Hedge Fund Performance Persistence: A Multinomial Approach Application to Asian Hedge Funds

Abstract

In this paper, we study the performance persistence of 206 Asian long/short equity funds that are listed in the EurekaHedge database over two and a half year period, from January 2004 to June 2006. We employ two main methods used by Goetzmann and Ibbotson (1994) and Brown and Goetzmann (1995) for testing performance persistence. We also extend the two independent binomials known as 2x2-contingency table, to a multinomial contingency table with 3 and 4 dimensional periods. Results obtained from the multinomial contingency table have shown greater level of predictability. All of our tests are carried out on quarterly raw returns with all related statistical results. However, our results show that there is substantial persistence in performance of Asian long/short funds in both tests. Results obtained from multinomial framework do bring further insight the pattern of persistence in performance of funds, which provide a certain predictability power on funds performance. Our study has the following Implications: (1) investors can make their investment decision for the second quarter based on the performance of funds in the first quarter. (2) Based on the performance of a fund in the first quarter alone will not provide enough information for predicting its performance in the third or the fourth quarters ahead. (3) Performance of funds in the future is influenced by their performance in the past periods to a certain level of extent. Thus, based on their performance in only one quarter to predict their performance in the following three quarters might not be possible.

Keywords: Asian hedge funds, Performance evaluation, Performance persistence, Parametric and non-parametric methods, Hedge fund risks and returns

JEL Classification: G11, G15