Optimizing Stock Portfolio with regard to Minimum Level of Total Risk using Genetic Algorithm

Risk and return are two main factors that have always been considered in the field of investment. Simultaneously with the advent of different models for portfolio optimization which the Markowitz model is the most important of those, the necessity to identify methods for solving these models gained great Importance. Genetic Algorithm is one of the most important metaheuristic methods used for the solution of the portfolio optimization models.This study aimed at evaluating the level of efficiency of this metaheuristic model in portfolio optimization. Therefore, in this study once we have calculated the optimal efficient frontier by the use of the genetic algorithm, and then we compared this optimal efficient frontier with the efficient frontier which was obtained through exact solution method. To achieve this purpose, 25 companies were selected from companies in Tehran Stock Exchange.
The results of our study shows that the optimal efficient frontier gained through genetic algorithm is equal to the efficient frontier obtained using the exact solution method, and thereby indicating the high efficiency of genetic algorithm in portfolio optimization. The other result of the present study is that the comparison of the optimal portfolio gained through exact solution with the systematic and unsystematic risk, also revealed that Stock diversity in portfolios with unsystematic risk is much greater than portfolios with systematic risk.