Large-scale portfolio optimization using multiobjective dynamic mutli-swarm particle swarm optimizer

Portfolio optimization problems involve selection of different assets to invest so that the investor is able to maximize the overall return and minimize the overall risk. The complexity of an asset allocation problem increases with the increasing number of assets available for investing. When the number of assets/stocks increase to several hundred, it is difficult for classical method to optimize (construct a good portfolio). In this paper, the Multi-objective Dynamic Multi-Swarm Particle Swarm Optimizer is employed to solve a portfolio optimization problem with 500 assets (stocks). The results obtained by the proposed method are compared several other optimization methods. The experimental results show that this approach is efficient and confirms its potential to solve the large scale portfolio optimization problem.