SIMHEURISTIC AND LEARNHEURISTIC FOR SOLVING STOCHASTIC AND/OR DYNAMIC PORTFOLIO OPTIMIZATION PROBLEMS

Abstract

Constructing portfolio by proper asset selection to maximize return and minimize risk has been considered an essential task for investment activities. Rich portfolio optimizations with realistic constraints are NP-hard problems and are commonly solved using metaheuristics. However, financial markets are characterized by their high volatility and uncertainty, and metaheuristics do not fully account for these random and/or dynamic components, which renders them unrealistic in the presence of heightened uncertainty and dynamism in financial markets. Therefore, this paper proposes a simulationoptimization approach specifically, a simheuristic algorithm to deal with the stochastic version of the problem and a learnheuristic algorithm for solving the dynamic version of the problem. Computational experiments are performed on a benchmark instance to illustrate the advantages of the proposed methodologies and analyze how the solutions change in response to a different degree of stochasticity, dynamism, and minimum required return. © SW 2023.All rights reserved