on March. The sovereign credit rating was still set at AA- level and did not decline until October. As in many other events of financial (banking) crises, it is extremely difficult to assess which model would be the most suitable one for Korea. The scenario of speculative attacks due to expected currency depreciation couldn't facilitate the run on Korea simply because the existence of tight regula-tions on currency forwards backed by corresponding cur-rent account transactions and the absence of currency futures market made this framework impossible [Mishkin, et al, 2000]. It was rather a consequence of creditors' herding behavior to withdraw their loans. The fact that Korea had a high level of short-term debt and only moderate interna-tional reserves favors this scenario. In retrospect however, and regarding the structural weakness of financial and cor-porate sectors this panic could be judged as rational. The set of macroeconomic indicators developed to mea-sure the probability of financial crises would blur the true pic-ture about the health of Korean economy. Here the problem is deeper and goes back to the early 1990's when the financial liberalization eased restrictions on short-term capital flows but left long-term ones in force. Together with the govern-ment's reluctance to give up its influence in most sectors of the economy, a series of bankruptcies of corporations under-mined investors' confidence in the quality of their assets. Thus, the Korean crisis was a result of interactions between financial and corporate sectors which failed to set prudential governance rules. The role the government played in calm-ing down the market was also of considerable importance since the policy was misdirected and instead of ceasing, it fuelled the crisis. The harsh tone and desperate announce-ments concerning good prospects of the economy gave a warning sign to investors. The fact that the authorities were bailing out collapsing corporations and banks raised doubts about the solvency of the whole economy. When the govern-ment announced the stock of official reserves at the end of November, the market confidence about the magnitude of usable foreign reserves was undermined and caused a further decline in the stock market index (regarding the course of events in Korea at the end of the 1997, see the overview of the chronology of selected news from the financial markets presented in Appendix 3). 6.5. The 1998 Recession and 1999 Recovery 6.5.1. The IMF Intervention in Asia On November 20, 1997 the exchange rate band was widened from 2.5 to 10 percent. The day after, the gov-ernment asked the International Monetary Fund to lend CASE Reports No. 39 support. On December 4, the IMF's Executive Board approved a Stand-By Arrangement with Korea of $21 bil-lion over the next three years based on the assumption of the GDP growth in 1998 of 2.5 percent. The World Bank and the Asian Development Bank provided an additional $14 billion as well as technical assistance. Individual coun-tries agreed on another $22 billion as a second line of defense credit. The total sum granted to Korea was equal to $58.4 billion, but the disbursement of money was based on certain conditions: 1. comprehensive financial sector restructuring with a clear exit policy, strong market and supervisory discipline, and independence of the central bank. Nine merchant banks were suspended; two commercial banks were cap-italised by the government; all commercial banks were required to submit plans for recapitalization, 2. the bases for corporate income and VAT widened in order to bear the costs of financial restructuring (fiscal measures were equivalent to about 2 percent of GDP and were consisted with the balance budget target), 3. an end to close the relationships between govern-ment, banks and corporations as well as an improvement in accounting (corporate financial statements had to be prepared on the consolidated basis), auditing (certification of external auditor) and disclosure standards, 4. the implementation of trade and capital account lib-eralization measures (liberalization of FDIs flows and opening money, bond and equity markets to capital inflows), 5. labor market reforms, 6. the publication and dissemination of key economic and financial data (IMF on line service). Nevertheless, the announcement of the program did not help to stabilize the exchange rate and the usable gross reserves fell to 8.9 billion US dollars at the end of December 1997. In the last quarter of the 1997, net cap-ital outflows were observed ($24.9 billion compared to $7 billion of inflows in the corresponding period of 1996). This was greater than the IMF initially expected (the pro-gram assumed net outflows of $11.1 billion in December 