Genetic Algorithm with an Application to Complex Portfolio Selection

Abstract

In this paper, a realistic portfolio selection problem is studied and genetic algorithm is designed to solve the corresponding quadratic mixed-integer problem. At first, a new portfolio selection model, as an alternative to the standard Markowitz model, is formulated for selecting portfolios with transaction costs and transaction roundlot constraint. In addition, due to these complex constraints traditional optimization algorithms fail to work efficiently and heuristic algorithms can be the best method, so a genetic algorithm is designed to solve our proposed problem. Finally, a numerical example is given to illustrate our proposed effective model and method.