2a A Dynamic and Adaptive Approach to Distribution Planning and Monitoring JFP Apr 2009.pdf

52 Journal of Financial Planning | AP R I L 2009 www.FPAjournal.org
Contributions
David M. Blanchett, CFP®, CLU, AIFA®, QPA, CFA, is a
full-time MBA candidate at the University of Chicago
Booth School of Business in Chicago, Illinois. He won the
Journal of Financial Planning’s 2007 Financial Frontiers
Award with a paper titled “Dynamic Allocation Strategies
for Distribution Portfolios: Determining the Optimal
Distribution Glide Path.”
Larry R. Frank, Sr., CFP®, a wealth advisor and author,
lives in Rocklin, California. He shifts people’s focus from
an income-centric to a wealth-centric viewpoint to help
them better understand how to live on their investments.
He can be reached at LarryFrankSr@BetterFinancial
Education.com.
D
istribution planning research is
entering its second generation.
The first generation of distribu-
tion research provided answers to rela-
tively static questions such as “what is an
initial safe withdrawal rate” and “what is
the best (constant) allocation for a distri-
bution portfolio.” Recognizing that distri-
bution decisions are not made only once at
retirement, an expanding body of research
is exploring retirement as a more dynamic
period, in which changes can be made as
situations warrant.
This paper will explore the question,
“What is a safe withdrawal rate?” not only
initially, but also currently. It will do so from
an adaptive perspective, where the with-
drawal rate is revisited annually based on
the performance of the underlying portfolio
or unforeseen expenditures. It will also be
revisited simultaneously with the effects of
the dynamic relationships of (1) constantly
decreasing distribution periods as the client
ages, which in turn allow for (2) an increas-
ing supportable withdrawal rate with a sim-
ilar probability of failure rate throughout
retirement. The study modeled the revisits
annually, but the data are displayed as five-
year slices through the data for simplifica-
tion of reporting purposes.
An adaptive approach to distribution
planning, where the withdrawal rate is fluid
and not constant, can dramatically improve
the probability of success of a distribution
strategy. Reviewing the withdrawal rate also
allows for the withdrawal amount to be
increased as situations warrant, which
ensures that a retiree is maximizing his or
A Dynamic and Adaptive Approach to
Distribution Planning and Monitoring
by David M. Blanchett, CFP®
, CLU, AIFA®
, QPA, CFA, and Larry R. Frank, Sr., CFP®
BL A N C H E T T | FR A N K
• This paper advances the “second-
generation approach” to the sustainable
withdrawal rate question.The study
evaluates the ongoing sustainability of
the withdrawal rate that is revisited
every year. The withdrawal rate itself
(not the dollar value) is increased,
decreased, or stays the same based on
the probability of failure for the
remaining target distribution period.
• This adaptive approach recognizes
that sustainability decisions do not
occur just once at retirement, but
should change as situations warrant
throughout retirement.To support
ongoing sustainability decisions, annual
probability of failure of the current
withdrawal rate is presented in this
paper, summarized in five-year slices
through the data.
• As a person ages, this allows for slowly
changing to higher withdrawal rates
associated with those shorter remain-
ing distribution periods. For example, a
15-year distribution period is more
appropriate for an 80-year-old than for
a 60-year-old retiree. Essentially, a
person “ages through the data” from
longer distribution periods to ever
shorter distribution periods.
• Revisiting the withdrawal annually
allows for higher withdrawal rates if
the portfolio performs well, for
unplanned or unforeseen additional
expenses, or for lowering withdrawal
rates if the portfolio is underperform-
ing. This is done through comparison
of the current withdrawal rate to
benchmark data to evaluate the asso-
ciated probability of failure rates of a
given portfolio mix and remaining dis-
tribution time.
• The revisiting approach introduced in
this paper is simpler than some of the
complex decision rules that have been
previously introduced, and is therefore
easier to implement and change as the
client ages and portfolio values change.
Executive Summary

Contributions
her lifetime income. As the client ages, his
or her remaining time dynamically gets
shorter. The adaptive approach in this
study demonstrates that the withdrawal
rate may be slowly increased as the client
ages through management of the client’s
exposure to probability of failure with his
or her current withdrawal rate and remain-
ing distribution time.
Previous Research
The assumption of a constant real with-
drawal amount from a portfolio is a consis-
tent theme in past distribution research.
The sustainable withdrawal rate is typically
defined as a percentage of assets where an
initial amount, adjusted for inflation, is
assumed to be taken from the portfolio for
the entire distribution period. For example,
a 5 percent withdrawal rate from a $1 mil-
lion portfolio would result in a $50,000
withdrawal in year one. The withdrawal in
year two, though, would not be based on 5
percent of portfolio assets; instead the
withdrawal would be $50,000 plus infla-
tion. The $50,000 withdrawals, adjusted
for inflation, are typically assumed to con-
tinue until the end of the distribution
period, where the strategy would either be
judged as “passing” (that is, it was able to
withstand the withdrawal for the entire
distribution period) or “failing” (in other
words, it ran out of money).
Recognizing that distribution planning is
more dynamic than just an initial with-
drawal decision, a number of studies have
introduced logic, or decision rules, to help
advisors determine how and when to
adjust a withdrawal amount over time.
Guyton (2004) introduced perhaps the
most well known study involving decision
rules, which were tested in a follow-up
paper by Guyton and Klinger (2006).
Guyton employs a variety of rules, such as
the Portfolio Management Rule, the Infla-
tion Rule, the Withdrawal Rule, and the
Prosperity Rule, to help an advisor deter-
mine how to adjust the withdrawal over
time to ensure the ongoing sustainability of
the portfolio.
While Guyton’s research provides valu-
able insight into distribution planning, it
takes a very “one size fits all” approach to
distribution planning. For example, he uses
a fixed 40-year period for his study. Forty
years is a relatively conservative estimate
for the distribution period, and each
retiree (or retired couple) will have a dis-
tribution period that is unique based on his
or her unique age, health, and family his-
tory. In contrast, the analysis conducted for
this paper considers nine different time
periods (10 to 50 years in 5-year incre-
ments) and takes a simpler approach to
adjusting withdrawals.
Bengen (2001) tested a variety of
performance-based withdrawal methodolo-
gies where the distribution rate was
adjusted during retirement in response to
changing portfolio conditions. One test
involved potentially increasing the real dis-
tribution rate by 25 percent or decreasing
it by 10 percent based on whether the
client was in a bull or bear market. For this
paper, the authors use a more precise
methodology than Bengen’s to determine
whether an adjustment is necessary.
Bengen’s analysis was also limited to 55
test “runs” due to his reliance on historical
time series sequence data; in contrast, this
paper takes a bootstrap approach and uses
100,000 runs per scenario.
Pye (2001) addressed the probability that a
withdrawal amount will need to be reduced
over various periods and for various with-
drawal rates. Stout and Mitchell (2006) took
a similar approach to Pye where the with-
drawal is potentially increased or decreased
annually, based on the likely sustainability of
the portfolio. Stout and Mitchell’s dynamic
model employs three types of controls—
portfolio deviation thresholds, withdrawal
adjustment rates, and absolute withdrawal
rate limits—in order to prevent overreac-
tions to short-term market movements.
Stout and Mitchell note that downward
adjustments should be more immediate than
upward adjustments, and this paper incorpo-
rates that concept. This paper could be seen
as an extension of Stout and Mitchell’s work.
Portfolio Ruin, Balancing Sequence Risk and
Longevity Risk
A key consideration when constructing a
distribution portfolio is how much to allo-
cate between equities and fixed income/
cash. The long-term importance of the
allocation decision has been well docu-
mented by Brinson, Hood, and Beebower
(1986), and more recently by Tokat, Wicas,
and Kinniry (2006). The potential benefit
of non-constant equity allocations for dis-
tribution portfolios has been noted by
Blanchett (2007).
Two key risks must be addressed when
making the allocation decision: sequence
risk and longevity risk. Sequence risk is the
risk, or really the implication, of starting
the distribution period in a bear market (or
a market with low or negative returns).
Sequence risk will affect clients differently
since people retire at different times. A
recent study by Watson Wyatt (Watson 2008)
found that retirees with a substantial portion
of their assets in defined-contribution type
investments are especially prone to
encounter sequence risk because they tend
to retire during market booms (that is,
when their 401(k)s are doing well). Market
busts tend to follow market booms, which
is the type of market these retirees are
likely to face shortly after they retire (think
mean reversion).
Sequence risk is directly correlated to
the market risk of the portfolio. Therefore,
more conservative portfolios with lower
equity allocations will have a lower likeli-
hood of encountering sequence risk. But
more conservative allocations increase
longevity risk, or the risk of the outliving
one’s resources.
As life expectancies continue to
increase, the need to create portfolios that
can sustain 40 or more years of inflation-
adjusted withdrawals is becoming increas-
ingly important. Studies by Cooley, Hub-
bard, and Walz (1998); Tezel (2004);
Cassaday (2006); and Guyton and Klinger
(2006) all confirm the importance of
equities in order to maintain an inflation-
adjusted withdrawal over a prolonged
period. Equities are important because
www.FPAjournal.org AP R I L 2009 | Journal of Financial Planning 53

