Generalized Tests of Investment Fund Performance

Authors

  • Márcio Poletti Laurini Departament of Economics - FEA-RP USP and associated researcher at CNPq
  • Antônio Zoratto Sanvicente Insper Institute of Education and Research
  • Rogério da Costa Monteiro Insper Institute of Education and Research

DOI:

https://doi.org/10.12660/bre.v31n22011.7173

Keywords:

Sharpe Ratio, GMM, Investment Analysis

Abstract

The paper discusses the use of statistical methods in the comparison ofinvestment fund peThe paper discusses the use of statistical methods in the comparison ofinvestment fund performance indicators. The analysis is based on the robuststatistics proposed by Ledoit and Wolf (2008), for the pairwise comparison offunds and two generalizations for sets of multiple investment funds. The multipleinvestment fund tests use the Wald and Distance Metric statistics, based onestimation by Generalized Method of Moments using HAC matrices. In orderto correct size limitations in the GMMestimation in the case of a large number ofmoment conditions, the test distributions are obtained through block-bootstrapprocedures. We applied the proposed procedures to daily return data for thelargest 97 actively managed equity funds in the Brazilian market, covering theperiod from July 2006 to July 2008. The results indicate that there are nosignificant differences in the performances of the 97 funds in the sample, bothin pairwise and joint comparisons, thus providing what is believed to be the firstBrazilian market evidence for the so-called herding hypothesis. rformance indicators. The analysis is based on the robust statistics proposed by Ledoit and Wolf (2008), for the pairwise comparison of funds and two generalizations for sets of multiple investment funds. The multiple investment fund tests use the Wald and Distance Metric statistics, based on estimation by Generalized Method of Moments using HAC matrices.In order to correct size limitations in the GMM estimation in the case of a large number of moment conditions, the test distributions are obtained through block-bootstrap procedures. We applied the proposed procedures to daily return data for the largest 97 actively managed equity funds in the Brazilian market, covering the period from July 2006 to July 2008. The results indicate that there are no significant differences in the performances of the 97 funds in the sample, both in pairwise and joint comparisons, thus providing what is believed to be the first Brazilian market evidence for the so-called herding hypothesis.

Downloads

Published

2011-12-02

Issue

Section

Articles