Fund performance raises specific issues

Unlike the returns of common stocks and mutual funds, hedge fund returns are generally not normally distributed.1,i This has considerable consequences on a number of hedge fund risk measures, as presented in last month’s article, which was drawn from the EDHEC Hedge Fund Reporting Survey2 conducted with the support of Newedge Prime Brokerage.

Simple measures, such as the Sharpe ratio,3 which are based on the normality assumption, are then not valid for assessing the performance and riskiness of a hedge fund. The same applies to all value-at-risk (VaR) measures that assume normally distributed returns. Non-normality of hedge fund returns is seen easily with statistical tests for normality. The most well-known test is the Jarque-Bera test.4 Another popular one is the Lilliefors test,5 which is in fact an adaptation of the Kolmogorov