An experimental metaheuristic approach for portfolio optimization problem with cardinality constraint

Abstract

The addition of cardinality constraint to the standard mean-variance portfolio optimization problem makes it investor-friendly. However, by adding the cardinality constraint, the exact methods fail in solving the portfolio optimization problem. Cardinality Constrained Portfolio Optimization Problem (CCPOP) is solved using two metaheuristics approaches in this work for a collection of thirty-one assets. Further, the two metaheuristic techniques are analyzed using seven performance metrics in their effectiveness to solve the CCPOP. Also, the changes in the efficient frontier on varying the value of the cardinality constraint are analyzed. The results are presented in tabular form, and a conclusion is drawn on the comparison of the metaheuristic techniques in solving CCPOP.