Optimal Portfolio Allocation Using Funds of Hedge Funds

Abstract

This paper compares different methods of optimization for a portfolio allocation that includes funds of funds. Optimization consists of minimizing risk measured by one of the following proxies: normal Value at Risk (VaR), adjusted VaR (adjusted using the Cornish-Fisher expansion), weighted historical simulation VaR, and semi-deviation. Results indicate that compared to the other proxies of VaR, normal VaR tends to underestimate portfolio risk. Moreover funds of funds improve the risk-return profile of the portfolio. This last result is interesting since funds of hedge funds exhibit less of the individual hedge funds' biases reported in the literature.