Otimização de portfólio de investimento com análise envoltória de dados e Markowitz

Abstract

The building process of an investment portfolio is one of the biggest challenges of the stock market. Balancing assets with excellent perspectives of valuation alongside low risk is a considerable challenge. To assemble a portfolio with good revenues requires much study because the stock market is variable and dependent on many factors, since worldwide economical factors levels to the investor’s emotions. To bypass the investor’s emotions some mathematical techniques were developed during the years with Harry Markowitz being a pioneer. This essay studied the model called mean-variance developed by Markowitz, combining Data Envelopment Analysis to compose different portfolios, under different aspects, as a different periods of analysis and different levels of risk acceptance with the intention to evaluate the performance of these portfolios. After the development of the models for the construction of eight portfolios, it was used statistics comparison tests. The conclusion is the utilization of such techniques for the construction e comparison of the portfolios should efficient for the investor, being possible to identify portfolios with meaningful differences in their revenues depending on the longevity of the data analyzed, the level of risk acceptance, and the technique used. Being the portfolio the one with the best performance that combined both optimization techniques, with the lower risk level that is possible and analyzing the latest data.