An evolutionary computation approach to scenario-based risk-return portfolio optimization for general risk measures

Abstract

Due to increasing complexity and non-convexity of financial engineering problems, biologically inspired heuristic algorithms gained significant importance especially in the area of financial decision optimization. In this paper, the stochastic scenario-based risk-return portfolio optimization problem is analyzed and solved with an evolutionary computation approach. The advantage of applying this approach is the creation of a common framework for an arbitrary set of loss distributionbased risk measures, regardless of their underlying structure. Numerical results for three of the most commonly used risk measures conclude the paper. © Springer-Verlag Berlin Heidelberg 2007.