they have historically increased the return
of a portfolio versus cash or fixed income.
Return is a key driver of portfolio success;
however, higher returns are typically
accompanied by higher variability, or
standard deviation.
Higher equity allocations, therefore,
decrease longevity risk but increase
sequence risk. Viewed differently, if a client
is unlucky and encounters poor initial
returns (sequence risk) during the distribu-
tion period, it is likely that the withdrawal
amount will need to be reduced in order for
the portfolio to survive. If a client is lucky,
though, and encounters high initial returns,
it is likely the withdrawal amount can actu-
ally be increased. The key is revisiting the
withdrawal to determine whether it is still
reasonable given the current value of the
portfolio. This is the primary concept that
will be explored in this piece.
‘Revisiting’ Methodology
Four different equity allocations were con-
sidered for the analysis because risk toler-
ances differ across investors and testing
only one allocation (60/40, for example)
would ignore this fact. The four different
allocations considered for the paper were
20/80 (20 percent equity and 80 percent
cash/fixed), 40/60, 60/40 and 80/20. The
equity piece of the allocation is split two-
thirds to domestic large equity and one-
third to international equity, while the
cash/fixed income allocation is split evenly
between cash and fixed income. For exam-
ple, the allocation for the 60/40 portfolio
would be 40 percent domestic large blend
equity, 20 percent international equity, 20
percent cash, and 20 percent intermediate-
term bond.¹
The withdrawal is revisited each year for
this study. Based on the underlying proba-
bility of failure for the portfolio, the with-
drawal amount can either be
increased by 3 percent,
decreased by 3 percent, or
stay the same. Note, this
change is in addition to a
potential increase due to
inflation. All withdrawal
amounts are considered to
be in real terms, eliminating
the effect of inflation on the
analysis. This was done by
subtracting the monthly
inflation rate, which was
defined as the increase in
the Consumer Price Index
for all Urban Consumers
(CPI-U)², from the monthly
returns used in the analysis.
CPI-U was used as the defi-
nition of inflation because it is the most
common definition.
The probability of failure of the with-
drawal is calculated each year based on the
portfolio allocation, the number of years
remaining in the target period, the previ-
ous year’s withdrawal, and the portfolio
value at the end of the previous year. The
withdrawal dollar amount is decreased by 3
percent if
• The probability of failure for the port-
folio is greater than 20 percent when
the target end date is 20+ years away
• The probability of failure is greater
than 10 percent when the target end
date is 11–19 years away
• The probability of failure is greater
than 5 percent when the target end
date is 10 years or fewer away
The withdrawal amount is increased by 3
percent if the probability of failure is less
than 5 percent. If neither of the above con-
ditions is met, the distribution dollar
amount does not change (except for infla-
tion or deflation adjustments).
The target period is defined as the length
of the assumed distribution period (30
years, for example). As the portfolio pro-
gresses over time, the remaining target dis-
tribution period, or planning period,
decreases. For example, if the target period
is 30 years, after 4 years the target period
would be 26 years.
To build a reference table where the
withdrawal rate (as percentage of current
assets) based on the equity allocation and
remaining period could be determined, the
probabilities of failure were calculated for
each of the four equity allocations (20/80,
40/60, 60/40, and 80/20) for periods
between 1 and 50 years (in one-year incre-
ments) and for withdrawal rates from 0
percent to 100 percent (in 1 percent incre-
ments). (A sampling of the data points
used in the reference table can be found in
Figure 2 on page 56.)
For the revisiting strategy, the probabil-
ity of failure was calculated for each year of
each run of each scenario to replicate the
dynamic approach an advisor would take
when working with a retired client as mar-
kets change. The probability of failure is a
very fluid number that can change a great
deal over time. As an example, Figure 1
includes the probability of failure for 50
runs of a Monte Carlo simulation with a 6
percent initial real withdrawal rate over a
30-year period for a 60/40 portfolio where
the withdrawal is adjusted during the dis-
tribution period based on the previously
described methodology.
The probability of failure at the begin-
ning (year zero) is the same for each of the
50 Monte Carlo runs, 39.01 percent. But as
the portfolio progresses through the distri-
bution period, the probability of failure
changes for each of the runs. In the aggre-
gate, the probability of failure tends to
decrease because the initial failure rate is
higher than the respective target probabil-
Contributions
“As the client ages, his or her
remaining distribution period decreases
and the client dynamically moves
through the ever-shortening distribution
periods. As a result, their current
benchmark withdrawal rate and
associated probability of failure
adjusts with time.
”

ity of failure (20 percent). This causes the
withdrawal amount to be reduced by 3 per-
cent a year until it falls within an accept-
able probability of failure range. Figure 1
demonstrates why it is important to regu-
larly revisit the likelihood of failure for a
distribution strategy, as the probability of a
portfolio failing (or succeeding) is always
changing over time.
The actual returns used for testing pur-
poses were created through a process
known as bootstrapping. This is a type of
simulation analysis where the in-sample test
period returns are randomly recombined to
create annual test returns. For the analysis,
monthly return information was obtained
on the four test asset classes from 1927 to
2007 (81 calendar years) and randomly
recombined to create hypothetical real
annual rates of return for the analysis. For
example, the monthly real returns for each
of the four categories for the same month
(such as June 1961) would be recombined
with monthly real returns from 11 other
months (such as March 1930, January 1995,
May 1979, and so on) to create each hypo-
thetical annual real return. A benefit of the
bootstrapping process is that no assump-
tions need to be made about the distribution
of hypothetical returns (for example, lep-
tokurtic and positively skewed).
Distributions from the portfolio were
assumed be taken once a year at the begin-
ning of each year. Each test scenario was
subjected to a 100,000 run bootstrap
Monte Carlo simulation. The simulator
used for this research was built in
Microsoft Excel by one of the authors. The
original simulator built for this analysis
used 10,000 runs; however, the simulator
was expanded to accommodate more runs
(from 10,000 to 100,000) due to the vari-
ability in the results of the 10,000 run
series. Over two billion Monte Carlo simu-
lations were performed for this analysis,
the majority of which were used to create
the reference table (Figure 2 shows a
sample of the data points) to calculate the
ongoing sustainability of a given with-
drawal rate.
The portfolios were assumed to be held
in tax-deferred accounts and therefore any
tax implications of the withdrawals are
ignored. Based on the bootstrapping
methodology, it is implicitly assumed that
the portfolios are rebalanced back to their
target allocations monthly. Any potential
costs associated with the rebalancing were
also ignored.
Nine target distribution periods (10, 15,
20, 25, 30, 35, 40, 45, and 50 years) and
nine real distribution rates (4, 5, 6, 7, 8, 9,
10, 11, and 12 percent) were tested for the
four different equity allocations (20/80,
40/60, 60/40, 80/20), for a total of 324
dynamic scenarios. Selecting the appropri-
ate initial distribution period is typically a
function of the planned length of the dis-
tribution period. For example, if you use
age 95 as the base mortality date for all
retirees (this methodology is discussed in a
paper by the authors titled “In Search of
the Numbers,” currently unpublished),
then for a client 65 years old the initial dis-
tribution period would be 30 years. As that
client ages, his or her remaining distribu-
tion period decreases and the client
dynamically moves through the ever-
shortening distribution periods. As a
result, their current benchmark with-
drawal rate and associated probability of
failure adjusts with time.
Results: Static Withdrawals for Comparison
Before reviewing the potential benefits of
revisiting a distribution portfolio see
Figure 2, which illustrates for baseline
comparison purposes the probabilities of
failure for a static distribution strategy.
After reviewing Figure 2, it is possible to
understand why 4 percent has widely been
noted as the safe initial withdrawal rate. The
probability of failure for a static 4 percent
withdrawal rate for a 60/40 portfolio over a
30-year distribution period was only 4.07
percent, and only 2.01 percent for a 20/80
portfolio. Viewed differently, approximately
1 of every 25 clients who take $40,000 a
year from a $1 million initial portfolio
(adjusted for inflation) is likely to run out of
money during the 30-year period. Even for a
50-year distribution period the probability of
failure for a 4 percent initial withdrawal rate
for a 60/40 portfolio was only 16.91 percent.
Higher withdrawal rates, such as 6 percent,
are commonly viewed as too aggressive
because the probability of failure is much
higher (such as 39.01 percent for a 60/40
portfolio with a 30-year distribution period).
But not everyone retires precisely at age 65
(age 95 minus 30 years of distributions),
and a 6 percent withdrawal is an incredibly
conservative withdrawal for a 15-year distri-
bution period.
Results: Dynamic Distributions
Contributions
4WUc`S ( =U]WU `]POPWZWbWSa ]T 4OWZc`S T]` # AO[^ZS ;]bS
1O`Z] @ca
`]POPWZWbg
]T
]`bT]ZW]
4OWZc`S
0OaSR
]
@S[OWWU
BOÙSb
S`W]R
3ZOâSR BW[S GSOà

$

# # # !
@ca BVOb OaaSR @ca BVOb 4OWZSR

Figure 2 includes a sampling of the infor-
mation used to create the reference table
to determine the ongoing success rates
when testing the dynamic strategies. As an
example, if a 60/40 portfolio with 20 years
remaining in its target period had a value
of $800,000 and a $40,000 real with-
drawal, the withdrawal rate, as a percent-
age of current assets, would be 5 percent
($40,000/$800,000), which corresponds to
a probability of failure of 2.07 percent.
Because the probability of failure at this
point is less than 5 percent, the withdrawal
amount for the next year would be
increased by 3 percent to $41,200 (from
$40,000). If, however, the portfolio value
was only $500,000, the withdrawal rate
would be 8 percent ($40,000/$500,000).
Because this corresponds to a probability of
failure that is greater than 10 percent
Contributions
4WUc`S ( `]POPWZWbWSa ]T 4OWZc`S T]` O DO`WSbg ]T AQSO`W]a 0OaSR ] O 4WfSR EWbVRÒeOZ @ObS

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
2Wab`WPcbW] S`W]R GSOà
`]POPWZWbg ]T 4OWZc`S
³#
,#³
, ³#
,#³
,³'#
,'#³

$
# '

%!
!$
!$

!

'

!!
'!

!
!

!
$
#%'
'#

#$
%#
'%%
'''#

!
#

%#
' !
'#%

%
!
$%

! %
##%
%#!%
$
#
##
%
$#

#'
$!$
%%%

$ %
%%#
'''
''

'
!

%%##
%'#
'%#
''#$
'''

%
!'
!%
## !
$''
'
'!%
'%#
$
%

$!$
%%
#!
%
!$
'
#

%%'
''
''''

!%
!!
!%
%#
'#
'%
''$
'''#

'
#$
# %
%

'!$#
'%$'
''
$
$
!#
!
$!%
$$!
!
'%
'#

'$
'$%
'''

!
%
$!%'
$
'$
''
'''
''''

%
$
!'
$#%
!
'%
'$#'
''
''#
$$
$#
! $%
#
$'
%$#
$%
'!$%
'$%
!'
%'#
''#
''''

$#!
!%$'
%%#!
'!
''
''%
'''%

%'
#

%!%%
%
'#
'%%'
''!
''%#
%
'
!'
#
%#!

'$!
'##
'%#
!'
' !
'''

'$%
'
'$$!
''#'
''$
''''

''
!
##%#
%%
'
'##
'
''#
''
'
$
!
$ !
%%%!
''
' $
'#''
'%'
'%
'$#
'''

%
#$#
'%
'%%$
''%
'''

!#!'
$
%%
' #
'$
'%
''$
''#

'
%$
$#!
%'
$%
'
'$#
'
$
'#%
''''

!
$#
'%
'!%
''
'''

$'
!'
$!%#
!%
'!
'%!
''!
''$$
''
#$
!#$
#
$%!
#
#%%
' $
'$
'
]PO
S (
4WUc` ` ]` O D
S T
OWZc`
OPWZWbWSa ]T 4 SO`W]a 0OaS
O`WSbg ]T AQ
D O
e
O
WbVR`
SR E
SR ] O 4Wf S
b
Z @O
'

%

#
$

e
O
WbVR`
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

#

OZ
e
eO
OZ

#$

#

%
%
'

%

!
g
#

S`W]R

b`WPcbW]
2Wa
'

#
'
$
%

''
''

%
'$
'

''
'''
''

%%

#

''

%%
$

!#

SOà
G
GS

'$

'
# !
'
''

'

'

''
''
#
'
!

''

''

'
%
!

#
#

%

''
$

''
'
'#
$#

'
%

'

'
$
%

#

$

$

$

$

$

$

$

$

$

# '

!

!$

!

#

'
#
%

#

'
''
%

%
'
%!

''
!
!

##

%

' !!
#
'

#
'
# $
$
''
#
''
$

''

'
%
!
$
%

%

'
''
#$
''

%
'
'
%
!

!

