ABSTRACT
This chapter provides a comprehensive overview of critical topics of predictive manager selection that include separation of luck from skill and performance persistence. It includes a framework for combining quantitative and qualitative factors within a Bayesian framework. It is unclear whether skill exists because luck can play a large role in outcomes. There are several issues associated with that. In statistics, inferences are made by comparing the value of a statistic to its distribution under the null hypothesis. A bootstrapping procedure creates a simulated distribution of the t-statistic of alpha that can be compared to the actual distribution to draw conclusions about existence of skill. Using the bootstrap approach, Kosowski, Naik, and Teo found strong evidence of positive skill relative to the Fung-Hsieh model among hedge fund managers. Hedge fund performance evaluation is very challenging because of having to rely on short track records to draw conclusions regarding skill.