1997 and net inflows of $3.3 billion in 1998). Revision of the program was necessary, but even then it failed to pre-dict the total net capital outflows of $14.8 billion in 1998. The IMF prescriptions for crisis management may be a source of concern given the long-term adjustment condi-tions imposed, but the role of the Korean government is also questionable. As the government was injecting capital to falling companies, investors did not really believe in its commitment to market restructuring. As a result, the exchange rate dropped by 40 percent between Decem-ber 7 and 14, the stock market declined by 17 percent during December 3–10. It was the sharpest reduction since the beginning of October. As the crisis accelerated, the defense of the won became impossible and the
- 114. 116 Marek D¹browski (ed.) exchange rate band was abolished on December 16. Eight days later, the IMF-backed program was adjusted, assuming further monetary tightening, speeding up the lib-eralization of capital and money market, financial sector restructuring as well as trade liberalization. Thanks to these interventions, by the time of the second biweekly review on January 8, some degree of exchange rate stabil-ity was achieved. On January 28, 1998 international cred-itors agreed on a plan to officially roll over the short-term debt of approximately $24 billion. The agreement was signed in March which reduced Korean short-term debt by $19 billion from $61 to $42 billion in the end of April 1998, and helped to increase the stock of usable foreign reserves. In a Letter of Intent of February 7, 1998, the macro-economic framework of the plan was revised with the GDP growth forecast of 1 percent for 1998. It also included additional measures to target the fiscal deficit to 1 percent of GDP (from a previous surplus of 2 percent) and further development of financial sector's reforms in order to stabilize short-term debt payments. Given the fragility of the exchange rate, the monetary policy remained tight. Within this program revision, foreign investors gained an increased number of financial instru-ments available. On the other hand, domestic companies had easier access to foreign capital markets and were expected to introduce a number of measures needed to improve overall corporate transparency (i.e. introduction of external audit committee or outside directors). On May 2 1998, the Korean authorities updated the second stage of the program – economic restructuring. This time the fiscal deficit for 1998 was set at the level of 2 percent of GDP. Other changes included the introduc-tion of measures to strengthen the social safety net, fur-ther easing of restrictions on foreign exchange transac-tions and foreign ownership of certain assets. Once again, the GDP growth rate forecast for 1998 was changed to minus two percent. Apart from setting a new policy framework, the review also noticed the successful emis-sion of sovereign bonds which, together with the current account surplus and notable capital inflows, increased the stock of usable foreign reserves to $34.4 billion. Even though the vulnerability of the currency market was miti-gated and the interest rate was lowered, monetary policy remained cautious about maintaining the exchange rate stability. The fact that the won slightly appreciated against the American dollar and was relatively stable (averaging around W 1290 per US dollar comparing to 1500–1800 at the beginning of the year) in the mid-1998, let interest rates to go back to the pre-crisis level. This decision was announced in a Letter of Intent from July 24. To support economic activity and strengthen the social safety net, a supplementary budget was prepared. This time the fiscal deficit of 5 percent of GDP was projected. Output was anticipated to contract to -4 percent. Since the inflation rate moderated, the average rate for 1998 was expected to be 9 percent. The huge drop in imports and the export recovery let the current account to turn into a higher sur-plus of 10 percent of GDP. In a light of a severe recession, further steps towards corporate and financial sectors restructuring were inevitable. According to Nam et al. (1999), 94 financial institutions were suspended or closed. As the NPLs were growing, the government provided CASE Reports No. 39 Figure 6-14. Call rates overnight 30 25 20 15 10 5 0 1997M 01 1997M 03 1997M 05 1997M 07 1997M 09 1997M 11 1998M 01 1998M 03 1998M 05 1998 M07 1998M 09 1998M 11 %% Source: IMF IFS
