Are Some Mutual Fund Managers Better than Others?

08/01/1997
Summary of working paper 5852

...there are substantial differences between funds managed by managers from low- versus high-SAT schools that cannot be explained by differences in behavior, expenses, or survivorship.

For years, economists have debated whether it is possible for mutual fund managers to "beat the market," either through superior stock picking abilities, or by correctly predicting the timing of overall market advances and declines. A related question is whether mutual fund performance is persistent over time: that is, whether some managers simply have "hot hands." In fact though, for most of the past decade, the average mutual fund manager has consistently underperformed the broader stock market indexes, once fund expenses are taken into account.

Now an NBER study by Judith Chevalier and Glenn Ellison suggests that the funds' returns can be predicted by at least some of the specific personal characteristics of their managers, such as age, average SAT score of their undergraduate institution, and whether they hold a Masters of Business Administration degree. Indeed, the results suggests that some groups of managers may be able to beat the broader market.

In Are Some Mutual Fund Managers Better Than Others? Cross-Sectional Patterns in Behavior and Performance (NBER Working Paper No. 5852), Chevalier and Ellison focus on managers, instead of funds: given the high rate of managerial turnover in the mutual funds industry, the distinction between fund performance and manager performance is not a trivial one. They then examine cross-sectional performance, instead of searching for correlations in fund returns over time.

During the time period they study, there is a strong correlation between fund returns and a manager's age, the average SAT score of his or her undergraduate school, and whether he or she holds an MBA.

To understand why these patterns exist, Chevalier and Ellison examine several factors that might explain them. These include fund size, expenses, differences in systematic risk (based on the return on a value-weighted composite index of the New York Stock Exchange, the American Stock Exchange, and the Nasdaq stock market, minus the risk-free return), management style (such as a higher propensity to purchase stock with above average market-to-book value ratios), and survivorship biases created by the absence of data on funds that may have been liquidated or merged into other funds. These factors appear to account for at least some of the perceived superior performance of certain managers. The advantage apparently enjoyed by MBA holders, for example, is almost completely explained by differences in systematic risk and managerial style: that is, by the higher willingness of managers with MBAs to hold so-called "glamour stocks," those with high market-to-book value ratios. A large part of the superior results achieved by younger managers can be attributed to the fact that they are more likely than older managers to work for funds that charge lower expenses, and to the survivorship biases inherent in the data.

Still, Chevalier and Ellison find that there are substantial differences between funds managed by managers from low- versus high-SAT schools that cannot be explained by differences in behavior, expenses, or survivorship. For example, a manager who attended Princeton (the school with the fourth highest SAT score in the sample) would be expected to outperform a manager who attended the University of Florida (with an SAT score close to the sample mean) by 60 basis points per year. But Chevalier and Ellison caution that this result could reflect a tendency of high-SAT managers to work for funds which have lower unreported expenses, better support staff, or superior compensation packages that induce greater work effort.

Chevalier and Ellison use data compiled by Morningstar Inc., a major publisher of information on the fund industry, to construct a sample of growth and growth-and-income funds that includes information on asset size, monthly returns, and expense ratios for 1988-94. Based on personal data also provided by Morningstar, they are able to estimate the age of fund managers and their educational qualifications, including degrees and the undergraduate schools they attended. Composite SAT scores are then calculated for those institutions.