How Investment Options Affect Portfolio Choice in Individual Accounts

Featured in print Bulletin on Aging & Health

More than two dozen countries around the world now have individual accounts as part of their public pension systems. While such accounts are not currently part of the U.S. Social Security system, many prominent policy makers and academics have pro-posed introducing an individual account component to the system. Moreover, millions of U.S. workers already have individual ac-counts in their employer-sponsored pension plans, as defined contribution or 401(k)-style plans have increasingly come to replace traditional defined benefit plans.

A key issue in designing individual accounts is determining what menu of investment choices to offer to participants. Standard portfolio theory suggests that it sufficient to offer just two options - the "market portfolio", an asset that incorporates all avail-able assets in proportion to their importance in the market, and a risk-free asset - and allow investors to select the combina-tion of these that matches their risk preferences. In reality, participants face a more complex menu of options. Employer-sponsored pension plans in the U.S. typically offer options from several asset classes, such as domestic equity, international, bond, and balanced funds, and often offer multiple funds within each asset class. In other cases, the menu is even more complex - in the Swedish social security system, account holders can choose from over 650 funds.

In Individual Account Investment Options and Portfolio Choice: Behavioral Lessons from 401(k) Plans, (NBER Working Paper 13169) researchers Jeffrey Brown, Nellie Liang, and Scott Weisbenner explore whether the mix of investment options available in an individual account plan affects participants' portfolio choices.

One challenge in examining this question is that the investment preferences of workers may differ across firms. If preferences are correlated with fund options - for example, because workers with a high risk tolerance ask their firm to offer more equity funds - this will introduce bias into the estimation. Previous studies have typically used data on a cross-section of firms in a single year, making it difficult to control for firm-specific differences in worker preferences.

The authors overcome this challenge by constructing a firm-level panel data set on pension fund options and contributions during the 1990s, using firms' 11-k filings with the Securities and Exchange Commission. By having repeated observations on firms, the authors can control for differences in investment preferences of workers across firms and over time, essentially using within-firm changes over time to identify the effect of fund options on portfolio choice.

The authors first report on fund options and contributions in their sample. The typical firm offers 7 or 8 funds - one company stock (all firms who make 11-k filings offer company stock), three domestic equity, one international equity, two bond, and one balanced fund. Most firms offer between 4 and 12 funds. The allocation of fund offerings and contributions varies across firms. While the typical firm has 38 percent of its fund offerings and 45 percent of its contributions in domestic equity, these figures can be as low as 20 percent at the 10th percentile and as high as 56 percent (for offerings) or 68 percent (for contributions) at the 90th percentile.

The central question examined by the authors is how portfolio allocation changes as fund options change. If participants follow an optimal portfolio strategy or a fixed allocation rule such as a 60/40 stock/bond mix, then increasing fund options, for example by offering a new domestic equity fund, should not have much of an effect on portfolio allocation. If, by contrast, participants follow a "naive" diversification strategy of allocating their assets evenly across all funds or randomly allocate all assets to a single fund, then increasing the share of equity funds among the firms' offerings will increase the share of the portfolio invested in equity.

The authors find that the mix of investment options has a strong effect on portfolio allocation. For example, suppose that a firm offering 3 stock and 2 bond funds adds a new stock fund, raising the share of its fund offerings in domestic equity by 6.7 percentage points (from 3/5 to 4/6). This change is estimated to raise the share of the participant's portfolio that is invested in domestic equity by 3 percentage points. The response to an increase in bond or balanced funds is similar, while the response to an in-crease in international equity or company stock funds is smaller but also significant.

The magnitude of these effects is smaller than what would occur if all participants engaged in naive diversification, but larger than what would occur if all participants followed an optimal strategy or fixed allocation rule. The authors also find that unobservable firm-specific factors are an important determinant of how assets are allocated, especially for company stock, since the results including these controls (discussed above) and excluding them differ substantially.

To explore the implications of these findings, the authors examine how fund offerings have changed in recent years. Between 1998 and 2002, the typical plan added four funds to its offerings, with two-thirds of the growth coming in actively-managed equity funds and only 8 percent in stock index funds like the S&P 500. As documented in other literature, actively managed funds charge substantially higher annual fees than do index funds, without earning higher after-expense returns on average (and perhaps per-form no better before expenses). This raises the concern that an increase in fund offerings may result in more money being allocated to high-cost funds that will on average yield lower net returns. A simple calculation by the authors shows that a 0.35 percent difference in annual returns can result in a non-trivial 7.5 percent difference in wealth at retirement, due to the power of compounding.

The authors conclude that their study "strongly suggests that average participants are not optimally allocating their portfolios according to standard finance theory predictions, but instead are following naive strategies that subject them to 'manipulation' by non-binding changes in the number and mix of investment options. A key policy implication is that the number and mix of investment options will have an important effect on overall asset allocation in the individual accounts."

The authors acknowledge funding from the U.S. Social Security Administration through a grant to the National Bureau of Economic Research as part of the SSA Retirement Research Consortium.