Skip to Content
Rekenthaler Report

Do Retirees Have It Backward?

The argument for owning more stocks with age.

A Radical Idea
This Monday, the fund-industry publication Ignites drew my attention to a paper called "Reducing Retirement Risk with a Rising Equity Glide-Path." The authors, Wade Pfau and Michael Kitces (the latter is an occasional Morningstar.com contributor), advocate just what the title suggests, that retirees would be best off increasing their equity allocation over time. To use the nomenclature of target-date funds, they should adopt a rising glide path. 

My immediate thought was Dow 36,000Sell the stock market gospel into a hot market by telling the audience what it wishes to hear. But such is not the case. Neither author has a history of publishing crowd-pleasers. Rather, the two typically write detailed and relatively arcane articles about financial planning, such as variable income withdrawal rates, the use of intermediate annuities during retirement, or how financial advisors should be regulated. Fast-talking marketers they are not.

To state upfront: The paper doesn't successfully make its case. That will be clear later in this column, when I post the numbers. However, it is a useful exercise that offers instructive points. Thanks to the authors for providing six pages' worth of excellent tables!

The intuition for a rising glide path is straightforward. Retirees who own a mix of stocks and bonds are safe if stocks perform reasonably well throughout the retirement period. They also are in good shape if stocks thrive during the early years of retirement, but then slide in the later years. The big bulge afforded by the early bull market plumps up the portfolio so that it can withstand the lean that follows. The primary danger to retirees, then, comes from an early bear market--when stocks are whacked shortly after the investor retires.

In which case, argue the authors, retirees are best off raising their stock allocation. The immediate market decline shrinks the size of the initial nest egg, so that annual income withdrawals become a higher percentage of the principal. The portfolio needs market appreciation to nurse it back to health. The more stocks, the likelier it is that health will be recovered. (This principle holds true if stock returns are fully independent--meaning that a series of losses improve the chances of upcoming gains--but the effect is even stronger if stocks are mean-reverting, which historically has been the case.)

The authors test the thesis under only high withdrawal rates--4% per year (expressed as a percentage of the initial wealth, adjusted each year by the amount of inflation) and 5%. Even the lower of the two is an aggressive target given today's stock and bond valuations.*

*Yes, I know, I have argued that 4% is a reasonable rate. But that argument assumes a flexible withdrawal strategy, rather than the fixed and immutable 4% per year that is used in this paper's model. 

The paper shows simulation results for these two withdrawal rates for three capital-market expectations: 1) the historical data from 1926-2011; 2) an estimate from financial planner Harold Evensky that slashes expected stock returns by 300 basis points per year and bonds by 80 basis points; and 3) an even-more conservative forecast that cuts stock returns slightly further and takes bonds' real returns to nothing.

Below are two tables that give a flavor of the results. 

Each table shows the success rate for the simulations under various conditions, with success defined as the initial wealth being able to fund the scheduled withdrawals for each of the 30 years. Next to the success rate is the amount of the shortfall when the simulation suffers a bad draw, defined as the 95th percentile. The shortfall is expressed in years. For example, 4 means that in a 95th-percentile simulation, the initial wealth is exhausted four years ahead of schedule, in year 26.

The first table shows the totals for a relatively high 45% average position in stocks during retirement. The rising glide path begins with 30% in stocks, increases equities by 1 percentage point per year, and finishes at 60%. The declining glide path is the reverse, starting at 60% and ending at 30%. Finally, there is a flat glide path that consists of 45% stocks and 55% bonds throughout the period.



The second table assumes a lower asset allocation, with a 25% average weighting in stocks during the retirement period. Here, the rising glide path starts at 10% and ends at 40%; the declining glide path is again the reverse; and the flat path is at 25%.



Obviously, these figures don't support the authors' contention.

At the higher asset allocation and a 4% withdrawal rate, the rising glide path works slightly better than either the flat glide path or the declining glide path for the Evensky and the historical market assumptions. However, with the conservative forecast the customary approach of reducing stock exposure over time is best.

When the withdrawal rate is increased to 5%, the traditional declining glide path yields the highest success rate for all three forecasts. The shortfall amount is higher as well, so one could argue that the declining glide path is not truly the better overall choice--but it's not worse, either. Call it a draw.

At the lower asset allocation of an average of 25% stocks, the traditional glide path yields a higher success rate across the board, for all three market outlooks and for both withdrawal rates. As most of the shortfall rates are similar, one must give the edge to current practice over the authors' counterproposal.

Of course, 45% and 25% average equity positions with 1%-per-year changes are merely two ways to test the proposition. One could measure the results for a 65% average stock position, or 35%, or many other amounts. One could change the asset mix by 1.5 or 2 percentage points per year. Some decisions would improve the relative showing of the rising glide path. None, however, would make the tactic anything like a universal rule. Rather, it is an approach that might make sense under some conditions, given certain assumptions.

Indirectly, the paper raises several other issues that are relevant for retirees. I'll cover those in tomorrow's column.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.

Sponsor Center