Building a Hedge Fund Portfolio with Kurtosis and Skewness

Abstract

Mean-Variance optimization has long played an important role in portfolio construction. These traditional methods have been applied to hedge funds, but recent research indicates that hedge fund return distributions are distinctly non-normal. That is, hedge fund return distributions exhibit considerable leptokurtosis and skewness. Consequently, traditional mean-variance optimization will lead to sub-optimal results. This article incorporates skewness and kurtosis into the hedge fund selection process and applies it to a “live” portfolio of hedge funds. The results show that multi-moment optimizations produced superior portfolios with higher Sharpe Ratios while explicitly accounting for kurtosis and skewness.