Assessing the impact of heteroskedasticity for evaluating hedge fund performance

Recently there have been a number of differing findings in the empirical evidence on fund performance. In this paper we suggest this difference could be explained by the treatment of the regression assumptions. The crucial question in this paper for investors is whether the presence of heteroskedasticity causes size distortion in testing fund performance. Our simulation findings indicate that heteroskedasticity can have significant impact on the evaluation of fund performance. We also apply a wild bootstrap approach to test a sample of hedge fund data. Our results suggest that one of the possible reasons for superior performance of hedge funds is that the bootstrap data generating process cannot fully account for heteroskedasticity. Overall, our results are consistent with the view that hedge funds are a heteroskedastic group and wild bootstrap is well suited to the performance measurement of hedge funds.