Mutual fund performance: the model for selecting persistent winners

Standard Fama-French-Carhart (FFC) models are widely used by academics to assess risk-adjusted fund performance versus market, size, style and momentum factors. However, they fail to reflect the industry standard, following which the performance of money managers is commonly evaluated relative to a corresponding benchmark and the peer group. In this paper, we introduce a new approach that augments the Carhart model and enables investors to identify the funds that outbid both the benchmark and the peer group. In addition, it allows discovering more certain winners by eliminating the under/outperformance of funds driven by the bias in the FFC factor construction. The application of our model is illustrated on Large Cap Value and Large Cap Growth US active equity mutual funds using contingency tables. The performance and persistence in performance are assessed by comparing our novel and the standard Carhart models. Our model identifies more winners than the Carhart; those winners earn higher returns net of benchmark and peer- group than the Carhart’s winners, and show persistence in performance 36 months ahead. The results are robust to different specifications of contingency tables, holding periods or style categories of funds.