PresentationF
- 2. MS Thesis Presentation Monetary Policy Transmission Mechanism: Role of Stock Prices Muhammad Ilyas Siddiqui 321-SE/MS Eco/S13 Supervised By Dr. Arshad Ali Bhatti Assistant Professor, IIIE International Institute of Islamic Economics IIU, Islamabad January 7, 2016
- 3. Presentation Plan Introduction Why this study? Research Hypothesis Literature Review Data and Methodology Econometric Model Results and Discussions
- 4. Introduction • Monetary policy is one of the two major policies (other being Fiscal policy), the Government uses to control the economy. • Monetary policy refers to any action taken by the Central bank, on behalf of the Government, to influence either the supply of money or the price of money, as given by the rate of interest (Moynihan and Titley, 2007) • Monetary policy transmission mechanism (MTM) explains the channels or paths through which changes in interest rate are transmitted into real economy. MTM channels: • Interest Rate Channel: It is core of Keynesian theory and works through changes in real interest rate as cost of capital. • Exchange Rate Channel: It works in face of flexible exchange rate through change in relative currencies value. • Credit Channel: It exits due to credit market imperfections and product of two sub-channels that is bank lending channel and balance sheet channel.
- 5. Introduction…. • MTM channels…. • Other Asset Prices Channel: It works through effect of monetary policy on relative asset prices and real wealth. • Expectation Channel: This channel works through the psychology of economic agents. Monetary policy actions may have an effect on the economy through their impact on the confidence and expectations of economic agents about the future outlook of the economy (Dabla-Norris and Floerkemeier, 2006). • Relevance of Stock Prices • Tobin’s q theory. Tobin (1969) defines q as the market value of firms divided by the replacement costs of capital at going rates. At a high value of q, it is favorable for firms to issue new stock rather to purchase new plant and equipment which implies that investment is more likely. • Keynesian theory. According to Keynesian economist, rise in interest rates make bonds more attractive for investors hence, they switch over from stock market to bond market. This situation induces reduction in equity prices. Therefore, lower equity prices causes a lower q which in turn, leads to lower investment spending resulting in reduced output. • Wealth effects on consumption when equity prices come down as a result of contractionary monetary policy, the value of real wealth with people decreases. They feel less rich and withdraw consumption spending. This affects aggregate demand adversely. Ultimately, aggregate output fall (Modigliani ,1971).
- 6. Introduction…. • Relevance of Stock Prices • Tobin’s q theory. Tobin (1969) defines q as the market value of firms divided by the replacement costs of capital at going rates. At a high value of q, it is favorable for firms to issue new stock rather to purchase new plant and equipment which implies that investment is more likely. • Keynesian theory. rise in interest rates make bonds more attractive for investors hence, they switch over from stock market to bond market. This situation induces reduction in equity prices. Therefore, lower equity prices causes a lower q which in turn, leads to lower investment spending resulting in reduced output. • Wealth effects on consumption when equity prices come down as a result of contractionary monetary policy, the value of real wealth with people decreases. They feel less rich and withdraw consumption spending. This affects aggregate demand adversely. Ultimately, aggregate output fall (Modigliani ,1971). • Monetary Policy and Stock Prices; Nature of Relationship • Negative- a positive shock in money supply will lead people to expect a contractionary monetary policy (Habibullah et. al., 1996; Sack, 2001). • Positive- money supply increases in response of higher money demand that is due the enhanced economic activity (Sellin, 2001; Alatiki and Fazel, 2008). • Depended- In case the investors anticipate fully the expansionary monetary policy, the stock market will not react; however, the stock price will increase if the shock is unexpected (Olivier Blanchard, 2011).
- 7. Why this study? • There is limited empirical literature that focuses on the MTM in Pakistan. • The empirical literature on stock prices as one of MTM channels is even much scarce. • We do not find any empirical study which examines the role of stock prices as a separate MTM channel in Pakistan. • To fill the literature gap, we analyze the impacts of monetary policy on output and inflation through stock prices.
- 8. Literature Review • There are three strands in the literature. • The first strand asses relation between monetary policy and other macroeconomic variables and discusses their mutual relationships. It covers a large space (Ando and Modigliani, 1963; Baumol, 1965; Bosworth, 1975; Ahmed, 1999; Hussain and Mahmood, 2001; Nishat and Shaheen, 2004). • The second strand gives analyzes stock prices as one of the channels of the MTM along with other asset price channels. Concerning studies are less in number (Mishkin, 1996, 2001; Agha, 2005; Aqeel, 2011). • The third strand concentrates on stock market as an only transmitting channel of monetary policy. Such studies are very few throughout the economic literature. (Chami, 1999; Sourial, 2002; Ehrmann and Fratzcher, 2004 and Seong, 2013).
