Volatility and the timing of cash flows make a
bigger difference on long-term returns.
Portfolio optimization has been the cornerstone of
financial planning for decades. That some 94% of an investor's success
comes from the asset allocation decision is the first article of faith
for most in the financial planning profession. The financial press has
served gallons of this mantra to individual investors, convincing them
that it is as precious as blood. But, there is something far more
important to successful financial planning.
Unlike institutional investors, whose context and needs were used to
prove the mechanics of Modern Portfolio Theory (MPT), individual
investors are primarily loss-averse, rather than "risk"-averse (MPT
tends to view risk as synonymous with volatility). In addition, whereas
institutions treat risk and reward as having an inverse relationship,
according to Nobel Laureate Daniel Kahneman, individual investors are
more likely to take on greater risk to avoid recognizing losses than
they are to pursue gains. In particular, Kahneman points to the
reluctance of individual investors to dispose of investments with
unrealized losses and their propensity to point with pride to their
realized and unrealized gains.
When viewed in the context of MPT, which typically
ignores cash flows in and out of a portfolio (and collateral effects
such as income taxes), the individual investor's propensities are
generally self-defeating. For example, it is hard to convince many
people to sell off their "gainers" and replace them with losers.
However, they are more easily motivated to sell "gainers" if doing so
allows them to accelerate spending on short-term goals when markets are
up or to defer spending on short-term goals when markets are
underperforming. As we shall see, these behaviors have a far greater
impact on successful financial planning than techniques such as asset
allocation.
Focus On Danger Zones
Let's examine the tale of three portfolios. All of
them have the same asset allocation, the same rebalancing, the same
cash flows, the same goals, the same tax rate; in fact, they are the
same in every way except for their outcomes-and the historical data
series that were used to project future returns.
Each simulation uses a 50-year data series
consisting of historical returns for asset classes for the following
sequences:
The first simulation uses data from the years 1960 to 2005, then from 1927 to 1930.
The second simulation uses data from the years 1980 to 2005, then 1927 to 1950.
The third simulation uses data from the years 1930 to 1979.
The performance statistics shown in Table 1 that
were produced by these data series cannot account for most of the
capital accumulation differences illustrated in Figure 1.
Based on constant mean return differentials between
the portfolios, the second simulation should have produced a value
approximately twice as large as the other two at the end of 50 years.
Instead, the second simulation produced an end value that was
approximately 90 times larger than the first simulation and 20 times
larger than the third. Clearly, the magnitude and consistency of
average returns only explains a small portion on the dynamics that work
to produce wealth, yet they are the cornerstone concepts that drive
most financial planning.
Two remaining inseparable variables explain the bulk
of the differences among the three portfolios: the volatility of
returns and the timing of cash flows in and, especially, out of the
portfolios. Figure 2 illustrates the volatility of returns for the
three portfolios. Figure 3 illustrates the frequency of downward
movements in monthly returns in a generic balanced portfolio. Figures
4, 5 and 6 illustrate the convergence of the timing of volatility of
returns and timing of cash flows. They plot the net cash flows in and
out of the portfolios against the annual returns earned by the asset
allocation that was applied uniformly to the portfolios.
In spite of the fact that the asset allocation
strategy used across these portfolios is diversified in an attempt to
achieve consistency of returns, volatility persists. Interestingly, all
three portfolios have approximately the same number of peaks and
valleys with apparent differences in timing, direction and magnitude.
The performance of these portfolios cannot be
considered as uncommon. We only think it is abnormal because of what we
believe about portfolio optimization. However, asset allocation only
works to dampen average volatility when investment strategies are
examined from the perspective of long-term holding periods. When
examined in "real time" even balanced portfolios experience frequent
downward movement in returns.
Figure 3 shows monthly changes in returns for each month between 1926
and 2002 for a portfolio that was evenly distributed among nine asset
classes, including fixed income (four categories of bonds) and equities
(five categories of stocks). The portfolio returns moved downward (e.g.
the return for the previous month was higher) 53% of the months, which
means that more than half of the time an individual investor was
exposed to the risk of spending at the wrong time.
