SEPTEMBER 2004


• • • • • • • • • • • • • • • • • • • • • • • • • •

Getting the Right Asset Allocation Mix

Defined contribution savings plans give individual employees more responsibility for their investment decisions-are they up to the task?

Research by Richard H. Thaler

What mix of fixed income and equity funds should companies offer in their 401(k) savings plans to prevent participants from investing too conservatively or too aggressively? How should plans deal with differences in risk aversion across the participant population? Should plans offer different funds based on the age of participants, allowing young employees to select aggressive, stock-rich portfolios of funds and older employees to gravitate toward fixed-income funds?

The array of funds offered to plan participants can strongly influence the asset allocation individuals select-especially given the growing popularity of defined contribution savings plans where investment decisions are made by plan participants.

In the study, "Naive Diversification Strategies in Defined Contribution Savings Plans," Richard Thaler, a professor at the University of Chicago Graduate School of Business, and Shlomo Benartzi of the Anderson School of Management at UCLA demonstrate that many investors hold naive notions about diversification and a pronounced lack of sophistication about portfolio asset selection.

To a great extent, participants choose their asset allocations based on the overall makeup of the funds in the plan. For a number of years, TIAA-CREF, the world's largest defined contribution saving plan, offered two investment options, TIAA (bonds) and CREF (stocks), and found about half of plan participants split their allocation right down the middle.

"It's particularly common for the proportion of assets invested in stocks to depend on the proportion of stock funds the plan offers," says Thaler. There is no shortage of data to back him up.

Putting Money on the Line

Using actual data from the Money Market Directories (MMD) covering 170 retirement savings plans with annual contributions of $3.23 billion and assets of about $50 billion, the authors found that 61.76 percent of the available investment options in the plans were equity options, and that the mean allocation to equities, defined as the combined allocation to company stock, domestic equity, and international equity, was 62.22 percent.

Plan sizes ranged from 100 participants to 237,500 participants. The database included a list of the investment options available to participants in each plan, and provided information about each investment option, including its investment style (i.e., money market, bonds, domestic equities), its assets as a percentage of the plan assets, and the year it was added to the plan.

Turning to a simple categorical analysis, Thaler and Benartzi used the relative number of equity funds to categorize the retirement savings plans in the database into three equal-size groups: low, medium, and high. In the low group, 37 percent of the options were equities; in the medium group, 65 percent; and in the high group, 81 percent. Calculating the mean allocation to equities, they found that the low group had an allocation of 48.64 percent, the medium group 59.82 percent, and the high group 64.07 percent.

"Participants will choose mostly stocks in a plan predominantly offering stock funds, and mostly bonds in a plan primarily featuring fixed income funds," says Thaler.

Contributions were then evenly divided across the number of options, resulting in a portfolio that had 75 percent of the total allocation in stocks. Over a 20- to 30-year time frame, however, such naive allocation strategies could prove costly.

First, investors might not choose from the best available alternatives, and secondly, they might choose the wrong risk-return tradeoff. The first type of error is not likely to cause too much damage, because even a naive allocation strategy will result in a well-diversified portfolio.

By contrast, picking the wrong risk-return tradeoff can be exceedingly costly. For instance, for someone with minimal risk aversion, investing too little in equities can lead to large "ex ante welfare costs," i.e., the anticipated costs of making a decision before the outcome is known.

The Bottom Line

To place the significance of the effects of naive investment behavior in context, Thaler and Benartzi found that a rational investor would increase his or her equity exposure from 50 percent to only 53 percent as the proportion of equity funds increased from 33 percent to 87 percent.

"This implies that the shifts in equity exposure are much more strongly influenced by the array of funds in the plan than would be expected," says Thaler.

One potential flaw in this analysis is the possibility that companies choose the fund options in their plans to reflect the desires of their plans' participants. Thus, the underlying risk preferences of the plan participants would drive the observed association between the relative number of equity funds and asset allocation, the authors note.

To test the likelihood of that possibility, the authors used a time series analysis of changes in the asset mix within plans. They obtained data from a pension consulting firm on one mid-size company that had twice changed the array of funds it offered during a four-and-one-half-year sample period. There were two opportunities to observe how employees' asset allocations responded to the changes.

As the company made additional stock fund options available to employees, the allocation to stocks increased. A later elimination of a bond fund also increased equity allocation.

Food for Future Thought

The results of Thaler and Benartzi's research suggest that the surge in private plan retirement funds invested in equities over the past decade may have much to do with the numbers of new equity funds added to these plans. That trend could continue, based on the ease of differentiating equity funds by firm size, style, industry or sector, and country or region, compared with the relative difficulty of differentiating fixed-income funds by anything other than maturity and risk. Their findings also have implications for the design of both public and private retirement savings plans.

While economists designing plans may worry about ex ante welfare costs, public or private plan administrators may also worry about ex post regret-the after-the-fact costs of a decision.

"A plan that by design encourages investors to put an unusually large or small proportion of assets in equities may suffer later if returns differ from historical norms," cautions Thaler.


• • • • • • • • • • • • • • • • • • • • • • • • • •

Richard H. Thaler is Robert P. Gwinn Professor of Behavioral Science and Economics at the University of Chicago Graduate School of Business.

>> View the study: Naive Diversification Strategies in Defined Contribution Savings Plans