Avaliação de performance de fundos de investimento no contexto brasileiro

Abstract

The international literature in finance offers a wide range of models for performance evaluation of individual assets, portfolios and, especially, investment funds. However, in the Brazilian academia, investment funds have received little attention from researchers and few production has been made on the subject. The objective of this work is to contribute to research on the topic, applying the major models for performance evaluation to ascertain the performance of Brazilian investment funds, as well as determine which variables affect their returns, which management style provides higher return, which benchmark is more attractive and also to study the influence of size effect and the ability of the manager. First, it was built a theoretical framework based on key international and national studies, with the aim of present basic concepts surrounding the subject and justify the use of the models, as well as providing a comparison with results found by previous studies. The theoretical framework also seeks to situate the reader within a historical perspective. The data used were gently supplied by the Associação das Entidades dos Mercados Financeiros e de Capitais (ANBIMA) and passed through filtering procedures to exclude funds with few observations, and merge with the time series from other sources, to ultimately create the variables that were used. The data refer to actively managed funds and passive benchmarks which are Ibovespa, IBrX and ISE. The models and variables used will be commented in the chapter of methodology. The results chapter starts with the descriptive statistics about the data and then bring the fruits of the appliance of theoretical models worked on, which are the performance indexes, the CAPM, the models that incorporate superior moments, the APT model and the model to evaluate the size effect and the comparison between different types of benchmarks and management styles. The results analysis led to the conclusions, showing that all performance indexes yielded similar results, which indicate that passively managed funds whose benchmark is the IBrX has superior return. Models that include co-skewness and co-kurtosis in the CAPM model are important, but it appears that using only the co-skewness may be more interesting than using jointly co-kurtosis. The APT model shows that macroeconomic variables are significant in explaining the returns of the funds, bringing the surprise that industrial production has negative relationship with such returns. No evidence was found that the equity size affects its financial results, but the analysis of funds whose benchmark is the ISE shows clearly that such investments generate returns inferior to the other.