Commodity Futures in Asset Allocation: A New Approach with the Extreme Value Theory

Abstract

This paper presents an alternative approach to address the issue of diversification benefit of commodity futures. Instead of running the mean-variance optimization to obtain optimal capital weights to different asset classes, a number of portfolio sets is constructed by adding different fraction of commodity futures into traditional portfolio with stocks and bonds. These portfolio's VaR level and Reward-to-VaR ratio are estimated using the Extreme Value Theory. Adding commodity futures into traditional portfolio can reduce portfolio's VaR and improve portfolio's performance measured by Reward-to-VaR ratio. However, diversification benefit of commodity futures is not identical for different initial traditional portfolios. The main benefit of adding commodity futures to traditional portfolio with most bonds is expected return improvement, while the main benefit for traditional portfolio with most stocks is potential risk reduction. The more stocks are included in initial traditional portfolio, the more commodity futures can be added into the portfolio to reduce portfolio risk.