On the robustness of persistence in mutual fund performance

Abstract

This paper analyzes persistence in US equity mutual fund performance over the period 1990–2015. We apply commonly used measures of persistence, which we test using a set of simulated passive funds. In the first stage we apply contingency tables and transition matrices in accordance with previous literature. Results show how these methodologies are biased towards finding evidence of persistence too easily. In the second stage, we take a recursive portfolio approach, which assesses the performance of investing by following recommendations based on past performance. Results show the importance of both estimating persistence by distinguishing among fund style groups, and considering the cross-sectional significance of recursive portfolios. In general, our results support evidence of persistence in mutual fund performance, especially for the case of the best mutual funds. However, this evidence does not hold for the most recent subperiod, 2008–2015. Empirical evidence of persistence is conditioned by the sample period, a result that could explain the inconclusive results found in the literature.