One of the core functions of financial planning is setting up clients portfolios in retirement so that resources are adequate to sustain the journey no small feat, given the uncertainties involved and the need to balance stability and safety against the risk of inflation, as well as the need for growth over the potentially long time horizon.
Conventional wisdom suggests that retirees should manage this challenge by having a moderate exposure to stocks at the start of retirement to help their portfolio grow and be able to keep up with inflation over the long run and then reduce equity exposure slowly over time as they age and their time horizon shrinks.
But recent research has suggested that the optimal approach might actually be the opposite start with less equity exposure early in retirement, when the portfolio is largest and most vulnerable to a significant market decline, and then slightly increase the equity exposure each year throughout retirement.
And as it turns out, an even better approach may be to accelerate the pace of equity increases a bit further in the earlier years (from an initially conservative base). After all, a slight equity increase in the last year of retirement isnt really likely to matter.
For instance, a glidepath might aim to increase equities in just the first half of retirement, until the target threshold is reached, and then level off. Instead of gliding to 60% equities from 30% over 30 years, glide up to 60% over 15 years then maintain that 60% equity exposure for the rest of retirement (assuming the 60% target is consistent with client risk tolerance in the first place).
Accelerating the glidepath reduces the time when the portfolio is bond heavy a particular concern in todays low interest-rate environment. And it may be even more effective to simply take interest-rate risk off the table altogether by owning short-term bonds instead. Such an approach leads to less wealth on average, but in low-return environments, rising-equity glidepaths that use stocks and Treasury bills can actually be superior to traditional portfolios using stocks and longer-duration bonds (say, 10-year Treasuries) even though Treasury bills provide lower yields.
In the original research that American College professor Wade Pfau and I collaborated on, showing the benefits of a rising-equity glidepath, we simply assumed that any retiree using a glidepath would make adjustments in a straight line throughout retirement. For instance, gliding equities to 45% from 30% during a 30-year retirement time horizon would require a shift of 0.5% per year.
Gliding to 60% from 30% in the same time horizon would involve shifting 1% per year.
Yet the reality in such situations is that, for someone who is spending down assets, the last 1% change in equity exposure (to 60% from 59%) in the 30th year is not going to impact the outcome. At that point, the retiree has either made it or not.
So we launched a follow-up study, testing the impact of an accelerated glidepath. In this case, instead of moving to 60% equities from 30% over 30 years, the retiree moves there in only 15 years (at 2% per year), and then plateaus.
To test the alternatives, we looked at how they would have performed historically compared with each other with a 4% initial withdrawal rate over rolling 30-year periods in the U.S., starting each year since 1871, assuming a combination of large-cap U.S. stocks and 10-year Treasury bonds that are annually rebalanced.
The results, shown in the How Fast a Glidepath? chart below, reveal that the accelerated glidepath over 15 years is superior to the 30-year glidepath. In most years, the difference is fairly small an improvement of the safe withdrawal rate of 0.1 to 0.2 percentage points but in the best years, the improvement was as much as roughly half a percentage point. The accelerated glidepath is ultimately better in all historical scenarios and improves outcomes in both high-return and low-return eras. Its only a question of how much.
A commonly voiced concern about our original rising-equity glidepath research was the fact that being more conservative with equities in the early years also means owning more in bonds. Thats not necessarily appealing in light of todays low interest rates and the fear that rates will rise at some point in the coming years.
Accordingly, in our follow-up research we also tested the impact of taking interest-rate risk off the table, by using portfolios of stocks and Treasury bills, instead of stocks and 10-year Treasury bonds. The benefit of using Treasury bills is that, because they mature in a year or less, they are reinvested annually, avoiding any risk that the retiree will need to liquidate bonds at a loss because of rising rates. The downside, of course, is that shorter-term Treasury bills generally have lower yields over time (at least in any normal, upward-sloping yield curve environment).
As shown in the Bills vs. Bonds chart below, there are times when Treasury bills help, and times when they hurt. The difference in outcomes between using Treasury bills and bonds is as much as a half-percentage point improvement in safe withdrawal rate, and as bad as a 2-point decrease.
It is notable, however, that in the worst of scenarios someone retiring in the mid- to late 1960s the use of Treasury bills offered a material improvement, lifting the safe withdrawal rate of the entire data set by almost half a point.
Heres why: It is clear in retrospect that the mid- to late 1960s marked the onset of more than a decade of rising interest rates, which led to some very significant losses for 10-year Treasury bonds. It was also an era of poor stock market returns a 15-year period of almost no capital appreciation, with the 1973-1974 bear market in the middle.
In such an environment, it did not pay to take the interest-rate risk of longer-term bonds on top of the volatility and sequence risk of equities.
By contrast, using Treasury bills produced worse results than using Treasury bonds both through the 1930s and again (and more drastically) in the early 1980s. These were times where interest rates and inflation had peaked and were falling (in the 1980s from high inflation to moderate, and in the 1930s from moderate inflation to deflation).
Ultimately, the results show perhaps not surprisingly that if interest rates are more likely to rise, its better to use equities and Treasury bills, while if interest rates are higher and more likely to decline (or at least stay flat), its preferable to use intermediate- or longer-term bonds that can generate capital appreciation as rates move favorably (and, of course, get a higher yield along the way).
While we are still in the preliminary stages of our follow-up research, there are already a few takeaways for advisors.
First: If a rising-equity glidepath is going to be used, the results clearly suggest that an accelerated one over 15 years is preferable to a 30-year glidepath. The improvement is modest, but it helps, and it helps consistently.
The results also show that in situations when interest rates are low and there is an elevated risk of rising rates (and associated bond price declines), it doesnt pay to take interest-rate risk. Safe withdrawal rates in the worst inflationary environments are better with Treasury bills than with bonds, even though Treasury bills may pay almost nothing at the beginning of such periods.
In the worst environments, while its appealing to generate a better return from bonds, compounding bond risk on top of equity risk doesnt pay. Instead, the better approach is to own bills and dollar-cost average into equities avoiding interest-rate risk in the process. The role of the bills is simply to act as ballast to stocks.
Its also crucial to note that a rising-equity glidepath is still a more conservative lifetime equity allocation. It forfeits upside to protect against some of the worst possible retirement scenarios, as is shown on the Comparing the Portfolios chart below.
The dark red line shows the difference between a traditional 60/40 stock and bond portfolio (using 10-year Treasury bonds) and an accelerated rising-equity glidepath moving stocks to 60% from 30% over 15 years (and using Treasury bills to fill in the rest).
Ultimately, it shows that the accelerated rising-equity glidepath detracts from safe withdrawal rates more often than it wins. It wins, however, in the situations when safe withdrawal rates are worst. In other words, the rising-equity glidepath gives up upside in most scenarios to protect against downside in the few worst ones. That makes it a poor wealth maximization approach, but an effective risk management strategy.
In other words, for retirees whose paramount concern is not maximizing wealth, but simply trying to maintain a stable standard of living through retirement from the highest starting point they can, the rising-equity glidepath appears to enhance sustainable retirement income when its needed the most.
Michael Kitces, CFP, is a partner and director of research at Pinnacle Advisory Group in Columbia, Md., and publisher of the planning industry blog Nerds Eye View. Follow him on Twitter at @MichaelKitces.
- Specialized Practice: Retirement Income Planning
- Combating Inflation in RetirementHelp Clients Reset Retirement Expectations
- Help Clients Reset Retirement Expectations