Static and Dynamic Asset Allocation with Higher Moments

Abstract

This paper illustrates a semi-parametric approach to static and dynamic asset allocation problems in terms of the moments of a multivariate distribution. By use of a general class of H-distributions, we reconstruct the portfolio density function from the moment sequence derived from the multivariate co-moments of its components. The information provided by these high order dependencies is used to capture tail dependency, diversification risk and asymmetries. The authors feel that the techniques suggested in this article offer a very useful set of tools for modern finance.