- 115. 117 The Episodes of Currency Crises in Latin... support of 41 billion won (10 percent of GDP) for their disposal as well as for banks' restructuring. In result most of Korean banks achieved capital adequacy ratio between 10 and 13 percent. An acceleration of the recession was so abrupt that the GDP growth rate was once again corrected to -5 per-cent in August 1998. A supplementary budget was intro-duced to increase expenditure needed for restructuring distressed sectors. However, the exchange market stabil-ity allowed further reductions in the overnight call rate, from 25 percent at the beginning of the year to 8.5 per-cent at the end of September. This help disturbed com-panies and reduced the number of bankruptcies from 3000 to 1400 during the same period [Nam et al., 1999]. 6.5.2. Macroeconomic Environment after the Crisis After the eruption of the crisis, Korea moved into a severe recession which was not expected even by the IMF staff. As indicated above, the growth forecasts were revised downward, along with almost all the other restructuring program's elements. The fiscal deficit was also adjusted in order to accommodate weaker econom-ic activity and costs of financial restructuring. High inter-est rates (reaching 25 percent in January 1998) necessary to achieve exchange rate stabilization imposed a high bur-den on the real sector. In the first half of 1998, the gross domestic product plunged to -5.3 percent comparing to the year before and declined to 6.7 percent during the whole year. The reduction in output was mostly the result of a sharp reduction in domestic demand. The annual percentage changes in private and public con-sumption were equal to -9.6 and -0.1, respectively. The gross fixed capital formation dropped by 21 percent, and was probably the most responsible factor of the total out-put squeeze. Exports had declined by 13 percent, yet the slide in import was sharper and equal 22 percent. This resulted in the current account surplus of 10.9 percent of GDP in 1998. The unemployment rate jumped to 6.3 per-cent during the 1st half of the year and reached 6.8 per-cent at the end of the year (during the pre-crisis period it did not exceed 3 percent). The consolidated central gov-ernment budget was in deficit of 4 percent of GDP com-paring to the pre-crisis surplus (the initial IMF program assumed 0.2 percent surplus). The annual CPI inflation in the first quarter of the year reached 8.9 percent, but in the last quarter was down to 3.9 percent, which was lower than before the crisis. The 1998 recession seemed to be over in 1999 with some signs of stability visible already in the second-half of 1998 which helped to fuel investments. In the third quar-ter of 1999, gross fixed capital formation increased by 4.8 percent compared to its 1998 level. GDP growth boomed and exceeded 10 percent at the end of 1999 – supported by a monetary and fiscal policy expansion. The fact that short-term interest rates were cut from 23–25 percent to 5 percent significantly diminished the costs of borrowing. Low interest rates also helped banks and cheabols to bear the costs of debt restructuring. The budget deficit was reduced to 3 percent of GDP. The growth in imports reduced the current account surplus to around 6 percent of GDP. The exchange rate was stable oscillating around W1200 per US dollar. Nevertheless, there was a couple of exceptions to this impressive recovery. One of them was an unemployment rate of 6.3 percent, which despite of deceleration was still in a sharp contrast to the pre-crisis average; the other was the collapse of the Daewoo group in the middle of the year. 6.6. Conclusions The financial turmoil which erupted in Korea in 1997 did not really fit into any group of theoretical models of financial crisis existing in the economic literature at that time. It is just recently when researches tried do develop so-called third generation models with new sets of funda-mentals. Broadly speaking, they aggregate macro and microeconomic indicators of economic soundness. The Korean case is a good example of the great importance of complex reforms once the economy is switching from the centralized to market governance. The healthy macroeconomic background is a prerequi-site, but the high rates of growth, on their own, are not a safeguard of the long-lasting economic success. In the Korean Republic, highly leveraged cheabols were vulnerable to shocks like a collapse in investor con-fidence. Inadequate liberalization instead of hosting long-term capital welcomed short-term inflows. Additionally, and even more importantly, cheabols were not efficient and facilitated family connections. Tight relationships among cheabols, banks and the government made the transparency of financial and corporate sectors virtually impossible. Even if, on average. the prudential rules on banks in Korea were similar to those in western coun-tries, existing exceptions created a scope for choosing the "regain strategy". The number of financial instruments available increased once the financial sector was liberal-ized, and made the government attempt to control the stock of foreign capital ineffective. In the last Letter of Intent dated August 13, 2000 which finished the IMF intervention in Korea, the mem-bers of the IMF Executive Board stressed the impressive Korean rebound from the 1997 crisis. They said it was possible due to "supportive macroeconomic policies and CASE Reports No. 39