''
#
%
'
'
%

%%

%
''

'
'!
%
!
'

#
''

''
$
'
!
'

''
%
''

#
''

'$$
$
'

!

'
''

%
'
'

''

#
%%
$
%
!
$#

'
%

''

''

'!

%
'
%
$#
$#
!

'
'
%

'
%$
%

'

%

'

%

#
$

$

$

$

$

$

$

$

$
$

$

!
'

!$ %
!$

!

%

!

#!
%
%
##
%
!

%
$

%

'#
'
!

!

'

## !

%
''
% $
'
'

!

'
%
!$#
'
'
'$#
'

%
$#
%

!'
#$
'

''
#
$
%
'
'
%

#

'
'
''
#$
%
!
'
'

$'

%
!
!

''
''
'

#
$

'
'$
#
#
'
##
'
'
#
!

''

''
!

'

%

%
%%
%
##
!
#

'

''

'
%
%
'

'

%

!
%

%

$'

#

!
$$

!
''

''

!

%
'
$
'
#
%
!
%%
%
!
$

'
!
'

$
''
%

#!

'
$
%

#

$
$
'

! $
#
##

!
'
!
#
$#

%
!$

%%
$
#

%

'
%
#

!

'

#
$$
% $#

!$
#!
!

%
!#
$!$ #

%%
#

%
%
'$

!$
'

%
$
!
$
%
!

'
$
%
! $
$
$
''

#
'

%
'
$$
!
$

$

''

'

%

#
%
'
'

$

'$
'
%
''

'
#''
'
' $
''
%

$ !

!

%
'
#
'
#
!

%!
%
%
!

$
'
'

''
#
%
'

#
'
$
'

#
%

#

'
!

%
%

''

$
%%
$#!
#$
'$
#

!#$
$

!

%
'

$#
'
#
%
$

#

''
#

$
'
%
'

'
,#³
³#
OWZc`
]POPWZWbg ]T 4
`
,#³
, ³#

S
`

#³
,'
,³'

#

#
'

BL A N C H E T T | FR A N K Contributions
(actually 55.25 percent), the withdrawal
amount for the following year would need
to be reduced by 3 percent, from $40,000
to $38,800. As a reminder, this calculation
was performed for each year for each of
the 100,000 runs for each of the 100 differ-
ent scenarios.
But when the withdrawal amount is
revisited on an ongoing basis, as it likely
would be when working with an advisor,
the actual real withdrawal amount received
by a client will likely change based on the
performance of the underlying portfolio
due to market forces. Figure 3 illustrates
the results of the five different percentile
slices from a $1 million portfolio over a
sample 30-year distribution period. The
initial withdrawal rate is assumed to be 6
percent ($60,000 from $1 million), the
target period is 30 years, and the portfolio
allocation is 60/40. The withdrawal
amounts are based on those runs that sur-
vived the entire distribution period.
As is evident in Figure 3, the range of
potential withdrawals changed over time,
primarily based on the performance of the
underlying portfolio—or viewed differ-
ently, the luck of the retiree. For example,
based on the information in Figure 3, and
the revisiting methodology discussed previ-
ously, those unlucky retirees (in the 95th
percentile or the worst 1 in 20), would see
their initial $60,000 withdrawal reduced to
$39,210 by the 30th year. But those lucky
retirees in the fifth percentile (or the best 1
in 20) would see their initial $60,000
withdrawal increased to $121,968 by the
30th year. The median expected withdrawal
at the 30th year was $82,133.
Revisiting the withdrawal amount also
reduced the likelihood of failure versus
using a static withdrawal amount. An
example of this is included in Figure 4,
which is based on the same assumptions
for Figure 3. Sequence risk is best con-
trolled by evaluating the current with-
drawal rate, since declining markets push
the current withdrawal rate up. (Sequence
risk is always present for all retirees who
take a higher withdrawal associated with
higher probability of failure.) Time does
not cure sequence risk unless near-term
rising market values (lucky retiree) reduce
the current withdrawal rate such that the
probability of failure is now lower.
It is important to note that using the
revisiting approach is going to result in
clients who take the same initial with-
drawal rate (say 5 percent) ending up with
very different withdrawal amounts during
the distribution period, depending on
their actual markets experienced. To give
the reader a better idea of the distribution
4WUc`S !( =U]WU @SOZ EWbVRÒeOZ /[]cba T`][ O $
]`bT]ZW] eWbV O ;WZZW] 7WbWOZ DOZcS T]` O !GSO`
BOÙSb S`W]R
EWbVRÒeOZ
/[]cb
3ZOâSR BW[S GSOà

$

'#bV
S`QSbWZS
# # # !
bV
S`QSbWZS
bV
S`QSbWZS
#bV
S`QSbWZS
#bV
S`QSbWZS
4WUc`S ( `]POPWZWbWSa ]T 4OWZc`S WT @SdWaWbWU 7a CaSR DSàca O
AbObWQ 2Wab`WPcbW] T`][ O $ ]`bT]ZW] eWbV O
;WZZW] 7WbWOZ DOZcS T]` O !GSO` BOÙSb S`W]R
`]POPWZWbg
]T
4OWZc`S
3ZOâSR BW[S GSOà
#

!#
!
#

#

#

@SdWaWbSR 2Wab`WPcbW]
# # # !
AbObWQ 2Wab`WPcbW]

Contributions
of withdrawal amounts using the revisiting
strategy, the withdrawal amounts at the
target end dates for the 95th percentile
(worst 1 in 20), 90th percentile (worst 1 in
10), 80th percentile (worst 1 in 5), 50th
percentile (median), and 20th percentile
(best 1 in 5) are included in Appendices
1–5. The corresponding probabilities of
failure for each of the scenarios is
included in each appendix to help the
reader easily reference the probability of
that revisiting strategy surviving the target
distribution period.
Revisiting, or adjusting, the withdrawal
amount throughout the distribution period
reduced the probability of failure signifi-
cantly. A static real withdrawal amount,
based on a 6 percent initial distribution (or
4WUc`S #( `]POPWZWbWSa ]T 4OWZc`S T]` O DO`WSbg ]T AQSO`W]a EVS`S bVS EWbVRÒeOZ /[]cb 7a
@SdWaWbSR /cOZZg

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
`]POPWZWbg ]T 4OWZc`S
³#
,#³
, ³#
,#³
,³'#
,'#³

$'
# '
%$!
!

!
$
!%'
% %
$#

%
#
%!
'
%!
$$!

$
!
'
'%
##%
!$
#'

'
#%#
$'
%$
$'!
'
''%#

%
'
#
%

!'''
%#
'

%
$
$
$#!
!
!$#
#
%%$

%!
'
!
%
$
%''

#$$$
$
!'%
$ '
%$
!$!

''!
'''

'
%
!%
%!#
!#
##!
%%$#
'!
'#
!'
#
!
'
#
!% #
#'
%$
#'
$'
%%
#
'!#

!%
'#%
$#'
%%
!#
#
$$%
%$
%$$%
' %
'''#

!#!
# '
!
!!%
%'
' $
'%'!
''%#
$
!$$
%
!
!!!#
#
%!#
$#
'!$#
$
'

$
''
!
#'!%
%$
$
#$#
#'
!!
#!
''
''

$
$
'!
!
$
# #
'$
''
'''
$#
#%
'!
!'
#
$
#
'
'$
!''
$#
%
!#
!!$%
#
$!
%%$
$$
#$
$$
%
% ##
'%!%
'''

!
'$

!''
%#
'#
'$
''#
'''
!%
#
!#
#'#
%%!
$%
!#
'!$
'$'#
#
%!%

%
!!
#$
%$
!$!
'
#%'
%
!$$
%
'$
'''

!'
#
$$
!#
%%%
'%
'
''%!
'''%
!
#!
$
!
##'
%$
#'%
'%
'%'
##
'
!
$
'
##''
%!
'
'!
$!
%!%
!$
%#
''!!
''''

!
#%
#!
$
%''
'!#%
'$
''%
'''
%!
' #
#
!$
##%
%$
$
'!'
'%!
#%
#
#
$!#
'
$
% #
$'
'$#
#$
#!
#
'!
''$
''''

!%!
$%
'
#'$

'!'
'%
''%%
'''%
#
%%
%
!#
##'!
%#$
$%
'!
'$
#'
' '
$'
'#
!
$
% %
'
'#$
dWa
S #(
4WUc`
S
@
]P
`
SR /cOZZg
aWb
]` O D
S T
OWZc`
POPWZWbWSa ]T 4 SO`W]a EVS
O`WSbg ]T AQ
D
DO OZ /[]
e
O
WbVR`
S bVS E
S` cb 7a
'

%

#
$

e
O
WbVR`
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

#
g

OZ
e
eO
OZ

'
#
'
'
!
# '

$ '
$
%
$
%
#

$

!
%
!
#

S`W]R

b`WPcbW]
2Wa
#$#
!#
'
!!
%
$
%
!
$
'

#
''
'
#
%
#
#
'
''

'
$$
%

%
$$
$ '
$
!
'
'

!$
%
!'

##

!#

''
SOà
G
GS
$

!$
''

'
'$
%
$$
'

''
'
%

!

%
$$
%
#
#$

''
!
%
'
%

%

!
#
#
''

#$
!
#
$

''
''
''
!!
'
#

#
%
!

'
'
%

$
!
%

$

'
$
%

#

$

$

$

$

$

$

$

$

#

!
!

'
#
!
'
'

''
%

# '
$

#
%
'
!

%

!

%

%

''
$'
%
%
!

$
%
$

!
!#
# '
#
!
!
$#
!

'
!

$
# #
%

$
''
'$

!
'
!
$

#
'
%
'

'

%
!!
!#

'

'
'
''

'
%%
##
!
%
%
!
%

''
!
'
'
'
'
'
!'

!
'$ #
$
'

'

%

!#
$

%!
'
'
#

'
'$
%
'
#
%

%%
#
'
$

#
!

''

'

%

'
!

%%

'
'$

%
$
'
''

%
'
$
!
'
%

#

'
!
$

%!
!
!

%
'
!#
'
#
%
#

'

%

#
$

$

$

$

$

$

$

$

$
$

$

!
$#
%
'

%
$
!
$

$#
!
'

#
!
$#

#
%
%

%!

#

%

%
$
#

!
'
$#
! !
%

#
!

!
!'
#
$
!#

$#
#

'

#
'
#
!$$
$
''
#
''
%

#
!!!
%
#

'
%

''
'#
'
%
'
%
'
#
%
!

'

!

'
!
''

'

'

!
%
##

$
'
%
%

##'

#!
!
%
'

'
!$
#'
!#
$

%
%
!
#
$
!#
#

'
''
'
%!
'
'
#
'

$

%!
%

#'

%
!

''

''

%
#
%%
%

$
'!
#$
!
%
'
!'
$
$
%

##'

#
!
$
%

#

%
#

'
''
'

''

$
%
#
'
%!

'
%
'

'
$
%

#

$
'
%
!$
'
!

'

%
!
#$$$

''
$
%
%

!

%!

