Table 2: Accumulation
Average Annual Return
Table 3: Distribution
Average Annual Return
Targeted Returns Are Not Predictors
of Retirement-Income Success
When accumulating wealth, a targeted rate of return will accu-
rately predict future value if present value and holding period are
known. But a future value cannot be calculated when taking with-
drawals, and higher returns do not guarantee greater wealth.
Table 2 shows a $1-million investment allocated 50/503 applied to
two different twenty-five-year periods, one averaging 7.2 percent
per year (1950–1974) and the other averaging 9.3 percent per year
The final column shows the account value after twenty-five years of
accumulating wealth. Not surprisingly, the period with the higher
return finished with more wealth ($5,973,634 versus $3,659,635).
Because a higher return leads to greater wealth, it creates an incentive to outperform an index or beat the market and invites criticism
of asset managers who fail to do so.
Table 3 shows why retirement income violates these axioms.
Table 3 shows the same investments and time frames as table 2
but with a 5-percent withdrawal from each portfolio in its initial
year (1950 and 1965, respectively), increased by 3 percent annually. Here the period with the lower return finished with a substantially higher ending value ($1,038,862 versus $79,192), illustrating that, for retirees, the timing of returns—especially
losses—has much more impact, assumed or targeted rates of
return do not accurately forecast future values, portfolio management is less predictable, and earning a higher overall return does
not guarantee greater wealth. This makes active management of
retirement-income portfolios more important and beating the
market a distraction to be ignored.
More Risk Does Not Guarantee
Greater Wealth Long-Term
In the wealth accumulation stage, it is axiomatic that you must
assume more risk to earn a higher return, and investors equate a
more-volatile risk policy with greater wealth. But neither a higher
return nor a more-volatile risk policy guarantees greater wealth
for distribution portfolios. Retirement-income success is a func-
tion of the timing of returns, especially losses, not the volatility
Figure 2: Optimal Long-Term Risk Allocation
Year-by-Year Account Values 2000–2013
$1m Investments, 5% Withdrawals Increased 3% Annually
Figure 2 illustrates an optimal long-term risk allocation. It compares the year-by-year values of a 70/30 distribution portfolio to a
less-volatile 30/70 distribution portfolio during 2000–2013, each
with a 5-percent initial withdrawal increased by 3 percent annually.
Figure 2 shows that, in some environments, a less-volatile investment policy is key to greater wealth long-term. Figure 2 shows that
after fourteen years, the less-volatile investment portfolio has
52 percent of the original investment but the more-volatile portfolio has just 26 percent. This is because withdrawals exacerbate market losses, and a catastrophic market loss can provoke a cascade of
principal erosion from which the portfolio cannot recover. This
makes a less-volatile risk allocation optimal long-term in some
environments. Overallocating to stocks to hedge premature principal erosion creates conditions most likely to lead to a catastrophic
loss that triggers a cascade of such erosion. The dilemma for retirees is that, in other environments, the more-volatile investment
would have been the optimal risk allocation long-term. But in
retirement, any set-it-and-forget-it approach (e.g., a static portfolio
or glide path) is potentially dangerous. A distribution portfolio
demands tactical management of downside risk.
The Risk Three-Step: Erosion, Cash Flow, and Panic Risk
In addition to accelerating principal erosion, a large loss early can
lead to a significant reduction of cash flow (cash-flow risk).