Aplicação das metodologias DEA (Data Envelopment Analysis) e SuperDEA na análise de eficiência de investimentos em ações negociadas na bolsa BM&F BOVESPA

Abstract

The objective of this research to outline the characteristics of weighting strategies for selecting stock portfolios using data envelopment analysis methodology (VRS / BCC). It used the Data Anvelopment Analysis (DEA) as a tool for analysis in the selection of investment portfolios based on stocks traded on Bovespa stock exchange. This methodology is based on non-parametric mathematical models that evaluate the performance of each Decision Making Unit (DMU) with a multidimensional perspective. This research selected a portfolio composed of 20 stocks, for which we obtained the efficiency indicators, variance, return and others, allowing the portfolio selection and comparison of different DEA portfolio weighting strategies (Naive 1/N, Markowitz and Sharpe) and SuperDEA. For purposes of analysis, we applied the Data Envelopment Analysis (DEA) method with the input oriented BCC model (BANKER; CHARNES; COOPER, 1984). The results showed a superiority of VRS DEA model in conjunction with the SuperDEA weighting method, as the cumulative weekly continuous returns between January 2011 and December 2014, as the lower values of average monthly volatility. It can be concluded that the adoption of DEA focused on portfolio management, together with the SuperDEA weighting methodology can be considered as effective in the portfolio optimization process in the financial market.