A New Evolutionary Algorithm for Portfolio Optimization and Its Application

Value-at-Risk (VaR) and Conditional Value-at-Risk (CVaR) are two of the most widely used and important risk measures in financial risk management models. Because VaR and CVaR portfolio optimization models are often nonlinear and non-convex optimization models, traditional optimization methods usually can not get their global optimal solutions, instead, they often get a local optimal solution. In this paper, the uniform design is integrated into evolutionary algorithm to enhance the search ability of the evolutionary algorithm. The resulted algorithm will has a strong search ability and has more possibility to get the global optimal solution. Based on this idea, a new evolutionary algorithm is proposed for VaR and CVaR optimization models. Computer simulations on ten randomly chosen stocks from Shenzhen Stock Exchange in China are conducted and the analysis to the results is given. The experiment results indicate the proposed algorithm is efficient.