Monday, May 26, 2008

An Experiment on Mutual Fund Fees

Apparently, the message that mutual fund fees matter is not getting through. In a recent paper, researchers Choi, Laibson, and Madrian describe an experiment where they give index mutual fund prospectuses to subjects and ask them to create a portfolio with $10,000.

The experiment was carefully controlled so that the only variable was the fees charged by each of four fund choices. Each of the funds was designed to mirror the S&P 500 index, and the subjects were offered incentives to maximize future returns. Some subjects were offered one month worth of upside on their $10,000 portfolio. Others were offered more modest rewards for good performance.

Before describing how the subjects performed, let me say a little about who participated. The subjects were Harvard staff members, Harvard students, and MBA students from Wharton. The students had SAT scores placing them in the top 2% of the population. So, you’d think that they would get above average results in this experiment.

However, the results were dismal. The best strategy was to put 100% of the money into whichever fund had the lowest fees. For subjects who were given just the prospectuses to read to make their decision, the mean fees they chose were more than 1% above the minimum. One group chose portfolios that averaged more than 2% above the minimum.

Even MBA students who understand the importance of fees chose portfolios with only 0.1% lower fees than other participants.

Some of the participants were given fee summary sheets explaining loads and expense ratios to make the comparison between funds easier. Even then, fewer than 20% of subjects chose the best portfolio.

It appears that mutual fund companies don’t have to work very hard to distract investors from their excessive fees.

No comments:

Post a Comment