Evaluating Mutual Fund Performance

We study standard mutual fund performance measures, using simulated funds whose characteristics mimic actual funds. We find that performance measures used in previous mutual fund research have little ability to detect economically large magnitudes ~e.g., three percent per year! of abnormal fund performance, particularly if a fund’s style characteristics differ from those of the value-weighted market portfolio. Power can be substantially improved, however, using event-study procedures that analyze a fund’s stock trades. These procedures are feasible using timeseries data sets on mutual fund portfolio holdings. THIS PAPER USES simulation procedures to study empirical properties of performance measures for mutual funds ~i.e., managed equity portfolios!. Recent studies of mutual fund returns ~e.g., Carhart ~1997!, Daniel et al. ~1997!! have moved beyond performance measures based on the Capital Asset Pricing Model, such as “Jensen alpha.” These studies account for nonmarket factors in the cross section of expected returns, such as size, book-to-market, and momentum. Applying multifactor benchmarks to performance measurement has been characterized as “simple” and “straightforward” ~Fama and French ~1993!, p. 54!. The basis for this view is that multifactor benchmarks are crosssectionally well specified. However, the power of multifactor benchmarks to detect abnormal performance of a managed portfolio has received little attention. Further, the specific method for implementing a multifactor benchmark could also affect the power of the tests. For example, there is reason to believe that regression-based benchmarks ~e.g., four-factor alpha! will have lower power than characteristic-based benchmarks, which form comparison portfolios using information on fund holdings ~Daniel et al. ~1997!!. However, the magnitude of the difference is an unresolved empirical issue. * Sloan School of Management, Massachusetts Institute of Technology and William E. Simon Graduate School of Business Administration, University of Rochester, respectively. We thank Doug Breeden, Charles Nelson, Wayne Ferson, Bill Schwert, Cliff Smith, René Stulz, two anonymous referees, seminar participants ~Rochester, Colorado ~Burridge Center Annual Conference! and Duke! for their comments, and Andreas Gintschel, Peter Wysocki, and Tzachi Zach for excellent research assistance. We are grateful to the Research Foundation of the Institute of Chartered Financial Analysts and the Association for Investment Management and Research, the Bradley Policy Research Center at the Simon School, and the John M. Olin Foundation for financial support. S.P. Kothari acknowledges financial support from the New Economy Value Research Lab at the MIT Sloan School of Management. THE JOURNAL OF FINANCE • VOL. LVI, NO. 5 • OCT. 2001