- 116. 118 Marek D¹browski (ed.) the competitive exchange rate; a wide range of structural reforms that addressed the weaknesses that contributed to the 1997 crisis and increased market orientation; a favorable external environment; and an improvement in confidence resulting from both the implementation of strong economic policies and the recovery and build-up in foreign exchange reserves". Despite the changes however, there is still a lot to be done especially in the area of structural reforms. In the special joint report the IMF and Korean Government agreed on major policies regarding further financial and corporate sectors restructuring in Korea. They identified at least a few problems, which despite the progress made in both sectors need to be solved. Bearing in mind all the reasons for the financial crisis in Korea, a number of them are key. Ongoing policy plans of the financial sector include the following measures: – while still being the owner of some commercial banks the government will let banks to operate on a fully commercial basis and will not be involved in the day-to-day management, – public funds should be only used to the extent nec-essary to facilitate the liquidation of failed institutions and restructuring of weak but viable banks, – financial transactions between affiliated companies or between institutions and their shareholders are con-fined to transactions in assets for which market price exist, – all related party transactions will be disclosed in audited accounts and reported to FSC, – exposures of commercial banks, merchant banks, specialized and development banks in excess of the new 20 percent or 25 percent limits will be subject to pro-gressive reduction. Further corporate sector restructuring consists of preparation of the consolidated statements assessing the overall financial structure as well as Daewoo resolution. Others encompass further improvements in financial transparency and accountability (Financial Supervisory Service) [10]. In spite of all this, the latest news from Seoul is not optimistic. The unsuccessful auction of Daewoo group undermined investor confidence and was immediately mirrored in the stock market decline. Additionally, the scale of corporate indebtedness before the crisis was so great that two years time after the crisis South Korea's conglomerates are still uanble to overcome it. The progress towards reducing bad loans has not been observed yet; the target of cutting the debt-to-equity ratio in the end of 1999 to 200 percent was not achieved. This begs the question about the efficiency and the speed of reforms carried out in the country. Indeed, one can question the IMF restructuring program applied in Korea, which failed to foresee the deepness of the crisis and caused the sharp output downturn, but in the light of current episodes one can also ask if the imposed condi-tionality was severe enough to bring about 'the second miracle. [10] Information about elements of banking system and corporate restructuring during 1998–99 can be found in Appendix 4. CASE Reports No. 39
- 117. 119 The Episodes of Currency Crises in Latin... Appendixes Appendix 1: 30 Largest Cheabols: April 1996 CASE Reports No. 39 Total Assets* (%) Debt/Equity Number of Subsidiaries 1.Hyundai 43.7 (6.94) 440 46 2.Samsung 40.8 (6.48) 279 55 3.LG 31.4 (4.99) 345 48 4.Daewoo 31.3 (4.97) 391 25 5.SK 14.6 (2.32) 352 32 6.Ssangyong 13.9 (2.21) 310 23 7.Hanjin 12.2 (1.94) 559 24 8.Kia 11.4 (1.81) 522 16 9.Hanhwa 9.2 (1.46) 712 31 10.Lotte 7.1 (1.13) 191 28 11.Kumho 6.4 (1.02) 480 27 12.Doosan 5.8 (0.92) 907 26 13.Daelim 5.4 (0.86) 424 18 14.Hanbo 5.1 (0.81) 648 21 15.Dongah 5.1 (0.81) 362 16 16.Halla 4.8 (0.76) 2457 17 17.Hyosung 3.6 (0.57) 362 16 18.Dongkuk 3.4 (0.54) 223 16 19.Jinro 3.3 (0.52) 4836 14 20.Kolon 3.1 (0.49) 340 19 21.Tongyang 3.0 (0.48) 305 22 22.Hansol 3.0 (0.48) 291 19 23.Dongbu 2.9 (0.46) 219 24 24.Kohap 2.9 (0.46) 603 11 25.Haitai 2.9 (0.46) 669 14 26.Sammi 2.5 (0.40) 3333 8 27.Hanil 2.2 (0.35) 581 8 28.Keukdong 2.2 (0.35) 516 11 29.New Core 2.0 (0.32) 1253 18 30.Byucksan 1.9 (0.30) 473 16 Total 286.9 (45.6) 669 *Figures in parentheses are the share of total assets of the corporate sector in Korea (629.8 trillion won as of the end of 1996). Source: Dongchul Cho and Kiseok Hong (1999).
