Defined-contribution savings plans sometimes permit participants to select their retirement savings investments. Many plans offer investors numerous funds, ideally allowing them to maximize their utility as measured by individual risk and return preferences. The authors survey employees at two organizations to see if they prefer the retirement portfolio asset mix they selected for themselves over other mixes. They find that retirement plan participants do not prefer their own portfolio when presented with various options, indicating that choice alone does not help them make an optimal allocation. Possibly, participants do not have enough knowledge of investments or of their own preferences to make optimal choices.
Defined-contribution savings plans often let participants determine for themselves how to invest their retirement savings. Such plans as 401(k)s and 403(b)s are popular in the United States, and similar programs are being introduced in other countries. These plans typically offer several investment funds for the participants to choose among when allocating their savings; one plan in Sweden, at an extreme, allows investors to select from 450 investment options. The authors want to know if these choices allow investors to maximize their utility by creating optimal portfolios for themselves; they also want to know if the portfolios selected maximize the participants' return relative to the risk taken.
To find out how choices affect utility, the authors surveyed employees at two institutions—the University of California at Los Angeles (UCLA) and SwedishAmerican Health Systems. At UCLA, the authors compiled information about the aggregate investment choices of the plan participants and then projected the range of retirement income from each participant's portfolio, from the portfolio with the average allocation of all plan participants, and from the median portfolio, based on the standard deviation of returns of all plan portfolios. They asked participants to rate each portfolio using the projected range of retirement income, not knowing which portfolio was their own.
Regardless of their personal risk preferences, most participants preferred the median portfolio to their own. The median portfolio was selected by 62 percent of those surveyed as the best, as opposed to 21 percent who selected their own portfolio.
The authors conducted a different survey at SwedishAmerican, where a consulting firm recommends portfolios for participants based on basic demographic information, especially age, but not risk preferences. Participants are alternatively allowed to select their own investment allocation. The authors asked those who opted out of the recommended plan to rate the attractiveness of three portfolios—the person's own portfolio, the average allocation of all participants in the plan, and the portfolio recommended by the consultant for the participant. The results are similar to those at UCLA: A full 61 percent of those surveyed preferred the recommended portfolio to their own.
The authors examine why those surveyed did not like their own portfolios. Reasons might include the failure to select efficient portfolios, differences of opinion about future stock market performance, unrealistic assumptions about the equity risk premium, changes in preferences from the initial plan enrollment, and mistakes in the initial asset allocation. It is this last explanation that the authors favor. In other words, defined-contribution retirement plan participants do not have the skill or the understanding of their own preferences to select appropriate investment portfolios. Furthermore, their choices can be skewed by the offerings given to them. In a plan offering many equity options, participants will disproportionately choose equity funds over fixed-income alternatives regardless of their true risk preferences. The authors conclude that choice by itself may not benefit retirement plan participants and may cause them to act against their own interests.