On Persistence of Performance Measures

In this paper, we analyze the persistence of several performance measures. Our work is radically different from others in the sense that we are not interested in evaluating if there is persistence in the performance of managed portfolios such as Mutual Funds but rather if performance measures have themselves some degree of persistence, that is, we focus on the inherent persistence biases of the measures. The main conclusions can be summarized as follows: i) Only three performance measures (Sharpe ratio, Reward-to-Semivariability, and Graham-Harvey measure) show some degree of persistence for almost all horizons and information sets. This means that the researcher should take into account that there is a bias in these measures to evaluate the true performance of the manager. ii) Performance measures seem to show approximately the same degree of persistence in a short term than in a long term. iii) With some performance measures, there is a high probability of reversals in performances. JEL Classifications: G23; C15; C63