%$

%
#

##
'

$

$$

'
!''
!$#
' #
$$
'

!
#

!#
%

%

%
$
%

%%

$
%

#
!!$

$
$
'$
$
$
%
!
'
#

!

'
$
%%
'
'
'

%%
$#
#
'

!
!
'
#

$

#
'

$
'

##

#

$
'!
'
%

##''
'
$
!

'
'
!

!
%
$

!!

'
%

%
'
#

!$

%
#$

%

!
%
#
'$'

'#$
!
#
'
$

'

%

'$#

'

%
#
$

'
#
$'
#
'
#
#'
%
'
!
$

'

$!

#
%
,#³
³#
OWZc`
]POPWZWbg ]T 4
`
,#³
, ³#

S
`

#³
,'
,³'

#

#
'

Contributions
$60,000 from a $1 million portfolio), had a
39.01 percent probability of failure at 30
years, while the probability of failure for
the revisited strategy was only 9.83 per-
cent. Figure 5 includes the probabilities of
failure for the same scenarios in Figure 2;
however, unlike Figure 2, the probabilities
of failure for Figure 5 incorporate the revis-
iting methodology where the withdrawal
amount was increased, decreased, or kept
the same based on the ongoing probability
of success for the portfolio. The revisited
strategy also had a consistently lower prob-
ability of failure as seen in Figure 6.
Some readers may question how it is
possible to have both a lower probability of
failure and a higher median withdrawal
amount when revisiting is used. This
occurs for two reasons. First, the with-
drawal amount was reduced with poor
portfolio performance. Based on the data
used to develop Figures 3 and 4, 88.47 per-
cent of the runs had withdrawal amounts
less than the initial $60,000 at year 5,
69.70 percent at year 10, 55.31 percent at
year 15. Reducing the withdrawal amount
as situations warranted better enabled the
portfolio to survive the entire distribution
period if the market returns were low.
Second, the dispersion of the ending
account values was much tighter for the
revisited methodology than the constant
approach. The revisiting methodology
ensures that the withdrawal amount is tai-
lored to the underlying portfolio; if the
portfolio performs well the withdrawal
increases, if the portfolio performs poorly
the withdrawal decreases. Contrast this
dynamic approach with the constant with-
drawal approach, where the same with-
drawal is taken regardless of the underly-
ing portfolio value.
It is worth noting that the probability of
failure actually increased for some of the
more conservative scenarios. For example,
the probability of failure for a 4 percent
distribution for a 20/80 portfolio over 25
years based on the constant methodology
was only .05 percent, yet was 3.54 percent
based on the revisit methodology. The pri-
mary reason for the increase was that a
probability of failure of less than 5 percent
was deemed acceptable when there were
ten or fewer years to the target end date
when determining whether to adjust the
withdrawal. For this scenario (4 percent
withdrawal, 20/80 portfolio, 25 year distri-
bution period), the 95th percentile with-
drawal amount (or worst 1 in 20) at the
25th year was $52,834. The failure rate in
4WUc`S $( `]POPWZWbWSa ]T 4OWZc`S T]` DO`W]ca EWbVRÒeOZ @ObSa 4WfSR ]` @SdWaWbSR T]` $
]`bT]ZW]

!

#
$
%

'

4 % 4 @D $ 4 % @D # 4 $ @D 4 # @D @D
`]POPWZWbg
]T
4OWZc`S
GSOà # GSOà ! GSOà !# GSOà GSOà # GSOà # GSOà
4 % 4 @D $ 4 % @D # 4 $ @D 4 # @D
@SOZ EWbVRÒeOZ @ObS ;SbV]R]Z]Ug
GSOà
# GSOà
! GSOà
!# GSOà
GSOà
# GSOà
# GSOà
@D
## !
%
!
%
'
' #
'!
!%
# %
$#%
%!%%
%%
%%
!%
#
!!!#
#
%%!
##'
##%
##'!
!'
#$
!'

##%#
$
$!%#
'
!
!'
#'#
!
!$
!#
%
'
$
#
!
!#!'
!'
!
%
'!
!#
$
#
%

%
%'
''

$'
#
!$$
#%
#
#!
' #
%%
!'
$
$#
!%
!
%!
#
1]``Sa^]RWU `]POPWZWbg ]T 4OWZc`S
³#
,#³
, ³#
,#³
,³'#
,'#³
]`bT
S $(
4WUc`

]P
`
]ZW]
T
O
]` D
DO
S T
OWZc`
POPWZWbWSa ]T 4 Sa
b
OZ @O
e
O
WbVR`
O`W]ca E ]` $
SR T
dWaWb
S
SR ]` @
4Wf

$

!

#
$
%

'

`]POPWZWbg
]T
4OWZc`S

SOà

SOà
SOà
SOà
# G
GS
!
! G
GS
# G
GS
G
GS ##
%
@D
D
% 4
4
GSOà # GSOà !

#
!
@
% 4
4
%

%%
! #

%!
%

#
!!!

!
%
%
$#

#

%
!
%

# 4
S ;
% @D
$ 4
! GSOà !# GSOà
'
b
OZ @O
e
O
WbVR`
SOZ E
@
# 4
D
% @
$ 4
#$
%
#
!' $

#

!

# '

'
'

!'
!

# @D
;SbV]R]Z]Ug
4
$ @D
GSOà # GSOà # G
#
#
4
D
$ @
!

!
'
%

%
! #

!
'

%

@D
D
D
GSOà
$
@ D
@
'
#
$#
%
!
%

!$$
!

#
#

!

'
SOà
SOà
SOà
SOà
# G
GS
# G
GS
G
GS
# G
GS
!
'
' #

!
##'
%%

%%
##'
%
##

%!
%

%
!

%

%
!
%
%

,#³
³#
'

#
#

#
]POPWZWbg ]T 4

'
!

#!
!
!$
!

#
#
!
!
#'

%
!
$
$
%
##
Sa^]RWU `
]``
1
,#³
, ³#
'
$'
#
#

S

''
'

#

%
!#
%

$
OWZc`
4

#³
,'
#
,³'
#
%
#
#!

#

!
%!

%
!

%
'

Contributions
the 24th year of this strategy was only .01
percent. In other words, the revisited
approach resulted in a higher lifetime with-
drawal amount, which is arguably each
retiree’s objective, and virtually every run
that failed did so in the last withdrawal year.
Figure 6 compares the table data from
Figures 2 and 5 for the portfolio composi-
tion 60/40 (other portfolios would yield
similar figures) for the withdrawal
amounts from 4 percent to 8 percent for
the 20- to 50-year periods. This figure illus-
trates the gap between Revisited (RV)
withdrawal rates, which have lower proba-
bility of failure rates relative to Fixed (F)
withdrawal rates, which is why the RV
columns are to the right of the F columns.
In reality, people withdraw dollar
amounts from their portfolios. Without
changing those dollar amounts (except for
increasing them for inflation), the with-
drawal rate is still constantly changing due
to the dynamic factor of fluctuating portfo-
lio market values. Advisors are able to
benchmark and compare their client’s cur-
rent withdrawal rate (current dollar with-
drawal amount divided by the current dis-
tribution portfolio market value) to Figures
3 and 6 to obtain an idea what the client’s
current withdrawal probability for success
or failure may be. This is especially impor-
tant during market declines where portfo-
lio values are less, which forces a higher
withdrawal rate from the portfolio.
A second dynamic factor is the effect of
aging where distribution periods are, in
fact, dynamically and continually shrink-
ing. An initial withdrawal rate for 35 years
remaining, then 34, 33, and so on, is quite
different from a sustainable withdrawal
rate when the retiree has 10 years remain-
ing. Withdrawal rates tend to be linear
when aligned for distribution periods from
20 to 40 years (ages 55 to 75) versus para-
bolic when aligned for periods under 20
years (ages 76 and older).
‘Safety’ of 4 Percent and Early Versus Later
Withdrawal Strategies
Distribution planning is not a “one size fits
all” exercise. Each client and retiree will
have different needs that are going to influ-
ence the sustainable real withdrawal rate
decision. Past research on adaptive strate-
gies has noted that 4 percent
is likely too conservative an
estimate for an initial with-
drawal rate, generally sug-
gesting a higher withdrawal
amount. Being able to take
higher withdrawals earlier
versus later has raised the
strategy of trying to reverse
this timing, or “smoothing”
withdrawal rates over the
entire distribution period.
Observe in the previous fig-
ures that,given similar proba-
bility of failure rates, a higher
withdrawal rate
correlateswith shorter distri-
bution periods, and vice
versa. Attempting to take a
higher withdrawal rate early in retirement
with the intention of changing to a lower
withdrawal rate later in retirement attempts
to reverse these findings. Considerations:
• It has been difficult to assess what rate
to use early on, unless the advisor has
relative probability of failure rates for
all the choices.
• Smoothing strategies require the
client to have the ability to cut expen-
ditures during poor markets. This is
difficult to explainunless the advisor
has relative probability of failures of
the client’s current withdrawal rate
(current annual withdrawal divided by
the current portfolio value).
• Higher initial withdrawal rates result
in still higher current withdrawal rates
even when the portfolio value declines
with poor markets (sequence risk).
• Portfolio value volatility accentuates
the sale of more shares. The higher
the smoothing rate over a sustainable
rate, the more the relative number of
shares are needed to be sold (negative
dollar cost averaging effect) versus the
non-smoothed rate.
• The negative dollar-cost averaging
effect has led to the strategy of placing
the first few years of distributions into
cash or more conservative portfolios/
buckets.
• Because the total value supporting dis-
tributions includes these conservative
buckets, this strategy is essentially
shifting the overall portfolio to one
more conservative.
• Figures 2 and 5 provide probabilities of
failure rates for different portfolio compo-
sitions for different withdrawal periods.
Sequence risk can be managed by review-
ing current withdrawal rates to ensure they
are still prudent given the relevant time
remaining. As time remaining is reduced by
client aging dynamics, the withdrawal rate
may increase over time. How to determine a
client’s time remaining is based on using a
common mortality-base age as discussed in
the white paper by the authors titled “In
Search of the Numbers.”
But each retiree can potentially incur
market declines at any time. Controlling
the risk of having to reduce a retiree’s with-
drawals is a function of setting the current
withdrawal rate lower, rather than higher,
at any given point. Benchmarking the cur-
rent withdrawal rate provides the ability to
assess the probability of failure over time.
A client can reduce the likelihood they
would need to reduce their withdrawals,
“Sequence risk can be managed by
reviewing current withdrawal rates to
ensure they are still prudent given the
relevant time remaining. As time
remaining is reduced by client aging
dynamics, the withdrawal rate may
increase over time.
”

