Data envelopment analysis and fuzzy theory: efficiency evaluation under uncertainty in portfolio optimization

Abstract

This article aims to analyze the behavior of a portfolio selected through Data Envelopment Analysis (DEA) associated with fuzzy logic and optimized using the Sharpe approach. As a basis for comparison, two other portfolios were used, one obtained through only the Sharpe approach. In this research study, a fuzzy DEA model was used to evaluate efficiency under uncertainty of the Brazilian Stock Exchange-Bovespa, by means of input and output indicators such as return, variance, earnings per share and price-earnings. The study reliably identified which efficient stocks and which were most sensitive to the effect of uncertainty. Through the comparison of portfolios, it was observed that the resulting combination of the fuzzy DEA models in which the stocks were considered efficient in both scenarios presented the best results.