Change Is a Good Thing (Summary)

Defined contribution (DC) plans play a vital role in the retirement outcomes of millions of employees and account for $5.6 trillion in assets in 401(k) plans today. Plan sponsors have a fiduciary responsibility to select and monitor the menu of investments offered to participants.

But in spite of the considerable time and resources plan sponsors spend adding and removing funds, there is relatively little evidence that these activities are fruitful. Strategies based on return chasing—that is, adding funds that have the best recent performance and eliminating those with the worst negative recent performance—have generally been found to be ineffective. However, there is some evidence from the literature that more sophisticated fund analysis tools may provide value.

The authors set out to investigate whether funds added to a DC plan perform better than the funds they are intended to replace over two subsequent periods: one year and three years in the future. They examine historical menus for a DC plan from January 2010 to November 2018, provided by four recordkeepers to a provider of managed accounts services. They identify 4,215 fund replacements—a sample size, the authors note, that is larger than has been used in previous similar studies.

The authors compare differences in attributes between the replacement funds and the deleted funds, including expense ratios, historical relative performance, and star and analyst ratings.

The investment style with the most replacements was “large growth,” and the most commonly replaced broad style group was “equity.”

They perform an event study—estimating raw return differentials at different intervals—to examine whether the future performance of the replacement and deleted funds differ. They then run ordinary least-squares (OLS) regressions to help explain what factors might be driving any differences, controlling for common metrics used by plan sponsors to evaluate funds, such as expense ratios, momentum, and style exposures. They also conduct an analysis to understand whether plan sponsors are better at adding or deleting funds by comparing the future returns of the replaced fund and the replacement fund to a representative style benchmark.

The authors find that compared with the deleted funds, replacement funds have lower expense ratios, substantially higher historical relative performance, and higher star and analyst ratings on average. The replacements significantly outperform the deleted funds on a relative basis over both the one-year and three-year forward-looking periods. However, the magnitude of outperformance is substantially lower in the future period than in the historical period. Whereas 86% of replacement funds have higher historical three-year performance, only 54% have higher future one-year performance, and only 55% have higher future three-year performance.

The most meaningful attribute related to the superior performance of replacement funds is having a lower expense ratio. Funds with higher expense ratios—both among the additions and deletions—tend to underperform. To a lesser extent, higher-quality analyst ratings among the replacements may also have been an explanatory factor. The authors also find that the benefits of the replacement decisions come predominantly from the removal of underperforming funds, rather than from the selection of new funds. In fact, both the eliminated and replacement funds underperform a style-matched benchmark after the replacement decision—but the performance of the eliminated funds is significantly worse.

These results suggest that plan sponsors do provide some value by monitoring and refreshing the menu of funds available to them. The authors conclude that fund replacements outperform deleted funds over both a one- and a three-year horizon—and that this outperformance is statistically significant. But the added value appears to be derived from the ability of plan sponsors to identify and remove underperforming funds rather than to select future high performers. For plan sponsors, the findings of this study suggest that fund replacements with lower expense ratios tend to result in better investment outcomes.