- 118. 120 Marek D¹browski (ed.) CASE Reports No. 39 Appendix 2: Foreign Capital Controls in Korea, June 1996 Outflows Inflows Purchases abroad by residents Sales or issues locally by non-residents Purchases in the country by non-residents Sales or issues abroad by residents Capital market securities and money market instruments (MMI) Permitted freely in case of securities. Prior approval required to issue; the sale of securities permitted, but approval required for MMIs. Maximum foreign equity holdings in listed companies 18 percent; approval required for MMIs with the exception of approved institutional investors. Issue of won-denominated securities subject to prior approval; otherwise reporting requirement only. Credit operations Commercial credits Financial credits By residents to non-residents To residents from non-residents By residents to non-residents To residents from non-residents No restrictions deferred-receipt of exports if under 3 years. No restrictions with exceptions. Prior approval required. Authorisation required, except for selected enterprises. Loan transactions Deposits accounts Bank related transactions Borrowing abroad Lending to non-residents Lending locally in foreign currency Residents’ foreign exchange accounts abroad Non-residents domestic currency accounts Reporting requirements for loans with maturities above one year if founds exceed a given amount Prior notification if loans exceed a given amount, otherwise ex post notification or freely permitted for loans below a given amount Loan ceilings according to economic sector. Permitted up to a certain ceiling for corporations and individuals; no restriction for institutional investors. Permitted for the purpose of converting funds into foreign currency and transferring them abroad. Source: Blondal and Christiansen (1999)
- 119. 121 The Episodes of Currency Crises in Latin... Appendix 3: Chronology of the Korean Crisis, 1997 – In January 1997, the 14th largest conglomerate Hanbo Steel Co. went bankrupt. The long-term rating of three Korean banks with the high exposure to Hanbo was low-ered. Yet, the sovereign risk for Korea did not deteriorate, indicated the clear separation of the sovereign rating and rating related to the private financial institutions' problem, – Between March and April there were further defaults including those of the top thirty chebols including Sammi Steel on March 19 and Jinro Group on April 21. In spite of that, there were no signs of broader problems, – In July the Kia Motors asked its creditors for the work-out agreement on its debt of $8 billion to avoid receiver-ship, – In August, in spite of the intervention, the National Bank of Korea was not able to defend the exchange rate at the level of W 900 per US dollar. At the same time the gov-ernment announced its readiness to guarantee foreign cur-rency liabilities of Korea's financial institutions. This materi-alized on the October 14, when the government injected with money Korea First Bank and some other merchant banks, – At the beginning of October, various credit rating agencies downgraded Korea (S&P, Euromoney, Moody), because of the governmental decisions to rescue Korea First Bank and undertake Kia Motors, – On October 27, Bloomberg said the free fall of Kore-an won raised the concern the country would need the IMF assistance, but the government denied it, – On October 29, to attract foreign investors Korean newspapers announced the bond market would be opened from 1998. Nonetheless, it did not prevent the currency from further depreciation, – On October 30, the foreign press suspected that the Bank of Korea official reserves of 30 billion dollars did not include dollars borrowed through forward market transac-tions as the Korean government ordered banks to stop sav-ing dollars, – November 8, government accused foreign press of making unjustified rumours about Korea and asked to stop to destabilize the market. Ironically, investors seemed not to believe this statement, – November 18, Bank of Korea made emergency loans to 5 major commercial banks worth $1 billion, however still denied it would need the IMF assistance, – November 20, the band was widened from 2.5 to 10 percent daily, – November 21, the Minister of Finance and the Econ-omy announced it would ask a rescue package from the IMF, – December 4, IMF Executive Board approved a $21 bil-lion stand-by credit for Korea which together with the World Bank, ADB and individual government loans consti-tuted $58 billion. CASE Reports No. 39
- 120. 122 Marek D¹browski (ed.) Banking System Restructuring In the banking sector, a credit crunch resulted from high-er interest rates that increased the stock of non-performing loans (net domestic credit of the banking system dropped by 50 percent on the annual basis). At the end of March 1998, the ratio of loans being three months plus in arrears to total loans was 16.9 percent for banks and 14.5 percent for all financial institutions (Sang- Loh Kim; 1998). In order to deal with this problem, in April 1998 an independent supervisory authority was established to apply international prudential standards. Additionally, to help banks resolving bad loans government committed W32.5 trillion to the NPLs' Resolution Fund and planned to resolve W100 trillion of bad loans via auctions, capital increase or subordinated bond issues. Five banks were shut down as well as 16 mer-chant banks in 1998. The Korea First Bank and the Seoul Bank