Contributions
hence cut their expenses, by using a with-
drawal rate appropriate for the time
remaining as well as a lower current with-
drawal rate relative to other rates possible
for that time frame remaining. In other
words, higher rates are generally possible
for smaller distribution periods (such as 20
years) versus longer distribution periods
(such as 40 years).
Conclusion
Because it is impossible to predict with
certainty the exact path each of your
clients will take during retirement, an
adaptive approach should be used when
determining the appropriate withdrawal
amount from a distribution portfolio. Past
distribution research has been based pri-
marily on the assumption where a con-
stant, inflation-adjusted withdrawal is
taken from a portfolio for the length of the
distribution period, regardless of the
underlying portfolio. The static methodol-
ogy ignores the dynamic needs of clients,
market fluctuations, and client responses
to those fluctuations, where the ongoing
value provided by advisors who regularly
meet with clients to ensure the future suc-
cess of the distribution strategy rests with
an ability to benchmark the client’s proba-
bility of success or failure. Revisiting the
withdrawal can materially improve the
probability of success for a distribution
portfolio and, therefore, is an essential
component of any distribution plan.
Endnotes
1. Data definitions:
a. Intermediate-term bond: defined as
the return on the Moody’s Seasoned
Aaa Corporate Bond Yield, assuming a
ten-year duration. Data obtained from
the St. Louis Federal Reserve: http:
//research.stlouisfed.org/fred2/.
b. Cash: defined as the yield on the
three-month Treasury bill. Secondary
Market Rate data obtained from
Tradetools.com (1927-1933) and the
St. Louis Federal Reserve (1934-
2006): http://research.stlouisfed.
org/fred2/.
c. Domestic large blend equity: defined
as the return on the “Big Neutral”
portfolio based on the 2×3 portfolio
return information publicly available
on Kenneth French’s Web site: http://
mba.tuck.dartmouth.edu/pages/fac-
ulty/ken.french/data_library.html.
d. International equities: defined as the
return on the Global Financial Data
World ex-USA Return Index, data
obtained from Global Financial Data
from January 1927 to December 1969
and the return on the MSCI EAFE
Standard Core Net USD from January
1970 to December 2007.
Because pure historical data is used for-
this analysis, as is common among distri-
bution research, the authors would cau-
tion the reader that if future returns are
lower than historical returns, the actual
result of a distribution portfolio may be
materially different from what this
research suggests.
2. Data obtained from the Bureau of Labor
Statistics.
References
Bengen, William P. 2001. “Conserving
Client Portfolios During Retirement,
Part IV.” Journal of Financial Planning 14,
5 (May): 110–118.
Blanchett, David M. 2007. ”Dynamic Allo-
cation Strategies for Distribution Port-
folios: Determining the Optimal Distri-
bution Glide Path.” Journal of Financial
Planning 20, 12 (December): 68–81.
Brinson, Gary P., L. Randolph Hood, and
Gilbert L. Beebower. 1986. “Determi-
nants of Portfolio Performance.” Finan-
cial Analysts Journal 42, 4 (July/August):
39–44.
Cassaday, Stephan Q. 2006. “DIESEL: A
System for Generating Cash Flow
During Retirement.” Journal of Financial
Planning 19, 9 (September): 60–65.
Cooley, Phillip L., Carl M. Hubbard, and
Daniel T. Walz. 1998. “Retirement Sav-
ings: Choosing a Withdrawal Rate that
is Sustainable.” Journal of the American
Association of Individual Investors 20
(February): 16–21.
Guyton, Jonathan T. 2004. “Decision Rules
and Portfolio Management for Retirees:
Is the ‘Safe’ Initial Withdrawal Rate Too
Safe?” Journal of Financial Planning 17,
10 (October): 54–61.
Guyton, Jonathan T. and William J. Klinger.
2006. “Decision Rules and Maximum
Initial Withdrawal Rates.” Journal of
Financial Planning 19, 3 (March): 49–57.
Pye, Gordon B. 2000. “Sustainable Invest-
ment Withdrawals.” Journal of Portfolio
Management 26, 4 (Summer): 73–83.
Stout, R. Gene and John B. Mitchell. 2006.
“Dynamic Retirement Withdrawal Plan-
ning.” Financial Services Review 15, 2
(Summer): 117–131.
Tezel, Ahmet. 2004. “Sustainable Retire-
ment Withdrawals.” Journal of Financial
Planning 17, 7 (July): 52–57.
Tokat, Yesim, Nelson Wicas, and Francis
M. Kinniry. 2006. “The Asset Allocation
Debate: A Review and Reconciliation.”
Journal of Financial Planning 19, 10
(October): 52–61.
Watson Wyatt 2008. “Influences on Work-
ers’ Asset Allocations in Defined Contri-
bution Accounts.” www.watsonwyatt.
com/us/pubs/insider/showarticle.asp?Ar
ticleID=18489.
B L A N C H E T T | FR A N K

Contributions
/^^SRWf ( '#bV
S`QSbWZS E]àb W @SOZ EWbVRÒeOZ /[]cba Ob BOÙSb 2ObS 3R EVS
CaWU @SdWaWbWU

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
1]``Sa^]RWU `]POPWZWbg ]T 4OWZc`S T]` @SdWaWb AbÒbSUg
³#
,#³
, ³#
,#³
,³'#
,'#³
c[PSà aV]e O`S POaSR ] O [WZZW] ^]`bT]ZW]
# '
$# !'
% $
'!!
!
#!%%
! $#
%!
'
# '
$# !'
% $
'!!
!
$%
''$
'%
'
# '
$# !'
% $
'!!
'!
'$'!%
'!!$
'''#
'
# '
$# !'
% $
'
''%
'!
'
%''$
'
$#
%#$ '
'%'
%!%'
%'
#%##
$#
%
%!
$#
%#$ '
%%
%!
$'$
#'
$#
%
%!
$#
%#$ '
% %#
$'%%
$
$ %!
$#
%
%!
$#
$$'
$#!''
$
$# !
$#$''
$#$%
%
%!
%
$ %
#%'
#
'
###
#$$
$$$%
##
%
$$$'
##'$
##!
$ !
###
#$$
$$$%
$% %
$!%%
#$!
#!!#
##
'''
###
#$$
$$$%
$% %
#$
# %
#'
# %
#!$
##%
#$
$$$%
$% %
# !
%%
!
!!$''
!#!
!!

$!!
O
#$!
%#'!
#

!#!
!!

# '#$
#%%%
#!
$$
%!
!$#
!'
$$

# '#$
#%%%
# #
#% %
%%
#$%%
$#
'%
# %#
###!
#%''
!$

! $
'!'
!!%!
!% %
# '
O
O
'
#
!%#
!!$$$
!! %%
!% %
!
#%#
# %#

%$
!'
!
!'
''
!#!
$'
''$
%'$
$
#
''
#
$!!

#%%'
##!$$
!%%!
!#
$$!$
#

! ''
##!%
O
O
!$ %
!##%%
! $
!#'
'!%
!
!##
#'
'!
#
!%$$
!#'
!#!$
!$'#
! '

%'%
$%

!'
!'!
!$ '

$!
#%

#'%$
!##
! #
!#
!
!'
!!'#!
O
O
O
!# $'
! $!
'
%$$#
%#'
'
! %#!
!'
#$
! '
!% '
! %#
! $
! #
!#!!'
!%#%
!'!
!'
! %
!##$
!$
!% %
! #
#%'
'
%#
#
!'!
$'$$
#
! #
$ $
!$'
O
O
O
! '%
'''
$!#$
#''
#%!
%$#
'
!$%#
# $
!'!%
!$
'%
'
!%
! $#
!#
!!#
!
!'
!
!% %
!##
!%'
!%$
#

#!#%
!'!
!$
%'#
$!%
'##
!!%
O
O
O
!'
%
%
!
!%!'
#

!!#
#''
! #'
!$'
!#
'$
'$
! %
! !
!#!%$
!%%
!$ '
!$
!!#
!!$
!# !
!$ %
!' #

!##
Ca
/^^SRWf ( b
'#
dWaWbWU
S
WU @
b W
]à
SbWZS E
Q
S`

bV
OZ /[]
e
O
WbVR`
SOZ E
@ S 3R
b
USb 2O
O`
b B
BO
]cba O R EVS
'

%

#
$

e
Ca
O
WbVR`
Ò
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

%%
!
'
#
dWaWbWU

OZ
e
eO
OZ
S
WU @
#
'!!
##
#$ '

'
$#
! $#

%
#
'

%

%
!
%

%
'
$
%

'
$
'

#!
%

$# !

#
!
#

S`W]R G
b`WPcbW]
2Wa
%
'
! !$
%
'

'

##
''
!#!
! $
'
!
%
#

%
!
!!
%

!!
!
'
!

!!$

%

#

$
#$
#

#

%
#
$

%

$
!

!#

SOà
G
GS

#

$
!

!

O
!
'
!
!#
#
##

O

#
!!'
! ''
!

!$

!

!
!
%
## !

%

'
$$
#
!
%%
!
!$
!
#
#
O

$!%
#
##

O
%

!!
'

'
$ $
! #
'
%
#
$'$$
'

!
!

$

!'

'
$
%

#

$

$

$

$

$

$

$

$

'
!
! $#
#

$#

$
$
$#

#$ '
!
%

'
#

'
$
!

!
%

'!!
%%

$
%

'
$
'
%

%

%
'
''
%
$
%

$# !

#
%

'
O
'
O
#
##'$ #
%
%
%%
!!$$$

#$

!#!
#
!
!!
%
#

%
!
!!

%
!

!
!

O

!
#

'
#

##
!

!
#
'
#
%

'

#$

##
!
$
%

$$
$
$
#$
#
$
##
$$
$

%

$$
$
$
#$

!

!

'

#
%%
$
! '
!

#'
$$#

! $!
'
O
O
O
'
#
%
!

'

%

%

'

'
! $
# $
!
%

O
O

O
O

O

#

#
##
!

!
%
%
!
'
%
#

!
$

##
!

!$

%
##!

!! #
!
O
O
!'
O

#''

!
'

#

#
$#
!
%
!
%

#$
%

''

O

O

O
%
$

'
%
%
#
$!
'
%
'

'

%

#
$

$

$

$

$

$

$

$

$
$

$

#
%
'!!

'

!
#

!!$

$

$#

%%

!
#
#$ '
%

%
$
$'
%

%

$
%

'
$
'
%

%

''
'
'
!
'$'
'!
%

$# !

#
%
'
%
##

''
#

% #$
$

#
!$# !
!'
#
#$

''
!
'
$
!#

$$
!'
'
!
!
%

$$

%

#

%
#
'
#

#
%

#$!
!
#

%%

%
#

'
#
$$
$
$
#$
#
'

#
!!
#
%%
!
$
%

%
$
$$
$
!
!
#
!
!
$ !
'
# !

!
!
#%
'
! #
$
#
'
! '
$
!'
'
!
%
%
!

#!!
!
!
#!$
%
!
'
%
!
$$
#
#
!
'
'
%

! '

'
!$

!
#
!

%
!

'

#
#

#
$
! !
#''
'
!!#
'
#
$
'$

%
%
#!
!
#
#

!
$#
'

!

$
!
$
! #
%
$
#
!
!
%

!
'
%

'
!
!
'

#
%
$

'
$
%

#

$
'
'
''$
%
!
'
#
!

