Heidorn, Thomas Kaiser, Dieter G. Muschiol, Andrea
Year of Publication:
Working paper series // Frankfurt School of Finance & Management 77
Hedge Funds are often considered as a possibility for optimizing traditional portfolios due to their alternative risk factors and sources of return. But as the return distribution of hedge funds shows negative skewness and excess kurtosis, using portfolio optimization techniques, based on the Markowitz mean variance framework, may lead to an overestimation of the op-timal allocation to hedge funds. In this study we use Polynomial Goal Programming (PGP) to incorporate higher moments of the distribution into the optimization. This article is the first to use the enhanced PGP method which incorporates investors' preferences to optimize a tradi-tional portfolio consisting of equities, bonds and commodities. Even when considering higher moments we can show that hedge funds can be useful in enhancing the return and reducing the risk of the overall portfolio. Moreover, we find evidence that applying the mean variance framework to hedge funds does not result in overestimated hedge fund allocations. On the other hand PGP leads to totally different asset allocations on the strategy level.