had been nationalized after they repealed 87.5 percent of their stock. The Financial Supervisory Commission (FSC) set a mini-mum target for capital adequacy ratios according with the BIS standard for banks and merchant banking corporations to be achieved by the end of 2000. Together with the Min-istry of Finance measures on improving the disclosure, accounting and accounting standards were introduced (accounting of securities was going to be based on the mar-ket instead of the book value). Starting from July 1, loans of at least three months in arrears but less than 6 months were categorized as non-performing. Prompt corrective action system was introduced imposing recommendations, mea-sures and orders on the unsound financial institutions (i.e. banks with the capital adequacy ratio lower than 8 percent minimum). Corporate Restructuring Corporate restructuring in Korea proceded on two sep-arate tracks. One was a debt workout for the smaller cheabols and other large corporations; the other included a package for the top five cheabols. Reform bills were passed by the National Assembly in February 1998 and among oth-ers included: – Tax Exemption and Reduction Control Act – tax breaks for company restructuring were provided, – Bank Act – the limit on bank ownership of a corpora-tion's equity increased from 10 to 15 percent, or higher with Financial Supervisory Commission (FSC) approval, – Corporation Tax Act – non-deductibility of interest on "excessive" debt was moved from 2002 to 2000, – Foreign Direct Investment and Foreign Capital Induce-ment Act – takeovers of non-strategic companies by foreign investors without government approval were allowed. For-eign investors could acquire 33 percent of shares without board approval (23 percent increase compared to the pre-crisis limit), – Antitrust and Fair Trade Act – new cross guarantees were prohibited. Elimination of existed cross guarantees was planed by March 2000, – Financial Supervisory Committee maintained relatively lax rules on accounting for restructured debt in order to help Korean banks to negotiate substantial rate reductions and conversions of debt into equity or low-yield convertible bonds [Mako, 1999]. CASE Reports No. 39 Appendix 4: Banking System and Corporate Restructuring
- 121. 123 The Episodes of Currency Crises in Latin... References Agenor P. , J. Aizenman (1999). "Financial Sector Ineffi-ciencies and Co-ordination Failures. Implication for Crisis Management". NBER Working Paper No. 7446, December 1999. Balino T., et al. (1999). "The Korean Financial Crisis of 1997 – A Strategy of Financial Sector Reform". Internation-al Monetary Fund, Working Paper No. 28. Bank For International Settlements, Annual Report 1997, 1998. Bank of Korea, various issues, http://www.bok.or.kr Blondal S., H. Christiansen (1999). "The Recent Experi-ence With Capital Flows to Emerging Market Economies". OECD, Working Paper No. 3. Cailloux J., S. Griffith-Jones (1999). "Global Capital Flows to East Asia. Surges and Reversals, Institute for Development Studies". University of Sussex, November. Chong Nam, Joon-Kyung Kim, Yeongjae Kang, Sung Wook Joh, and Jun-Il Kim, Corporate Governance in Korea, OECD Conference on Corporate Governance in Asia: A Comparative Perspective, March 1999. Corsetti, et al. (1998). "What Caused the Asian Currency and Financial Crisis?". Part I: Macroeconomic Review, NBER Working Paper No. 6833, http://www.nber.org/papers Dooley M.P., S. Inseok (2000). "Private Inflows when Crises are Anticipated: A Case Study of Korea". Financial Supervisory Service, Monthly Review, Volume I No. 8., August 2000 http://www.fss.or.kr Hahm Joon-Ho and Mishkin F. S. (2000). "Causes of the Korean Financial Crisis: Lessons for Policy". NBER Working Paper No. 7483. International Monetary Fund, International Financial Sta-tistic, various issues. Kaminsky G., C. Reinhart (1996). "The Twin Crises: the Causes of Banking and Balance of Payments Problems". International Finance Discussion Paper No. 544, Board of Governors of the Federal Reserve, March. Lane, et al (1999). "IMF-Supported Programs in Indone-sia, Korea, and Thailand. A Preliminary Assessment". Inter-national Monetary Fund, Occasional Paper No. 178, Wash-ington. Mako W.P. (1999). "Corporate Restructuring in Korea and Thailand". OECD Conference on Corporate Gover-nance in Asia: A Comparative Perspective. Mishkin F.S. (1996). "Understanding Financial Crisis: Developing Country Perspective". NBER, Working Paper No. 5600. OECD Economic Surveys, 1997–1998, Korea. Park D., C. Rhee (1998). "Currency Crisis in Korea: How Has Is Been Aggravated". Sang-Loh Kim (2000). "Current Trends and Prospects of Korean Economy and its Restructuring". Yoon Je Cho, Changyong Rhee (1999). "Macroeconom-ic Views of the East Asian Crisis: A Comparison". Paper pre-pared for the World Bank Conference, "Asian Corporate Recovery: Corporate Governance and Role of Govern-ments", Bangkok, Thailand. CASE Reports No. 39
