We tested a variety of return assumptions, investor savings schemes and risk tolerances using simulated and historical data. We examined both average statistics (such as mean and standard deviation) as well as percentiles (such as the median or 95 percent lower bound) to measure risk and return. In none of these comparisons did we find compelling evidence for outstanding superiority of a particular glide path. Instead, we found a fixed stock/bond ratio investment to be surprisingly robust to potentially uncertain market conditions. We are concerned that retirement investors, who may have little information or expertise about investing, are being steered into a false sense of security investing in glide-path funds. We have presented evidence that even a simple glide path is no guarantee of lower risk at retirement, including a historical analysis showing the 80 to 0 percent glide path worsened both risk and return compared with a static 40 percent stock strategy—the standard deviation of wealth being $47,000 higher and the mean wealth being $193,000 lower. The highly sloped glide paths popular in many target-date funds fare little better in our simulations.
Default glide paths correspond to a particular risk tolerance, which may not match the investor’s circumstances or preferences faithfully, especially over time. The true level of risk taken over the course of savings is somewhat obfuscated by the glide path, and the very fact that it can be matched in risk characteristics with many other paths without greatly harming the retirement wealth distribution shows the arbitrariness of the glide-path selection. Offering glide-path funds as a qualified default alternative may give a false impression that fiduciary review has been completed. We believe this is bad policy and may constrain cross-sectional suitability of the portfolio.
Empirical evidence from real-world TDF investors also bears out the possibility that glide-path funds are not always matched well to market climates—TDFs can lose value just like any other funds if they are exposed to assets with falling prices.12 Losing one’s retirement savings as the target date approaches is a shame, especially if a prudent review of the portfolio or greater diversification could have prevented some of the loss. For investors sticking to a long-term and consistent investment plan, TDFs have dubious value. For those investing a lump sum, they’re inferior to an equivalent target-risk portfolio.
Concerns About Matching On Standard Deviation
Wealth distributions tend to exhibit positive skewness, with a few wealthy individuals exerting disproportionate influence on the mean and standard deviation. If the matched glide paths shown, for example, in Figures 1, 3, 4, 5 and 6 have nonmatching skewness, the comparisons may be invalid. To investigate whether the standard-deviation matched comparisons are valid, we performed both a visual exploratory analysis and a computation of skewness values for all the glide paths. The glide paths shown in Figure 1 have computed skewness values from 1.97 for the maximally descending glide path to 1.49 for the maximally ascending glide path. The static 40/60 portfolio has a skewness of 1.53, and the glide path with maximum expected return has a skewness of 1.58. There are clearly differences in skewness among the glide paths. However, it is unclear if they are substantial enough to invalidate the comparisons since they are not great enough to clearly sway an investor’s preference, given the differences in expected returns.