Optimum portfolio selection using a hybrid genetic algorithm and analytic hierarchy process

Purpose - The purpose of this paper is to present a multi-objective model to the optimum portfolio selection using genetic algorithm and analytic hierarchy process (AHP). Portfolio selection is a multi-objective decision-making problem in financial management. Design/methodology/approach - The proposed approach solves the problem in two stages. In the first stage, the portfolio selection problem is formulated as a zero-one mathematical programming model to optimize two objectives, namely, return and risk. A genetic algorithm (GA) with multiple fitness functions called as Multiple Fitness Functions Genetic Algorithm is applied to solve the formulated model. The proposed GA results in several non-dominated portfolios being in the Pareto (efficient) frontier. A decision-making approach based on AHP is then used in the second stage to select the portfolio from among the solutions obtained by GA which satisfies a decision-maker's interests at most. Findings - The proposed decision-making system enables an investor to find a portfolio which suits for his/her expectations at most. The main advantage of the proposed method is to provide prima-facie information about the optimal portfolios lying on the efficient frontier and thus helps investors to decide the appropriate investment alternatives. Originality/value - The value of the paper is due to its comprehensiveness in which seven criteria are taken into account in the selection of a portfolio including return, risk, beta ratio, liquidity ratio, reward to variability ratio, Treynor's ratio and Jensen's alpha.