%
!
''
!
''
%

%

$#$
$#$
$#
$
$#!
$
$$'
'
$
'
%

%

'

'
'

'
%

$# !

#

%
'$
'$

$
##!$$
#$
%
!$

'

%%
#
$!!
$#
''
#
'
%%
#
!
#

%
%

#
'

'
%
#
###
%

#
%
'

#$

%

'
%
#
#
$
%%

%
#
%

%
$
$$
$
%

##
#
%

#
#
%

#
#
%

%
$

!

$
$
!
!

#
'
!
!

'

!
#
#
'
'
! #
%
!$
##$
%
'
$
%

%
#

!

%
!
!$ '
!
!

!

!
$%
%
#'
$

%
#
$

!'

'
!

$
!##
#
!!$
!$
%
!$ '

%
!!#
# !
%

'
!
#
!$
$
!
'
##
%

'

%
!
!
#
#!

%

%
!
!
%

!

e O
c[PSà aV]
Sa^
]``
1
,#³
³#
]Z
S POaSR ] O [WZZW] ^]`bT
O`
S T
OWZc`
]POPWZWbg ]T 4
^]RWU `
,#³
, ³#

ZW]
SUg
b
O
b`
dWaWb A
S
]` @
T

#³
,'
,³'

#
'

Contributions
aa
/^^SRWf ( 'bV
S`QSbWZS E]àb W @SOZ EWbVRÒeOZ /[]cba Ob BOÙSb 2ObS 3R EVS
CaWU @SdWaWbWU

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
³#
,#³
, ³#
,#³
,³'#
,'#³
# '
$# !'
% $
'!!
!
%!
'%
$!$
' !#$
# '
$# !'
% $
'!!
!
%!
!!
!!'
''
# '
$# !'
% $
'!!
!
%
$$%
!$
'$'
# '
$# !'
% $
'!!
'%% '
'
'#
'%%'
'%%
$#
%#$ '
$
%%%$
%$$
$!$
$#
%
%!
$#
%#$ '
% '
%'!
%#
$'%
$#
%
%!
$#
%#$ '
%'!!'
%$!
% '
%!
$%
%
%!
$#
%! $
%!$$
% ##
% $'
%'
%!$!
%$''
%#
%
$#'!
$!%
#$%'$
'
###
#$$
$$$%
##
%
$$%'$
$ ''
# #
# $
###
#$$
$$$%
$% %
$''!$
$!%$
$%
#%#%
#%!
# '
##%
$$$%
$% %
$#!
$
$!
$!!
$#
$!$
$#$$
$%#
%!
#$%
# '
'
!' '%
!#!
!!
!#
$!!
O
$!#'
#!'%!
#%
$#
##
!!

# '#$
#%$
$ '!
#' #
#'!
#'
#!!
# #
#!%
#$!$
#'%#
#'#
##$
# '
#!!
#!'!$
#%!!
$
$ ##%
$$!#!
%$
##%
!%
! !
!!%!
!% %
# '
O
O
#%%
%$#
'!$
!'#$
!%%
!'##
'%
#$%
# $$
#!
'% '
%
$'
#$
%%
#%
# '
#$!!
# $%
#!
'!'
!
##$
# %
#$%
#'%
$!$$
%
!#
!
$'

!!''
##!%
O
O
#!!
%
! $
!#
!#!%
!%!
!'#
#!
# #%
'#
#'!'
#
$
$!
%$
$$'
' '
#!%
'''$
%

$##'
'$
'$
#!
##
#'!%
!'!
!!%$'
$%
!%!
'!%
!$$
O
O
O
$ %
!%
!%%%
! ! !
! '
!#
!%$'
!'$
#!$!
%$%
$$
#!
!%'
!''#
$%

#%$
'$
'
'%
$#
#'
$%!
%!
''$
# !#
#$%'
!##$
!$'
!$!%
'

!'!
O
O
O
!'
!$!
! %
!'$
! $'
! %%
!!'
!'!!
##!
!'
!' '
!$#
!$% !
!%$'$
!'
!$
%
%
#
#
'#
##
$!
$
'$
#
#!'
!! #'
%!$
%

$!
!!
O
O
O
!%
!!$#
!
!!

!
!' %
!!
%%$#
#
$
!'
!#!%
!$
!%$
!' #
%
##$
$
$$#'
%
!#$
%%
#!$

#
#
Ca
/^^SRWf ( '
dWaWbWU
S
WU @
b W
]à
SbWZS E
Q
S`

bV a
OZ /[]
aa
e
O
WbVR`
SOZ E
@ S 3R
b
USb 2O
O`
b B
BO
]cba O R EVS
'

%

#
$

e
Ca
O
WbVR`
Ò
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

!
'!!
!
#
dWaWbWU
#

#$ '
$

$$

$#

OZ
e
eO
OZ
S
WU @
$
!
$
%

%

%%

$
$
%

'
$
'
%
'
%
%

$# !

#
S`W]R G
##
%
$
! %
'
'
'$ %
'
!#!
!
!
%
b`WPcbW]
2Wa
!
#

#
#

%
!
!!
%

!!

!
%
!
##
%

!

'
'
!
'

#
%
#$

$
#$
#

%
#$
!
$
'
$#

%

!
#

O
'
%

$

!

!

SOà
G
GS

!#
!$$

%
!
%
'
!
!

''
!
'

%
!
$
!
$
%
!!
!#

'
!

%
## !

!!
%

$
!
%

!$
'
#
O

$!
#

O
!!
!

'
%

%
%

'
!! #
##$

'

!
!$
$

'
$
%

#

$

$

$

$

$

$

$

$

'

#$
$!$

#

!

$#

#
!!
$#

%
!
'
#$ '
!
%

'
$
%

!
'
%

'!!
%

$
%

'
$
'
%

%

! !

%

%

$# !

#
!
'
%
'
$ ''
%%
O
'
O
!$
%
##
$#
!!
#$

#$
$#
%
##
%
%
# $
'

'
!
%
!
'
!
'

%

!
#
!#'

O

!
#
#
'
#

##
$
#
%

#
%
!'
#
'$
$

##
!
$
%
!

$$
$
$
#$
#
#

#
%
$$

%

$$
$
$
#$
!!
'
!
'
O
O
O
!
!
!$
!
!
$
'

! '
! ! !
%%%
%
!
%
'
!
!
$
%
!

#
!
!

!
! $
!
$

!!

O

O

#
#
'
!
%

%
!
#
%
#!
!
%
#
!

$
%

#
#

O

O
%
##!

%
!
O
O
O
'$
!!$#

'
!!

!
!
!'

!

!

%
!

O

O

O
!
'
!'
%%
!
$

%

$!
'

'

%

#
$

$

$

$

$

$

$

$

$
$

$

!$
%
!
'
'
!

#$ '
'
'!! $!
%

!
#
%

%
$
%

'
%

%

!!
'
%

$
%

'
$
'
%

%

$$

%

$# !

#

'
!!

%
%
$$

% #$
'

# '
!
'
'
$
#$!$

'
%
#
#
%

%

#
#
!!
$

%

!
%

'

#

!
#
!
#
$
%
#$
%
!
#
#

#

#
#%
'
#
#'
$
$ '
!$
%
#

'
#
$$
$
##
%
#
%
#
%
$

%
!
$
'
'
$
%

%
$
$$
$

#

'
' '
!$
! !
!
'
$

!
#
'
#!
!

!
!
$
%
#
'

'
%
$

$
!''

%
!
$

#
$$
'
%
$
%

#!$
%

'
!
!
'

$$
%
%

$
$
'
!
#'
#

'

#
#
$
#

%
$
#
$#
#
!
$

!$

$

%
'
!
!$
%
!

'$
%
#!
!
!

'
!

'

'
%

%

!
!
!

'
$
%
%
$
$#
'
!
##
!
'

'
$
%

#

$

'#
'
'

'
#

#
!$!
##

''
!
!$$
'
! $
%

$
%

%

%

'
$
%

%

'!!
%

$
%

'
$
'
%

%

%
'
%

%
'
'
%

%
'
%

$# !

#
'$
!
#
!$
!!
#$!!
%

#!
!$$
$$
%
%

#
!!
#
!
!$
!
'
#$

$
!

'
#
%

#$

%

#
#
!
'

'
$

#
#
#
$$!
$ ##
$

%
#
!'
#
!
#
#
!
##$

'
#
$#!
%

'
#
%

%
$
$#
$

$
$

$
$
%

%
$

%
%

#

#
$

$
!
%
%

#
'
#
'$
!
!
#'
#
%

$
%

#$
!%
!

#
'

%

$
%
$
$
'
$

'

'

''$
'

'
#
$
##
!
#
'

'

$##

'
%

'

%
#!

#$
'
##$
$
#!$
%
%%

#
'

#

$

!
!
#

#
$$#

!
#

'

$

$
#
'

#

%
e O
c[PSà aV]
Sa^
]``
1
]Z
O`
,#³
³#
S T
OWZc`
]POPWZWbg ]T 4
^]RWU `

ZW]
,#³
, ³#

SUg
b
O
b`
dWaWb A
S
]` @
T

#³
,'
,³'

#

#
'

Contributions
/^^SRWf !( bV
S`QSbWZS E]àb W # @SOZ EWbVRÒeOZ /[]cba Ob BOÙSb 2ObS 3R EVS
CaWU @SdWaWbWU

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
³#
,#³
, ³#
,#³
,³'#
,'#³
# '
$# !'
% $
'!!
!
%!
#%'
!#
$!
# '
$# !'
% $
'!!
!
%!
#%'
'!
%!%
# '
$# !'
% $
'!!
!
#'
'$#
'
' #
# '
$# !'
% $
'!!
!
%'
%%!
'%
''
$#
%#$ '
'%##
###
'%!
% %
$#
%
%!
$#
%#$ '
'%##
$'%
'%
%!'
%'#
%
%!
$#
%#$ '
%!
#$
'%

!!'
''
%!
$#
%#$ '
'
!
%#
!%'
!#
$%%%
'$
%
%'
$#%$
$ !!
#'
###
#$$
$$$%
##
%
%##
%!%
$#
$!!
$'
#
$$$%
$% %
%
%
%
$$
$$'!
$%% '
$'$
%$!!
%#
%
%'#
%%
%%'%
%
%!%
%#%'
%#
!!
$! %
#%
#!!
# #
!%#
!!
#$$
$!!
O
%!#
$ !
#'
#
#
# %'
#%'
##
#'%
%!'
$####
$$
$!'
#'#!
$!'
$ %%
$$'
%$$
%%
$#%%
$!!
$#
$!'
$%'
$'!'
% !
%%!
# '!
''
!%#
!%
!!
!% %
# '
O
O
$!!
##$!
#
$'$$
#%%!
%
#!!
#$%
###$
$#%
$#%
#%!
#!'
#!#%
#$#
#' !
$!''
$$'
$# #
$!#
$
#'%'
$ %!$
$!'
$#'$
% !%
%
$
! !
!$!'%
!'$
!$
!#%
##!%
O
O
#!'%
''$
$%
#
%#

%#!!
# !
#'!
$'!
#
#!
# %
# #'
#!!$
##'
# '%
$%$
$#'!
$!
$!$
#'#
$$
$ '
$ $
$$'$
%%
!#
!
!#
%
'$$$
!#$
O
O
O
#%
%%$%
%
!'%%
!'
%#
%
#'!
#!'$
$#!$
#'#'
# !#$
'%
$!
#!
#!!%%
#'
#'!#
$#!
#''!
$ %
#''$
#'#
$%
$ %!
$#%
$'%#
!'$!
!# #
!
#
%$%
!#!'
O
O
O
#!
# $
%'
!! '
!$%'
!' $
!
'$!
#'%%!
#%
# ##

%#
'
'
#
##$
#'$
$'!
$$
$#'
$!
$
$#$
$! %%
$''
$!
!%%!!
!
##!
!'
#%$
!!!
O
O
O
!$

!%
!#!'
!% '
!$'%%
!%%$%
#%
# $
$#'
#$'
#!!!'
$#
%!
#%
'
#!'
##$$%
$
$#%!
$'
$%!
$##$
$$
$!
$#''%
$!$%
Ca
/^^SRWf !(
dWaWbWU
S
WU @
b W #
]à
SbWZS E
Q
S`

bV
OZ /[]
e
O
WbVR`
SOZ E
# @ S 3R E
b
USb 2O
O`
b B
BO
cba O EVS
'

%

#
$

e
Ca
O
WbVR`
Ò
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

!
'!!
!
##
#
dWaWbWU
#
#$ '
%
###
!