The timing of drawdowns on portfolio assets relative
to the occurrence of downward movement in portfolio returns is more
significant to the success of financial planning than is asset
allocation. Figures 4, 5 and 6 illustrate the "Draw Down Danger Zones"
(circled in red) that occurred for each simulation. A danger zone
occurs when cash flows go out of a portfolio at the same time that
portfolio returns move downward.
These are considered to be danger zones because a
spending event that takes place in a down market exhausts a higher
percentage of a portfolio than it would if it took place when the
portfolio value was higher. Spending events have the same effect as
negative returns in that they require higher positive returns in
subsequent periods to restore a portfolio's value.
For example, if you lose 20% of your $1,000
portfolio you have a remaining balance of $800. To get back to $1,000
the next period you have to earn 25%, not just 20%. If you spend $200
from an $800 portfolio it is the same as incurring a 25% loss. But, if
you spend $200 from a $1,000 portfolio it is the same as a 20% loss.
In the first simulation (Figure 4) ten danger zones
occurred and were distributed throughout the simulation period (1960 to
2005 plus 1927 to 1930). Overall wealth accumulation could be improved
by either accelerating or deferring cash out flows from the portfolio
into a higher earning period.
It is no coincidence that the second simulation (Figure 5, with a
simulation period running from 1980 to 2005 plus 1927 to 1950) ended
with a much greater value than the other two, since only seven danger
zones occurred and the average magnitude of the drawdowns was
relatively small. Since these danger zones occurred early in the
simulation, deferring cash outflows would have significantly improved
wealth accumulation.
Ten danger zones occurred in the third simulation
period as well (Figure 6, with a simulation period running from 1930 to
1979). In this case wealth accumulation could be improved by deferring
cash outflows from the first four danger zones to a period of improved
returns.
Strategies To Cope With
Drawdown Risks
Since short-term downward movement in portfolio
returns occurs frequently, even in balanced portfolios, the risks of
untimely drawdowns will be high for most individual investors.
Consequently, successful financial planning will include anticipating
these risks and having strategies in place to mitigate them. The
following are suggestions on how to focus your planning in the
direction of drawdown risks.
Integrate cash flow and portfolio growth
projections. These projections will help you anticipate the relative
timing of cash outflow events and downward moving returns and they help
pin-point the causes of, and therefore the responses to, danger zones.
For planning purposes projections should be prepared using short
intervals (monthly or shorter). Short-term intervals make it easier to
see the danger zones which can be caused by any event or combination of
events that result in negative cash flows from a portfolio. These
events can include spending in excess of income and cash reserves,
including paying taxes or management fees, as well as portfolio
rebalancing.
Use planning projections to manage the danger
zones. Execute drawdowns when the "current portfolio return" moves
upward. The "current portfolio return" is the percentage change in
value of a portfolio from one point in time to a following point in
time. For management purposes the interval between points is not as
important as the direction of the current return. That is because your
goal should be to execute events on an uptick rather than a downtick.
Use planning projections to identify "safe portfolio
values." These are the values at which the portfolio has enough funds
that future spending goals and lifestyle needs can be achieved without
exhausting the portfolio, even with very adverse market performance.
These "safe portfolio values" can be used as targets at which point
money can be withdrawn from a portfolio in advance of a need, to be
held in cash reserves for future spending. Although the cash reserves
would likely earn a lower return than the same money might earn in the
portfolio, the danger zone risks would be avoided and the success of
the financial plan could be assured.
Leverage your clients' real world instincts. Help
them adopt behaviors that are useful in mitigating the risks associated
with "Draw Down Danger Zones." Their aversion to recognizing losses can
make it easier for them to defer spending when markets are
underperforming. Their pride in being savvy investors can be preserved
in the tangible evidence of achieving goals ahead of schedule, made
possible by selling winners during market upticks.
Successful financial planning by individual investors will depend
largely on their ability to anticipate and manage drawdown risks.
Anticipating these risks will be aided by integrated financial
projections that pinpoint "Draw Down Danger Zones." Managing these
risks will require understanding the events that cause the danger zones
to occur and implementing appropriate strategies.