ALGORITMOS GENÉTICOS PARA OTIMIZAÇÃO DE PORTFÓLIO DE INVESTIMENTO

Abstract

Cash equity does not mean long-term financial protection. Some currencies can depreciate and thus investing them becomes a good option. In addition to protecting, the investor can make a profit, that is, increase the amount of capital invested. With technological progress, the number of investment opportunities has increased dramatically, making investors face two questions: “In which assets should I invest? And how much should I invest in them? ”. Therefore, investors and researchers aim to develop models using mathematical/computational tools to solve such questions. Markowitz (1952) says that the first question is more related to the investor’s behavior, while for the latter statistical concepts can be used for solution. The present work aims to answer the questions, bringing a solution to the investor, embedding real features in theoretical models. Furthermore, we propose a Genetic Algorithm meta-heuristic to solve the same models with real embedded features, as long as they have a high computational complexity. The proposed algorithm presents good results in reduced computational time.