- 9. Literature Review • The major studies in first category are : • Thorbecke (1997); Filer (1999); Sack (2001); Shahbaz and Ahmed (2008); Foresti (2006); Mahmood (2001); Ajit Singh (2008); Saleem et. al. (2013); These studies analyze the interrelationships between monetary policy, stock prices and economic growth. Most of the studies use M1, M2, money market rates, Repo rate , six month treasury bills, USA Federal Fund rate, GDP, KSE-100, exchange rate, CPI, IPI as indicators. Major techniques used are OLS regression, ECM, Granger causality and vector auto regression. Findings do not endorse single result. All types of relationships- positive, negative, significant, insignificant, unidirectional, bidirectional.
- 10. Literature Review • Important studies in the second category are: • Mishkin (2001); Sack (2002); Agha (2005); Baig (2011); • • Important studies in the third category are: • Chami et. al. (1999); Sourial (2002); Seong (2013); Ehrmann and Fratzscher (2004)
- 11. Data and Methodology • We use quarterly data from 1991:Q1 to 2010: Q2 • 1-Policy (independent) variables: They are the basic instruments through which monetary policy works. • Narrow Money (M1) • Short Run (Interbank) Money Market Rate (MMR) • 2-Intermediate (target) variables: By using these variables monetary policy hit the target variables. • Nominal Exchange Rate (EXR) • Private Sector Credit (PSC) • Stock Market Index, KSE-100 (SMI) • 3-Target (dependent) variables: They are the ultimate target of monetary policy and among the macroeconomic variables through which economy is controlled. • Real Gross Domestic Product (GDB) • Consumer Price Index (CPI) • 4-Exogenous block: These variables represent the shock originating in the external world. • World (IMF) Commodity Prices (COMP) • Federal Fund Rate (FFR)
- 12. Estimation Methodology • Following the literature, we utilize vector auto regression (VAR) as estimation technique. • Since the work of Sims (1980), VAR models are common in estimation of MTM • They can summarize the dynamic relationships among variables (Bernanke and Gertler, 1995) • They are appropriate both for advanced and emerging countries as they can generate multivariate forecasts and their data requirements are less demanding. • Problems of large-scale econometric models can be avoided with VAR. •
- 13. Estimation Methodology • Second Stage of Estimation • A structural VAR (SVAR) is estimated for Pakistan at the second stage to confirm the results of base line VAR model. • SVAR is important: • 1- to identify a monetary policy shock based on economic theory. • 2- give allowance of imposing enough restrictions to identify an exogenous policy shock, without having to specify complete or large-scale models of the economy. • 3- it takes into consideration money aggregates and short run interest rate jointly and help to identify monetary policy shocks appropriately. • 4- it is considered suitable for small open economies such as Pakistan. • 5- it takes no account of the properties of time series data. (Perera, 2013).
- 14. Econometric Model Following Perera and Wickramanayake (2013), we use VAR in two stages. First Stage of Estimation The unrestricted - baseline VAR model based on assumption that system is recursive and therefore Cholesky decomposition can be employed for identification. We can specify our baseline model in following matrix form yt = k + A(L)yt-1 + Bxt + ut (4.1) where yt = vector of endogenous variables k = vector of constants xt = vector of exogenous variables ut = vector of serially uncorrelated disturbances ( zero mean and a time invariant covariance matrix) A(L) = coefficient matrix (a matrix polynomial in the lag operator) In the baseline specification, the vector of endogenous variables such as real gross domestic product (GDPt) and the consumer price index (CPIt). The measure of monetary aggregate that is narrow money supply (M1t) and the domestic nominal short-term interest rate given by interbank money market rate (MMRt). We write it as follows: y, t = (GDPt CPIt M1t MMRt) (4.2)
- 15. Econometric Model The general form of structural VAR model of order p is as follows: Aoyt = co +∑p i=1 Aiyt-1 + εt (4.3) yt = 7*1vector of endogenous variables yt = (COMPIt, FFDRt, GDPt, CPIt, M1t, MMRt, SMIt) Ao = 7*7 contemporaneous matrix Ai = 7*7 autoregressive coefficient matrices εt = 7*1 vector of structural disturbances assumed to have zero covariance. multiplying both sides of (4.3) with A-1 o to get the reduced form as: yt = ao + ∑p i=1 Bi yt-I + et Where ao = A-1 o co Bi = A-1 o Ai et = A-1 o εt (εt = Ao et) We can derive structural disturbances by imposing appropriate restrictions on Ao
