Multi-objective indicator based evolutionary algorithm for portfolio optimization

Abstract

Portfolio optimization is a standard problem in the financial world for making investment decisions which involve investing into a variety of assets with the aim of maximizing yield and minimizing risk. Modern portfolio theory is a mathematical approach to the problem that endeavors to accomplish a plausive portfolio by giving best weighting of the assets. In this study, an indicator based evolutionary algorithm (IBEA) has been compared with two well known evolutionary algorithms-Non-dominated Sorting Genetic Algorithm II( NSGA- II) and Strength Pareto Evolutionary Algorithm (SPEA-II).The results reveal that IBEA outperforms the other two algorithms in terms of its closeness to the true pareto front. Also, a diversity enhanced version of IBEA (IBEA-D) is proposed, which is found to be providing more diverse solutions than IBEA.