- 122. 125 The Episodes of Currency Crises in Latin... Comments to Papers on Asian Crises by Jerzy Pruski* The comparison of the four papers on the 1997–1998 financial crises in Asia clearly indicates that Korea, Thailand, Malaysia and Indonesia fall in the group of relatively homogenous economies. 1. After the Second World War, all these countries were characterized as under-developed economies. They man-aged to change this situation due to very high rates of eco-nomic activity accompanied by very high investment and saving rates. The civilization leap might not have been achieved without the prominent role of the state. As the years passed, government interference in economic life gradually diminished but still remains relatively strong. In a market economy, the government used to significantly influ-ence the investment allocation and consequently determine the development of the selected branches of economy. Therefore, it had to target the specific financial instruments in order to push the private sector into the preferred area of economic activity. As a result, the Asian economies are characterized by strong interdependence of the private sec-tor and government spheres. 2. The aspiration for high and sustainable rates of eco-nomic growth forced all the countries to liberalize their domestic financial markets. As a result of financial market liberalization, accompanied by weak regulation, the newly established financial intermediaries engaged excessively in financing projects characterized by high branch and credit concentration. Then the Asian countries became increasing-ly aware about the high costs of not participating in global-ization of the world economy. Due to very specific and structural reasons, they liberalized short-term capital flows far before liberalization of FDI and long-term capital flows. The method of liberalization of both the financial markets and capital accounts significantly contributed to the scope and intensity of the Asian crisis. 3. The extreme state interference in economic life caused excessive concentration of market power and own-ership. The powerful, family-owned conglomerates were characterized by unbelievably high rates of economic expansion accompanied by insufficient improvement in pro-ductivity. The low economic efficiency was caused by a number of factors. The most important is the imperfect organizational structure (e.g. in Korea – the role played by chairman office), weak corporate governance, excessive output diversification (preventing conglomerates from effective competition in a few carefully chosen branches) and growing domestic and foreign competition. Moreover, the Asian countries experienced a relatively slow process of democratization. With high ownership concentration, intensive state interference in economic life and young democratic institutions, the countries exhibited poor trans-parency of public finance and government relationships with private sector. 4. The strong desire to preserve the existing structure of ownership led conglomerates to finance their rapid expansion through credits from financial markets with a negligible role played by capital markets. In the presence of weak financial regulation and high rates of economic growth, such a practice was responsible for conglomerates' uncontrolled indebtedness on the domestic market and excessive credit risk concentration in the financial sector. Moreover, liberalization of the capital account allowed firms to incur new debts on international financial markets and continue their economic expansion. Many companies financed their activity as if they didn't face hard budget con-straint. As a result, the share of FDI and long-term invest-ments in total foreign debt was small but the ratio of debt to equity and ratio of short-term foreign debt to foreign official reserves reached high and unsustainable levels. 5. Imperfect and opaque markets emerged as a result of the state interference in the economy and high concentra-tion of private ownership. Under such conditions, investors (especially foreign ones) couldn't obtain all the necessary information for a correct risk assessment. Moreover, there * The £ód¿ University CASE Reports No. 39
- 123. 126 Marek D¹browski (ed.) is some evidence that even the conglomerates were unable, at least to some extent, to assess the total risk of their rapid expansion. 6. The analysis of the main features of the Asian economies leads to the conclusion that ineffective micro-economic foundations, coupled with opqaue markets, and the resulting asymmetry of information were responsible for the 1997–1998 financial crises. That line of reasoning is sup-ported by the fact that Korea, Malaysia, Thailand and Indonesia demonstrated good economic performance in terms of the GDP growth, inflation, unemployment and public finance deficit. The relatively high current account deficits were the only exception to overall positive macro-economic picture. 7. However, in the second half of the 1990's, globaliza-tion and higher openness exposed the Asian economies to higher competition, appreciation of local currencies and deterioration of terms of trade. Global markets severely verified the microeconomic efficiency of industrial corpora-tions and financial institutions and, to some extent, improved the transparency of particular markets. 8. The above remarks emphasize the role of structural factors and microeconomic inefficiencies as the main rea-sons behind the financial crises in Asia. In general, the situa-tion should improve, due to deep structural and institution-al reforms. However, the positive effects of supply-side reforms usually require long time. But it is known that the Asian countries recovered from crisis quickly and since then their economies have performed surprisingly well. Econo-mists have explained many aspects of financial crisis, but no doubt there is still enormous scope for additional research work. CASE Reports No. 39