Over the past few decades, risks associated with providing for financial security in retirement have increasingly shifted from employers to employees as employer-provided pensions have shifted from defined-benefit to defined-contribution (DC) plans. Recent work in behavioral finance suggests that investors do not make optimal investment decisions in their DC plans. The authors designed and administered a pair of mutual fund choice experiments to over 1000 survey respondents who participate in the RAND American Life Panel. Their analysis sheds light on the question of how mutual fund investors respond to variation in fees in a hypothetical scenario in which fees should be obvious to the investor. The results show that some aspects of individual behavior are consistent with rational wealth-maximization and the majority of the respondents are able to provide estimates of fees that lie within a benchmark range. However, they find that respondents tend not to minimize expected fees and are more averse to backend load fees than to front-end loads. The trade-off between expense ratios and loads is found to be somewhat sensitive to the expected holding period in a manner consistent with expected-wealth maximization, but investors may tend to be too averse to loads. Differences in measured financial literacy predict differences in behavior, with lower rates of literacy among women accounting for differences in choice behavior by gender. They also find that financial literacy mediates individual responses to the presentation of information intended to enhance decision making.
Dominitz, Jeff Emmett, Angela A. Hung, and Joanne K. Yoong, How Do Mutual Fund Fees Affect Investor Choices? Evidence from Survey Experiments. Santa Monica, CA: RAND Corporation, 2009. https://www.rand.org/pubs/working_papers/WR653.html.
Dominitz, Jeff Emmett, Angela A. Hung, and Joanne K. Yoong, How Do Mutual Fund Fees Affect Investor Choices? Evidence from Survey Experiments, Santa Monica, Calif.: RAND Corporation, WR-653, 2009. As of October 26, 2021: https://www.rand.org/pubs/working_papers/WR653.html