$#
'

OZ
e
eO
OZ
S
WU @
%

%
'

%

'
$
%

'
$
'
%
#
%
%

$# !

#
'
!!
S`W]R G

##
$
% !
'
#
'
'
$ !! #
!!
%
b`WPcbW]
2Wa
!
#

#
#$$

%
!
!!

%
!
%
!
'

'

#

#
%
!
#
#

%
#

!
$

$
#$
#
#
%
$#

%

%

!
'

O
!'$

!
#

!

SOà
G
GS

!#
#$
$$$

!#
!
!
!#

!
'
!$
%

%
! !
$
%
## !

%

#
!
%

'
!$!

!!
#
O
!
!'
!

#

O
!!
#!'
$
%
#
%

#
##
!
!
# #
%

%
!
$!

!
$
%
!
'

'
$
%

#

$

$

$

$

$

$

$

$

!
!
$#
#
'
#

#$ '

%
%
'
#
'
!
##
!
%

'

%

!
%

'
!
'
$
'!!
%

'
$
%

'
$
'
%

%

'

%
#
%
%

$# !

#
$

!
%
#
!!
O
'
O
##$!
%

%
%
$'$$

!

' !!
'
%
$
#
#

%%
#
#

$

O

!
#
#$$
##
%
#
%

#
'
#
!!
#
$#

'
#
$ !

!#
%

##
!
$
%

$$
$
#
$
$
!
%

##
%

%

$$
$
$
#$

O
O
O
#
'
!
'$

!
# '!
%
!
#

%

!

!!
%

'
!
#
%%

'
!
#
%

$
%

%

%

#
%

O

O

#
#
%

!

%
%

!
$
%
$
'

!
!'
#
!

O

O
%
##!

%
O
O

O

! '

'
'
'
%%
%
#
!
$
%

%
!
!
!$'
$
%
!
#!
!
%
!
'

# $
!$
!

O

O

O
'$

'

%
$
%

#

'

%

#
$

$

$

$

$

$

$

$

$
$

$

'
!
!%
!
!
#
#$ '
'!!

''

'

!

!!

%
'
#$
%

$
%

'
$
'
%

%

'

$#
'
#

%

$# !

#
%

'

!!
'
$
!
###$
%
%
%
!
' !'
%
#$#
$$'
%%
''
!
!
$
'
#
'
!#
#
!
#
%
#
$
#

$
$####

$#
'
%
$
#
$
!
$
#
'
#
!
$
$$
$

!
%

%

'
#

##
$

%

'
$
%

%
$
'
$$

%

%

%

%

%
$
$$
$
##

'
#
#

%

'

'
#
%%

#!
!

#$
#
#'
!$
!'$
#'!
#
%
!!
#
'
!!$
$

%

'

%
!
#

'
#

#
$
!
#
!
!
# '
'
##
!
#
#
#
%

#
'
!
#
#
'
$
%

'
#
$
#
!'
$$
%

$
'
'
'
!
$#
#
$
'

#

%

'
#
!!!
#

#$
##
#

$
#
!
%
#
!
#

'

%

%

%%

'
'$

'
$
%

#

$

!
%
!
'
#
#
#$ '
#
$
!
'
%%%
!#
!
'
'
$

%
!
%

'!!

$
%

'
$
'
%

'
'
'
%

%

%

%

$# !

#
'

#
%
%
'
$# #

' $#'$

'
%
$$

'
'
!$
'

%

!
!

%

!
'
!
$
%
$
%

'
#
$
$!!
#
$!

$$
%%

%

!

'
$

%
$
$!
$#
%%
$#
%

%

%

!!
%

%
#
%

%!

%

'
%%

%

%

#
'

%

%

#
%

$
$
$
# '
$
$
$

'
$
%
$
$
$
#
%
!

#
'$
%
!
$
$#!
#
%

'
$
%

$#
$$'$
%
$
$ $
%

$
'
'
#
$
'
#
#
$
!$
'
'
#
!
!
#
!
'
#
$

%

$

$

'
#
$

$
'
$#
%
$

%

##$
!
!
$
%
'
%
$!$
$!
$#''
''
!
$
%%
$
$
#$
$

%

$
!

$
#'
%
$#
$
!
##$$
$
!

$

$

'
'
e O
c[PSà aV]
Sa^
]``
1
]Z
O`
,#³
³#
S T
OWZc`
]POPWZWbg ]T 4
^]RWU `

ZW]
,#³
, ³#

SUg
b
O
b`
dWaWb A
S
]` @
T

#³
,'
,³'

#

#
'

Contributions
/^^SRWf ( #bV
S`QSbWZS ;SRWO @SOZ EWbVRÒeOZ /[]cba Ob BOÙSb 2ObS 3R EVS
CaWU @SdWaWbWU

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
³#
,#³
, ³#
,#³
,³'#
,'#³
# '
$# !'
% $
'!!
!
%!
%%
'%$
#'
# '
$# !'
% $
'!!
!
%!
#'%
$#'
!#$
# '
$# !'
% $
'!!
!
%!
%%
% '
$%'
# '
$# !'
% $
'!!
!
%!
%%
#$$
' #
$#
%#$ '
'%##
'
'$
#
%!$!$
%
%!
$#
%#$ '
'%##
''%
'$'$
'!
'#%
'%
'
$#
%#$ '
'%##
#
''!
''$$$
''!%

'!
$#
%#$ '
'%##
'%$'

'
$#
$!
!
%
!#
%# %
%
$$
#'!#!
#$
$! $#
##
%
%$%#
#!%$
$
%%'%#
%%!
%%%
%'##
!$$
%
%$%#
%$'
$%
%'
$'
#%
'#'
'#%
%
#
'#$
'#'!
'!!%
'#%
'#%$
''$'
!'
%'%
$%!
$!$'
#$$%'
# !!!
##!!
#!%$
$!!
O
!
%'%'
%#!%
%#%
$'%
$'
%%
%$
%$%
!
$$
''$
$

$
!$%
$##
'%
!
%'!
%
'
%$%
' !
''%
'$$
'''
$!#
#'%$
##!
%'!$
#'
$' !
# '
O
O
$$
% #
$'
$! %
$
$!''
$$!%
$%!$$
%
!
#%
!!
%$'!$
%$#
%'
#$
'#%
'!
!'
' $
''%
$
'%$
'!
'!'#
''
$#
#$##
# $$
$#'!
#%
'%
!!
##!%
O
O
%#
$'$!!
$##$
#'!'
#'
#'#!
$#
$%#'
%#
'% '
#%
%''!
%'!
%$%
%%!%

!
'#
'!'#
'#%
'!
'
'!#!
' %#
'%
'%#

#!'!

'#
!%$
!'%
!
O
O
O
% %#$
$##
$ !#
#'
##
#'%!
$'
$ %%
$%
'
'#'
$
%%$
%!%%
%%'$
'
!#
!
$
'$'$
'
'%%!
'$!
'
'%$#'
'''#$
#$'
''
#$$
!!
!$
!%!
!$#$
O
O
O
%#
$% #!
$#
#%!'
#!!
##%
#%##
$#%'
$'%
'#!
#

%%##
%%%%
%%$
!
#$
%
'!
$

!%'
'$
!#!
!'!
#
'
'
#
!##
!!
!! #
!!'
O
O
O
%!$
$# %
$% !
#
# %
## #'
#$ #
#'!
#% %

'!#
'''#
%'$'

%

$$%
%
$''
'
!
%#$
%!$'
$
'!

#!
Ca
/^^SRWf ( #
dWaWbWU
S
WU @
S
SbWZS ;SRWO @
Q
S`

bV
OZ /[]cba
e
O
WbVR`
SOZ E S 3R EVS
b
USb 2O
O`
b B
BO
O
'

%

#
$

e
Ca
O
WbVR`
Ò
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

!
'!!
!
#
dWaWbWU
#

#
##
!$

$
%%

OZ
e
eO
OZ
S
WU @
#$ '
! $
%

'
'
%

'
$
%

'
$
'

%
%

$# !

#
!#
'
S`W]R G
!
!

%
$$
$
'
!
'
!!!
!$
!
'
b`WPcbW]
2Wa
!
#

#
$
$'
!!
#
'
%

#
#
%

'
#
!#
$
%
!
#
#$
#
#
#
%
#$$
!$
$

%
$
'

%

#
!
'
#
%

%
#
%

%

#
'

O
!$
%
$

!
! !
!
SOà
G
GS

!#
!

$
#

!'!

!!
%
!'

%
!
'

!

$$
#
#$##
%
## !

!
%
'

#

$
'
$#
# #
'
#
O

##
#
#

O
'
!!
$#$

!!
!
!!
$
!
!
#$$

'

%
!
'

'
$
%

#

$

$

$

$

$

$

$

$

'

#'
$
%%
#

#$ '
##

'$'$
!$
!
#'%
!
%
!

%
'

#
'
'
!
%

''
'!!
%

'
$
%

'
$
'
%

%

!$
%

$ #
%
%

$# !

#
%

'

$
O
'
O
#
$
##
!''
%

$
%
$
%
!$$

%
$
%
$$!
$
!
$

'
$
%

%

%'
$

!

O

!
#
$

%

%%

'
$

$'
#
#

%

$
#!
%

%'

#

##
!
$
! $#
%
!
#
#$
'
%

%

%
%
!

