The phrase “past performance is no guarantee of future results” (or some variation thereof) can be found in most funds’ literature, and for good reason: a wealth of studies show a lack of long-term performance persistence among actively managed mutual funds. However, many investors appear to believe that winners persist: past performance and related metrics remain important factors in manager selection.
Since 2002, S&P Dow Jones Indices has published the SPIVA® U.S. Scorecard, measuring the percentage of active managers that beat their benchmarks across various equity and fixed income categories. Its sister report, the Persistence Scorecard, shows the likelihood that a top-quartile manager maintains its status in subsequent periods.
By marrying the two reports, this paper studies the degree to which outperforming funds from one period continue to beat their benchmarks thereafter. Specifically, we first identify funds that beat their benchmarks, based on three-year annualized returns, net-of-fees. We then examine whether these funds (the “winners”) can continue to outperform during each of the next three one-year periods.
Our results show that among equity funds that beat their benchmarks over the three-year period ending Sept. 30, 2016, the performance persistence among domestic and international equity categories in the following three years was worse (in general) than a random draw. In other words, past performance did not typically help identify superior performing managers in advance.
DATA AND METHODOLOGY
The University of Chicago’s Center for Research and Security Prices (CRSP) Survivorship-Bias-Free US Mutual Fund Database serves as the underlying data source for our study. The universe used for the study only includes actively managed domestic U.S. equity funds. Index funds, sector funds, and index-based dynamic (leveraged or inverse) funds are excluded from the sample. To avoid double counting multiple share classes, only the share class with greater assets is used. At each measurement period, the universe consists of over 2,300 active equity funds, on average (see Appendix I).
Based on the earliest availability of Lipper style classifications, our study covers the period from March 31, 2000, through Sept. 30, 2019. On a quarterly basis beginning on March 31, 2003, we compute the trailing threeyear annualized returns for each fund in our universe, as well as for their benchmarks. We then identify funds that beat their benchmarks and track their relative performance in each of the next three years. By identifying funds that beat their benchmarks as winners and those that do not as losers, this approach applies the “winner-winner, winner-loser” methodology developed by Brown and Goetzmann (1995) and examines if winners in period t are also winners in t + j, where j = Year 1, Year 2, and Year 3.