The transient and persistent efficiency of hedge funds

During the 1980s and 1990s hedge managers generated excellent performance, drawing the attention of wealthy investors who were willing to pay high fees for high returns (typically an annual fee of 2% of funds under management and 20% of performance). In light of the performance of hedge funds, investment dollars flooded into hedge funds and many new hedge funds opened. This expansion of the industry reduced both the investment opportunities and the quality of managers available to funds; both factors should have reduced mean returns. Furthermore, the efficient market hypothesis (EMH) argues that it is not possible for a fund to persistently earn above average returns. The literature is unsettled as to whether hedge funds returns are persistent. For example, Agarwal and Naik (2000) argue that persistence in hedge funds performance exists while Capocci and Hubner (2004) found limited evidence of persistent performance. In this paper we use the stochastic frontier model of Filippini and Greene (2016) to estimate the transient and persistent efficiency of hedge funds. The estimated frontier is a random effects model with two skewed normal error components, which allows the transient and persistent efficiency of observations to be identified. If the persistent component of inefficiency is zero, then hedge fund managers do not consistently beat the market. We hypothesize that (1) mean efficiency declined as the industry grew, (2) “star” funds experienced transient efficiency, and (3) no funds were persistently efficient. A large sample of hedge funds over almost 20 years is used to estimate our model and test our hypotheses. Robustness checks include using different measure of hedge fund performance and different sub-samples of funds.