Research Investigation of Hedge Funds Performance Persistence

Abstract

This article examined the performance persistence of 736 hedge funds monthly observations of nine style categories, as they include the largest number of hedge funds measured from January 1990 to January 2003. The sample is free from survivorship bias as it includes funds that has terminated, merged or unitized. The whole sample consists of 773 hedge funds. The methodology that we have used is contingency tables. In order to examine performance persistence we rank the returns of hedge funds from the period 1990 to 2003 using average returns figures. We categorize the funds based on their returns as winners and losers using positive and negative return values. In each category, we mention the total number of funds. Funds based on their total returns were classified as winners – winners, WW, losers – losers, LL, winners – losers, WL and losers-winners, LW. Our methodology is similar to the one used by Tonks (2000), Allen and Tan (1999) and Carpenter and Lynch (1999). The persistence of winners and losers according to the above results could be justified by the skill of the fund managers. The results are mixed. There is evidence of strong persistence of winners – winners, WW and losers-losers, LL over 1-year period. In contrast, winners –losers and losers-winners show mixed results of weak persistence over 1-year period. Winners’ hedge funds that continue to be winners in the next year shows that fund managers have stock picking ability, skill and not luck. In contrast, losers’ hedge funds that continues to be losers in the next year shows that fund manager’s lack of skill and stock picking ability.