One of the most important dilemmas individuals will tackle at retirement is to decide how much income should be withdrawn from their portfolios. The answer has become more difficult, especially in recent years, because changing rates of return on the various investments favored by retirees are much lower today than in the past. It wasn’t long ago that interest payments from fixed income investments such as treasury bonds, municipal bonds and certificates of deposit sported returns of 8%, 9%, 10% and more. Today’s rates of 4%, 5% and 6% make counting on these returns for retirement income a much more uncertain affair.
Because these investments alone can no longer be counted upon to provide enough return to support a reasonable standard of living (especially after taxes and inflation), other investments with different types of risks must be considered. The alternate investment of choice is common stock. Stocks have shown superior rates of return over time, relative to the fixed income investments mentioned above. The bulk of these superior rates of return, however, have come from the investment appreciation of the underlying securities. This appreciation can vary considerably year-to-year, making it difficult to know how much income can be taken from the portfolio’s rate of return in any given year, in contrast to stock dividends, which are paid directly to shareholders on a quarterly basis. The problem is further advanced by the fact that the percentage of dividends paid to shareholders has substantially declined in recent years. How, then, do we assemble a portfolio that will deal with this uncertainty and produce enough income to support us comfortably in our retirement years?
The answer is now clearer based on a study conducted by three professors of finance at Trinity University in San Antonio, Texas. In this study, Professors Philip L. Cooley, Carl M. Hubbard and Daniel T. Walz measured the various rates of return earned by stocks and bonds over the period of years between 1926 through 1995 and matched those returns to different withdrawal amounts over distinct periods of time ranging from 15 to 30 years.
Finding the right mix
It is better for this exercise to think of portfolio income in percentage terms rather than in dollar terms. For example, the percentage of income a portfolio may reasonably support over a given number of years may be 3% per year ($30,000, based on a $1,000,000 portfolio) or 7% ($70,000). Next, the effect that a particular withdrawal rate had on different combinations of stocks and bonds over varying periods of time was analyzed. In other words, the result of taking annual income of 5%, for example, for a period of 15, 20, 25 and 30 years was analyzed. Put another way, “Will a portfolio consisting of 50% stocks and 50% bonds support a certain level of income better than a portfolio consisting entirely of stocks?” These questions were examined (see Table One) by determining the effect of withdrawals on portfolios containing:
- 100% stocks
- 75% stocks/25% bonds
- 50% stocks/50% bonds
- 25% stocks/75% bonds
- 100% bonds
Portfolio success was characterized by calculating the percentage of time periods in which the portfolios had amounts remaining in excess of $0.00. Of course, planning to have a portfolio worth $0.00 by the end of a certain period is a questionable goal, so the actual dollar amounts remaining at the end of each period were reported.
The Results
For purposes of simplicity, the results are divided into three categories of success:
- Very high
- High
- Moderate*
It should be assumed that any result not in these three categories is not worth listing.
The Bottom Line
As the information in Table One indicates, the study showed that portfolio success rates decline with increases in the length of the payout period and increases in the percentage paid out. As professors Cooley, Hubbard and Walz state in their study, “If history is any guide for the future, then withdrawal rates of 3% and 4% are extremely unlikely to exhaust any portfolio of stocks and bonds…”
The most surprising finding was that bonds added to the success rate of some of the portfolios. The authors believe that the diversification provided by bonds adds to the stability of the portfolio. This reduces the negative effect of withdrawing money out when the market is in steep decline. As a matter of fact, for withdrawal rates of 7% or lower, a 50% stock/50% bond allocation had higher success rates than portfolios with greater amounts of stock.
What about inflation?
Since the previous illustration does not account for the effect of inflation, calculations were made to determine inflation’s effect over these time periods. To do this, one must extract less income from the portfolio than is actually earned. (In this example, the term “income” is used to mean a combination of dividends, interest and capital appreciation). This leaves the “inflation” amount in the portfolio producing more income in the future.
The study’s results showed a decline in success rates for mid-level and high withdrawal rates. Three percent and 4% withdrawal rates continued to produce very high success for stock-dominated portfolios and 5% percent withdrawal rates produced very high and high success rates for all payout periods. However, the 6% and 7% rates perform reasonably well for short periods only. All withdrawal rates above 7% perform poorly for all payout periods.
The rest of the story
As mentioned before, defining success as having more than $0.00 at the end of a period of time is hardly satisfactory. If we knew when we were to leave this earth it may be acceptable, but for most of us, this is obviously not the case. What, then are the likely amounts of money we would have left under these scenarios? Chart One shows the various ranges of values left at the end of the selected time periods. Calculated are the minimum, maximum and average amounts.
To use this chart, look at the percentage indicated on the horizontal scale, then look at the bar directly above it. The diamond at the bottom of the bar represents the least amount remaining; the box in the middle, the average and the triangle at the top, the most. In this example, the first entry indicates 4% for 15 years. The least amount remaining (for a hypothetical $1,000 investment) is around $500; the average, $3,000 and the most, around $6,000.
Conclusion
What is the appropriate annual withdrawal rate from a portfolio during the retirement years? The answer clearly depends on the mix of stock and bonds and the amount of time one will need to withdraw the money.
The study states:
- Early retirees who anticipate a long payout period should plan on lower withdrawal rates.
- The presence of bonds increases the success rate for low to mid-level withdrawal rates. Most retirees would most likely benefit from a 50/50 mix of stocks and bonds. As stated earlier, for withdrawal rates of 7% or lower, a 50% stock/50% bond allocation had higher success rates than portfolios with greater amounts of stock.
- If one wishes to account for inflation, one must accept substantially reduced withdrawal rates from the initial portfolio.
- Retirees with stock dominated portfolios and low withdrawal rates will most likely leave large inheritances to future generations and may want to consider early tactics to protect the inheritance from estate tax.
- For payout periods of 15 years or less, withdrawal rates of 8% or 9% from stock-dominated portfolios appear to be sustainable.
