Avaliação de fundos de investimento utilizando o Índice de Sharpe AR-GARCH-M

Abstract

The Sharpe ratio, a well known and widely used measure for ex post facto performance evaluation of diversified financial asset portfolios, has an important limitation, the assumption of normally and independently distributed (NID) daily excess log-returns. This paper proposes an econometric Sharpe ratio and an estimation method derived from an AR-GARCH-M model. The proposed Sharpe ratio is compared to the standard Sharpe ratio. Daily log-returns of mutual funds in the Brazilian market were used and Sharpe ratios were estimated for each. The Spearman rank correlation coefficient between proposed and standard Sharpe ratios was found to be positive, but the ratios led to different conclusions about the performance of the funds. The proposed Sharpe ratio relaxes the NID hypothesis for log-returns and includes as stylized facts conditional heteroskedasticity, autocorrelation and dependence between log-return and