Detecting structural breaks and identifying risk factors in hedge fund returns: A Bayesian approach

An advanced perspective on the predictability in hedge fund returns

“…This result contradicts to the recent study of Meligkotsidou and Vrontos (2008) who find that accounting for break points improves predictive accuracy. At first, this finding seems surprising since our in-sample CUSUM tests detect break points in time periods also spotted in previous studies.…”

“…Regarding the ''Long/Short Equity" strategy our results are mostly in line with the breaks reported by Meligkotsidou and Vrontos (2008), who report that the HFR equity hedge index is subject to structural breaks in the end of 1998 and, depending on the method used, somewhere in the course of 2000. Also corresponding to the results of the mentioned study, our CUSUM test detects a structural break at the end of 1999 for ''Emerging Markets" hedge funds.…”

“…, s}. Also notice that our method adjusts regression parameters based on the segment of data after the last break date z, as opposed to the break point models applied in previous hedge fund studies of Meligkotsidou and Vrontos (2008) or Fung et al (2008), which for each time period in between breaks calibrate independent models. But in view of the fact that the size of the interim segments might be very short, we believe that the concept of adjusting rather than substituting model parameters is more accurate.…”

“…In their hedge fund pricing studies Fung and Hsieh (2004) and Fung et al (2008) illustrate that structural breaks entail changes in regression coefficients and their significance. Also Meligkotsidou and Vrontos (2008) have identified and considered break points for the derivation of monthly hedge fund return forecasts. However, all these studies identify structural breaks based on a singular analysis of all available return observations.…”

The performance of hedge funds and mutual funds in emerging markets

“…In order to analyze extreme market events and changing return patterns more closely, we follow Fung and Hsieh (2004) and Fung et al (2008) and use a modified CUSUM test to find structural breakpoints in factor loadings (see Meligkotsidou and Vrontos (2008) for a more detailed analysis of structural breaks in hedge fund returns). Fung and Hsieh (2004) as well as Fung et al (2008) find that structural breaks coincide with extreme market events (in their case the collapse of Long-Term Capital Management in September 1998 and the peak of the technology bubble in March 2000) and conclude that these events might affect managers' risktaking behavior.…”