A Metaheuristic Approach to Multiobjective Portfolio Selection in Tehran Stock Exchange (TSE)

This paper presents a novel metaheuristic method for solving an extended Markowitz portfolio selection model. In the extended model, the objective function has been modified to include realistic objectives and four additional sets of constraints, i.e., bounds on holdings, cardinality, minimum transaction lots, and liquidity constraints have been also included. The first set of constraints guarantee that the amount invested (if any) in each asset is between its predetermined upper and lower bounds. The cardinality constraint ensures that the total number of assets selected in the portfolio is equal to a predefined number. The liquidity constraints reflect the investor's tendency to invest on those stocks that are more quickly tradeable.
The extended model is classified as a multi-objective mixed-integer programming model necessitating the use of efficient heuristics to find the solution. In this paper, we propose a heuristic based on pareto combined ant colony optimization and simulated annealing approaches. The performance of the proposed approach is compared to some other approaches using Tehran Stock Exchange data. The computational results show that the proposed approach effectively outperforms other approaches subject to the computationtime needed and the quality of the obtained solutions especially in large-scale problems.