- 16. Components VD of GDP Time Period VD of CPI MMR (baseline) 0.210728 1.378362 3.124916 5.113285 Q4 Q8 Q12 Q16 0.168485 2.810061 4.561628 4.239366 M1 (baseline) 5.842879 9.610432 12.77320 15.29505 Q4 Q8 Q12 Q16 0.236158 3.194236 10.83325 20.41191 SMI (stock market channel) 3.750656 5.686760 7.136972 8.245667 Q4 Q8 Q12 Q16 1.564091 0.986778 2.220572 6.091740 MMR (stock market channel) 0.227681 1.291607 3.065665 5.210945 Q4 Q8 Q12 Q16 0.126579 2.506933 4.535215 4.337248 M1 (stock market channel) 3.506075 4.75628 5.915295 6.958337 Q4 Q8 Q12 Q16 0.382503 0.382503 2.323669 4.749184 SMI (all channel endogenous) 2.882958 4.383710 5.801023 7.069753 Q4 Q8 Q12 Q16 1.876468 1.295246 5.727449 13.90800 Table 5.3: Variance decomposition (VD) of GDP and CPI - Baseline Model versus SM channel
- 17. Components VD of GDP Time Period VD of CPI EXR ( exchange rate channel) 0.419714 1.039137 1.361452 1.479931 Q4 Q8 Q12 Q16 4.182306 18.13292 28.89163 33.82420 MMR ( exchange rate channel) 0.608441 1.740702 3.052492 4.430022 Q4 Q8 Q12 Q16 0.613067 0.446562 0.325828 0.428118 M1 ( exchange rate channel) 5.858769 5.858769 13.46811 16.40583 Q4 Q8 Q12 Q16 0.312333 3.159039 11.10071 20.36989 PSC (bank lending channel) 19.18298 19.25027 18.97339 18.46213 Q4 Q8 Q12 Q16 0.281123 0.352633 1.470962 4.030985 MMR (bank lending channel) 0.579082 1.745838 3.642293 6.079923 Q4 Q8 Q12 Q16 0.224509 2.571515 3.857057 3.564473 M1 (bank lending channel) 8.299370 9.652282 11.36304 12.94446 Q4 Q8 Q12 Q16 0.286655 4.649080 11.27792 18.13594 Table 5.4: Variance decomposition (VD) of GDP and CPI - EXR and PSC channels
- 18. Components VD of GDP Time Period VD of CPI SMI (SVAR) 0.533167 2.463111 3.506124 4.073808 Q4 Q8 Q12 Q16 0.402664 0.664915 1.828915 5.235307 PSC (SVAR) 20.39558 18.98588 17.80314 16.72251 Q4 Q8 Q12 Q16 0.548890 0.246428 0.217958 0.197210 EXR (SVAR) 0.748846 1.335725 1.592848 1.724218 Q4 Q8 Q12 Q16 3.117706 10.41930 17.89436 21.60731 Table 5.5: Variance decomposition (VD) of GDP and CPI – SVAR models
- 19. Conclusion Both baseline unrestricted VAR and structural VAR confirm the significance and potency of monetary policy in Pakistan during the research period. Our estimation results show that Stock market channel is the least effective. Also it does not bear any Granger causality toward any variable. This is in accordance with Sourial, 2002, who declared it as a future channel. The overall findings of the study suggest that, the stock market as an MTM channel is observed as an active channel only when working in collaboration with other channels; it has significant impact on prices in long run. This is more likely due to strong correlation between output and SMI In response of tight monetary policy, the aggregate price levels responds significantly in first six quarters and then dissipate and keep on rising for the rest of (4 quarters) period. Output shows V- shaped response, bottoming out after third quarter and then keeps on declining for rest of (7 quarters) period.
- 20. Conclusion Exchange rate channel is the most significant channel in Pakistan. Both in VAR and SVAR, it has bi-directional causality with inflation. Also it has uni- directional causal relationship with interest rate in both VAR and SVAR The second most significant candidate is bank lending channel. It has bi-directional causality with output and un-directional with inflation the findings about stock market channel call for caution because the share ownership is not still very common in Pakistan rather firms prefer bank credit on equity sharing for their financing needs (Agha, 2005).
- 21. Policy Implications Stock market channel needs more attention. As in collaboration with other two channels, stock market gave much better results which is signal for policy makers to understand the situation. The stock markets of Pakistan have been very sensitive to political upheavals and law and order situations in Pakistan. Fortunately, both these factors are in process of substantial improvement in Pakistan due to military operations. Therefore, it has now become easier for policy maker to exploit fully the stock market channel in Pakistan. Exchange rate channel can be focused for further improvement in Pakistan economic conditions. The liberal exchange rate regime is already in operation and volume of remittances has increasingly improved. Both these aspects have strengthened the exchange rate channel in Pakistan. Further reforms in baking system are needed to enhance the performance of bank lending channel. Past history of Pakistan witness that the process of reformation in banking system has boosted the economy.
- 22. Data Sources • Monthly reports and bulletins of State Bank of Pakistan, especially working paper no. 54, 2013. • Websites of International Financial Statistics (IFS) • Publications of Karachi Stock Exchange (KSE) • Various data sources of International Monetary Fund (IMF) • Pakistan Economic Survey (Different Issues). Ministry of Finance, Government of Pakistan.