%
%
%
'
%%

%
#!
%
$
%

%

$
#
$
%
O
O
O
$
##
#
!!
$
!
%%
!
#

'
$ !#
$##

#$
$
'
'

$
$#
%

'
#
!
##
'
#
!'
#$
%

%

#

O

O

#
%
$
$

#
'
#
'
#

'
#
%
$#

$
'
$
#

%

$

O

O
%
##!

#$ #
!
O
O
!$
O
%
'
!
!!
#

'
'
#
'
##
## #

%

#
%

$
#

$#
!
%

#

O

O

O
%
#
%
#
%
#
!
%
#

#
%
$
%

'

%

#
$

$

$

$

$

$

$

$

$
$

$

'
!
!
#$
'

#
#$ '
##
'!!
!

%%

#

$$$

%
!
''
''

''

%

'
$
%

'
$
'

'

%

'

%

%
%

$# !

#
!
$#
'
'
#%
'

!
## $

' !!
$ !$
$#

!$$
$##
%
%
#$
#
'

'
%
$
%

$'
%

''$
%
#
$$
!
%

%
$
!$

#

%
%$

$$

%

#
'
$

%

%
$
$
%

%
$
%

%

!
'
%

%%

'
#
$

%
%
#
'

'$

%%

$
'
'

$

%%
!

!

%%

!
%

$
%%

!
!
'#

'
'

%
$
#
$
'
%

$
%
!
%%

%
%$

'

%

$
'
'
%

!
%
#
%

'

%

#
%
$
$
%

%

!

#

'
$$
#$

!
%

$

%%
$
'
%

##
''
'

!#
'

!

%
#

'
#
'
%

%%
%%

#
#
%

'
%
#

'
$
%

#

$
#
%%
$$
!
!
'
#

!
'

#
!
'!
##

'
#$ '
$
$

$
%
'
'!!
%

'
$
%

'
$
'

'
#

%
%

$# !

#
%
$

!'
$
'
%
!
!
'

$
#
$#
'
!

!

$
$
!
%

'''

'
'$$
'
!
'
''%

'
%

'

'

'
%

'
!
!
'

'
%
$
%

'

#$
'
%

#

%

'

$
''
%
#
'
#
'
%
!!
'
'
'#
'

%

%
#
'

'

!
#

'
#

'
#$

'
'
'$!
!

'$'$
$

$
#

'
''
#
$#
%
'
#
!
%

%
'

'
!
%

#
!
#

#

%
'
%
'
%

'
!#
'
'
'
%

#
'
'
!
'
'

!
!
#$
'

$''

'
#!
'

'

!

$
!
!$
%

'$
%

'
!

$
'!

%

#
!'
#
!

%
!

%

e O
c[PSà aV]
Sa^
]``
1
]Z
O`
,#³
³#
S T
OWZc`
]POPWZWbg ]T 4
^]RWU `

ZW]
,#³
, ³#

SUg
b
O
b`
dWaWb A
S
]` @
T

#³
,'
,³'

#

#
'

/^^SRWf #( bV
S`QSbWZS 0Sab W # @SOZ EWbVRÒeOZ /[]cba Ob BOÙSb 2ObS 3R EVS
CaWU @SdWaWbWU

$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$
$

]`bT]ZW]
/ZZ]QObW]

#
$
%

'

#
$
%

'

#
$
%

'

#
$
%

'

7WbWOZ
EWbVRÒeOZ # # ! !# # #
³#
,#³
, ³#
,#³
,³'#
,'#³
# '
$# !'
% $
'!!
!
%!
!%%
!
' #
# '
$# !'
% $
'!!
!
%!
!%%
!#$
!% #
# '
$# !'
% $
'!!
!
%!
!%%
!#$
!$!
# '
$# !'
% $
'!!
!
%!
!%%
!#$
!$!
$#
%#$ '
'%##
'#$
%%
''$%
$
#%
%
$#
%#$ '
'%##
#
'!

$'%
#%
' '
$#
%#$ '
'%##
#
'$
%%
$ '
'$$
' $
$#
%#$ '
'%##
#
!'#
%'#!
$$$
###
!!
%
%$%#
$!%
$$
%$%$!
% ! %
%!
$##
##
%
%$%#
'$
'%%'
'''
'!$#
'#$#%
'#

%
%$%#
##
$
$'
%$
#
!!'
%'!
%
%$%#
$
%$!
!
%! %
'!'
!%
!'#
!
%'!
%$'
$'
$%$
$!'
$#%#
$!!
O
!
'#
'! %$
'
%$
$%
' '
''
' !'
!
'$%'
!#$$

!
#%
$$%

%
!
'$%'
%$#
%'
!#'
##
#%
%$
' #
%$
%!##
$##
#''
#%'
#%$
# '
O
O
' $!
'!#$#
'!!
!$
!
%$%
#'#$
$
' $
' $!
$
$'
%
!'
##
%%%
!
''%#
' $!
'!%
#!
%%%'
#!%
!%!
$ %
! '#
!$$'
$
$'
#
#!#
#! '
$%
##!%
O
O
'%'
' !
%#%
##
%' %
%'!$
$$$
'$
''$
#%

'##%
'%#
'# !
'##
%%
#!%'
#'
!
!
#$$
$$!%
!!
!'%
!$#
!%
#!
$$ $
#'%
#!
!$
'#!
$!#
O
O
O
''$%
'!

#%
%%!$%
%!'
##
$ $
%%#

!!!
% $
!
$
#!
$%!
%%$
'#
#$'
' '
!%$
!%$%

!!'
!$#
$
##'$
$ $%
#%!$%
#$ %
#$
!!

O
O
O
$''
'#$
#!!

%

%$!
''
%
$'
!%$
!'
%'
$!
%
!''
!$
'%'
!#
!' $
%
#$!
#!#!
#
#%
$!
$#!
$
#%
%$
!
%
!#%#
O
O
O
%'!
'$!$
'#$
'%
%'!%
%
''%
'
$%
%
!''%
!%$
%'
$!#
%!
%!
!#
!
#%
$'$%
$ $
$#
$$
$%
$''%
%$#
!#
Ca
/^^SRWf #(
dWaWbWU
S
WU @
b W #
SbWZS 0Sa
Q
S`

bV
OZ /[]c
e
O
WbVR`
SOZ E
@ S 3R E
b
USb 2O
O`
b B
BO
ba O EVS
'

%

#
$

e
Ca
O
WbVR`
Ò
E
7WbWO
bW]

/ZZ]QO
]ZW]
]`bT

!
'!!
!
#
dWaWbWU
#
#$ '
%

##
$
#$

%%

OZ
e
eO
OZ
S
WU @

''$
%%

'
%

'
$
%

'
$
'

!
%
%

$# !

#
#
S`W]R G
!
'
%
%
!
!
##
$
$$
$
'
'
'
b`WPcbW]
2Wa
!
#

#
#

%
#
%
#

'
#
$##
!
%

$
%

%
$#
!
'
!
$
%
$
'
$
$
%

'
%

%

!
%

%$
%$

!
$
%
$
%

%

#
$

O

#
#

!#
'

SOà
G
GS

!#
#
!
!$
!

$

!
$
%
#
'

! '
#
#
%

'
#
'
$$
$
%
## !

$
$
#
#
'
$

#
O
#
!
$

#

O
%
#
!

%

!!
#$
%

%
%
#
%

$
%

$

!$
%
#
$

'
$
%

#

$

$

$

$

$

$

$

$

$

#
$
!
%%
#

#$ '

#
!

%
%%
!
##
%
%
#
'
$

'
!
'!!
%

'
$
%

'
$
'

%
#

#
!

!
%
%

$# !

#
'

!
!
O
'
O
!#$#
'$ !
' #
%
!$
%
' '
'
!
$
' #$

$
#'
$
%

!
'

$
'
#
' $
!

O

!
#
#
'

%
$
!$#
%

''

'
%
!
'
'
#

##
!
$
$##
%
$#
!
#$#
'
'
'
%

%
'
%
$
%

%

%

O
O
O
#

!

%
$ $
##
'

%
%
'
'!
%

$
!
%

!$
!$
%%

%
#
##

$

''
'

O

O

$
'

$$
$
'
%

%
'
%

%
#
%

!

'
#
%
'
!

O

O
%
##!

%

O
O
!
O
'$!$

#$

%
'

'
''
!
%

!
'
%

%
'
'
!!
#$
'
%

''

O

O

O
'

$
%

%
#
$

'

%

#
$

$

$

$

$

$

$

$

$
$

$

$
!
'!!
!
#
$
'
#
#$ '
'$
$$
%%
##
#

$ '

%

'

%

%

'
$
%

'
$
'

'
%
#
#
!

!
%
%

$# !

#

%
!
#
!
'
$
#
'
$
!
'# ''

## $'
%
##
%
!'
%%%

!

!

!#$$
$
'
' $
!
'

$
$$
%
#

$

%
'$
#

'
!
'

!!

%

$

%
$
%

%

!

!

!
%
$
!
!
$

$
!!!

#
$ $
%

%
'
%
$
%%
#
##

!
#
%
%

%

%

%
#!

%
'
#
'

%

'
%
##
'

%
#
!
''$

$
'

!
!

%

%

'
#
!
!$
%

''
%

%
#
$!
!
%

'
$
%
!
'
!''
$

$'

%
$

'

'
!
$
%
!

%
!
%

'

'
$
%

#

$
!
%%
$
!
!
!
#
#$ '
!!

#

$$$

###
$
##

#
!
#
'
%
!'
'!!
%

'
$
%

'
$
'

'
!$
#
!

!
%
%

$# !

#
!$
#
'
%
!
#
!
!
%
%

'
! '#
'
!
##

'
'
'
%
!
#!
!$$
#
$
$
%
%
!
!
#
%%

%
$#
!
'
' $
!
%
'
'
'
%

#
!#

%

%

$
%
'$
#

%
'
!
%
!
!

'
!
%
$
%

%
$
%

%

'

%
$
#
!
%
#
!
#
$
%
'
'
#
#
%

! #
##'$
$
!$#
'

%
$
'
!
'
#
!

$#
!!
%

!
$
%
!
%
%

!
#$$
'
!
#$
!
'
'
#
'
%
!
!
!'
!
!
$$
!

#
#
!#
%
$#
%

#

!
$$

$
%
!
'
'
$

%
$

!
$!
$ $

'$
$
$
%
#
'
$#!

$
%
#
#
!#

%

!'
#
%

'
e O
c[PSà aV]
Sa^
]``
1
]Z
O`
,#³
³#
S T
OWZc`
]POPWZWbg ]T 4
^]RWU `

ZW]
,#³
, ³#

SUg
b
O
b`
dWaWb A
S
]` @
T

#³
,'
,³'

#

#
'
Contributions

2a A Dynamic and Adaptive Approach to Distribution Planning and Monitoring JFP Apr 2009.pdf

More Related Content

Similar to 2a A Dynamic and Adaptive Approach to Distribution Planning and Monitoring JFP Apr 2009.pdf (20)

More from Better Financial Education (15)

Recently uploaded (20)

2a A Dynamic and Adaptive Approach to Distribution Planning and Monitoring JFP Apr 2009.pdf