Opposition-Based Modified Imperialist Competitive Algorithm for Portfolio Optimization

Portfolio Optimization is a well known problem in area of finance and its aim is to minimize the risk of a portfolio containing different amount of assets. Different model with non-linear objective functions and constraint have been developed for this problem. As it is very time consuming procedure to solve these kinds of models, meta-heuristic methods can be applied here. In this work we use Imperialist Competitive Algorithm (ICA), a new meta-heuristic method with respect to multiobjective optimization principals to solve this problem. We also used a new scheme in ICA, known as Opposition-Based Learning, to speed up convergence of this algorithm and improve its accuracy. A new creativity that can be seen here is that, we try to improve the characteristic of imperialists to increase our chance for getting better solution. At last, we apply our model for 20 selected stocks from Tehran Stocks Exchange. All results show improvement in our proposed algorithm, call it Opposition- Based Modified ICA (OBMICA), in comparison with ICA