Brain Storm Optimization for Portfolio Optimization

In this paper, Brain storm optimization (BSO) algorithm is employed to solve portfolio optimization (PO) problem with transaction fee and no short sales. In addition, simplified BSO (SBSO) and BSO in objective space (BSO-OS) are also utilized in the same model. The potential portfolio proportion is regarded as ideas that individual generated. In the experimental study, three cases with different risk aversion factors are considered. Simulation results demonstrate that both original BSO and modified BSOs obviously outperform PSO and BFO in PO problem. In particular, BSO-OS, which not only saves the computation time but also finds the optimal value, shows an extraordinary performance in all set of test data.