|
Why do hedge funds stop reporting performance? Authors: A. Grecu, B. G. Malkiel, and A. Saha Source: The Journal of Portfolio Management Date: Fall 2007 |
Why do hedge funds stop reporting performance? According to Grecu, Malkiel, and Saha, there are two possible explanations. First, funds stop reporting when they underperform their peers. Second, they stop when they do not need to attract additional capital. In fact, hedge funds can stop reporting because performance reporting is not mandatory. The Securities and Exchange Commission, for example, does not require it.
The authors begin with a simple statement: studies that provide a measure of the survivorship bias report a positive bias, ranging from 0.16% to 4.50%. These figures might indicate that funds that stop reporting underperform other funds. Consequently, the authors opt for the first explanation—funds stop reporting when they underperform their peers.
A comparison is made between the performance over the last months before the exit from the TASS database and the performance prior to this period. Data come from January 1996 to April 2004. Whatever the length of the period before funds exit (3, 6, or 9 months), and whatever the performance indicator (geometric mean of the fund’s return or Sharpe ratio), the performance during the last months is lower than in the prior period. In other words, before they stop reporting, funds experience declining performance, and poor performance leads hedge funds to stop reporting. For this reason, the authors discount the second possible explanation—that funds stop reporting when they do not need to attract additional capital.
Following a study by Gregoriou (2002),{1} Grecu et al. scrutinise hedge funds’ time to failure. Here, failure occurs when a fund stops reporting. They underline an important difference in the methodology. While Gregoriou uses a Weibull distribution to model the change in failure rates over time, Grecu et al. use a log-logistic distribution. This choice is motivated by the log-likelihood values provided by Akaike’s information criterion. The log-logistic distribution better fits inverted and non-inverted U-shaped hazard functions. After a rapid rise the hazard rate reaches a peak at the 66th month with 1% per month, and after this peak it falls gradually but remains high (from the 66th month to the 120th month, the exit risk falls by only 11.73%).
In the last part of the article, Grecu et al. attempt to identify factors impacting the likelihood of failure. Two models—a Cox semiparametric hazard model and a parametric log-logistic survival model—are used. Both models exhibit consistent results for Sharpe ratios, volatility, and assets under management taken as factors. It turns out that high Sharpe ratios are associated with a lower risk of failure, higher volatility increases the risk of failure, and larger funds display a lower risk of failure. The model results confirm the authors' expectations, and, as a result, they conclude that when hedge funds stop reporting performance, it isn't because they are successful.
Footnotes
{1} Gregoriou, G. 2002. Hedge fund survival times. Journal of Portfolio Management 3 (3): 237-252.
