Multi-stage stochastic mean–semivariance–CVaR portfolio optimization under transaction costs

Abstract

In this paper, we present a model for portfolio selection, characterized on the basis of three parameters: the expected value, semivariance, and Conditional Value-at-Risk (CVaR) at a specified confidence level. In order to solve the proposed model, we design a hybrid of genetic algorithm (GA) and particle swarm optimization (PSO) algorithm. Because the effectiveness of meta-heuristic algorithms significantly depends on the proper choice of parameters, a Taguchi experimental design method is applied to set the suitable values of parameters to improve the hybrid algorithm performance. Finally, some numerical examples are given to illustrate the effectiveness of the proposed algorithm.