Foolish Payouts

The Trinity Study discussed in our article How Much Are Ya Gonna Take? provides a useful methodology for looking at withdrawals from a Foolish portfolio in retirement. A Foolish portfolio may be constructed in all kinds of ways, but for the purposes of this article we'll be using data assembled on the Foolish Four stocks covering the period 1961 through 1998. Using this data, we can compare five portfolios consisting of:

  • 100% Fool Four
  • 75% Fool Four and 25% long-term corporate bonds
  • 50% Fool Four and 50% bonds
  • 25% Fool Four and 75% bonds
  • 100% bonds
Both the Trinity Study and the one done by the Retire Early website indicate inflation-adjusted withdrawals fare poorly at higher withdrawal rates. Therefore, to make things interesting we will adjust the initial withdrawal from each portfolio by the inflation rate in all succeeding years.

Payout rates will range from 3% to 12% of the starting portfolio value of $100,000, and we'll measure five payout periods: 15, 20, 25, 30, and 35 years. Our success rate for each payout period will be measured by the ratio of the number of ending portfolios with a value greater than zero to the number of possible payout periods in the years 1961 through 1998. We'll compare those results to a similar analysis that uses the S&P 500 stocks as was done in the Trinity Study. Return data for the Foolish Four stocks is from the official 1998 Dow Spreadsheet maintained by The Motley Fool. Return data for the S&P 500, long-term corporate bonds, and inflation is as reported by Ibbotson Associates.

Success rates for the FF portfolios are as shown in Table One. Note that there is a 100% success rate for all payout periods for withdrawals of up to 6% in the 100% FF portfolio. Success rates remain reasonably high for all payout periods through an 8% withdrawal rate, but decline significantly at higher payouts. In the 75% FF portfolio, we again see a 100% success rate for all payout periods through the 6% payout rate. At a 7% withdrawal rate there is a degradation in success rates as compared to the 100% FF portfolio, and that degradation continues at higher withdrawal rates. In a 50% FF portfolio, the 100% success rate only goes through a 5% withdrawal rate, with a marked decline in success at higher payout rates. A 25% FF portfolio could sustain a 100% success rate in all payout periods only at the 3% and 4% payout rates. The 100% bond portfolio failed miserably in comparison with all other portfolios.

Results for the S&P 500 portfolios are shown in Table Two. Of interest here is the fact that a 100% success rate for each portfolio containing the S&P 500 could be sustained for all payout periods only at a 3% withdrawal rate. Use of a 4% withdrawal rate had a reasonable probability of success only in the 100% S&P 500 and the 75% S&P 500 portfolios. None of the portfolios had a good probability of success using a withdrawal rate of 6% or higher save for the shorter payout periods of 15 and 20 years. Even then, they were mediocre at best and significantly lagged the FF portfolios.

Wow! Pop the keg and bring out the noisemakers. The Foolish Four smothered the S&P 500 in all portfolios. And the addition of bonds ran contrary to the Trinity Study. In the latter, adding bonds to the portfolio increased the stability of the inflation-adjusted withdrawals with little decline in success rates up to a 50/50 mix of stocks and bonds. But in the 1961 to 1998 period, their addition significantly degraded success rates. What went on over that 38-year span to cause that effect? My guess is two things happened. The first is the down market of 1966 along with the bear market of 1973-1974, neither of which was offset significantly by bond returns. The second is the double-digit inflation years of 1974, 1979, and 1980, along with a generally high inflation rate throughout the 1970s. Ibbotson tells us that in the years from 1970 through 1979, the average annual total return of the S&P 500 was 5.9%, while long-term corporate bonds averaged 6.2% per year, and inflation averaged 7.4%. In short, it was an era rife with disaster for the unwary retiree seeking inflation-adjusted income. Unless, of course, he was a Foolish Four retiree. That's because the Foolish Four stocks enjoyed an average annual return of 17.3% in that 10-year period.

Table 3 shows the ending portfolio values for each payout period over the range of 100% FF down to 50% FF. Withdrawal rates range from 5% to 8%. The ending values reflect what was left after 1999 income was taken for each payout period. Note that even at a 7% initial withdrawal rate, some serious estate planning issues could have arisen for retirees who used these portfolios. The majority of those who used lower payout rates definitely would have required the services of an estate planning attorney. I'll take that hardship, though. Certainly those attractive ending values indicate that rigid application of the payout rules may not be necessary when flush times are at hand. Current income could have been increased in some years with nary a danger of running out of income.

Were the past 38 years indicative of the future? Only a fool (little f) would say that's a certainty. On the other hand, there's no indication we are headed for a depression, either, though that's also possible. Regrettably, Foolish data doesn't go back to 1926, so it's impossible at this time to say what the addition of 36 more years of data would reveal. The nature of the value-oriented approach used by the Foolish Four as well as its superb performance during the nasty decade of the 1970s -- especially during the bear market of 1973-1974 -- and in the face of inflation as opposed to deflation makes me reluctant to say the Foolish Four methodology would experience a marked decline in success rates should we include those years. We should be aware, though, that it could. Additionally, we should look at the few portfolios that failed in this study to determine why that happened.

The failures shown in Table 3 occurred in the 20-year or longer payout periods at payout rates of 6% and 7%. Those failures were a matter of timing because the payout periods began in 1965 and 1966. Portfolios starting then were faced with the market drop in 1966 and followed by a sustained bear market in 1973 and 1974, all the while trying to increase annual payouts by the inflation rate. As we saw before, that is an excellent prescription for disaster. Faced with such large market setbacks, prudence says we should curtail withdrawals in lean years by keeping income steady or (perish the thought) even reducing it for a year or two. In reality, we probably would do exactly that. Nevertheless, the constraints of this analysis don't allow for reductions anymore than they allow for taking income beyond the annual inflationary increases.

The lesson we can take from this effort is that the use of the Foolish Four investment approach in a retirement portfolio may lead to an estate planning problem. Indeed, follow the consensus approach outlined in our previous article of a 4% to 6% fixed withdrawal rate instead of the inflation-adjusted withdrawals we used here, and that problem could very possibly get out of hand. That would really make the kids mad because our estates would have to give too much to Uncle Sam, thus leaving less for them. But I certainly would like to wrestle with that problem. How about you, Fool?

Food for thought, certainly. To discuss questions regarding this article, visit us on the Retirement Investing message board.