An analysis of patterns in the reported returns of hedge funds can be used to assist in the identification of funds that are engaged in such reporting violations as overvaluation and misrepresentation or that are running Ponzi schemes.
The authors identify a number of performance flags based on the patterns of reported hedge fund returns. A survey of hedge funds charged with regulatory breaches or investor lawsuits shows that such funds were more likely to have reported performance that triggered these flags than were hedge funds with an unblemished regulatory record.
How Is This Research Useful to Practitioners?
Such agencies as the U.S. SEC do not have adequate resources to deter hedge funds from reporting fraudulent performance figures. The flags identified by the authors provide a relatively simple, low-cost method of screening hedge funds to identify those funds that are possibly misreporting performance. Many of the flags identified are designed to capture such unreasonable incentives of hedge fund managers as a perceived need to report a low number of negative returns.
For current and potential investors, funds that are flagged as suspicious may require further detailed investigation to better understand the drivers of the performance. In addition, funds may need to confirm the strength of their reporting and governance frameworks as well as provide evidence of clean audit reports.
How Did the Authors Conduct This Research?
The authors identify a series of flags based on patterns or characteristics of reported performance over the period from January 1994 to December 2008. Flags based on return characteristics include (1) various correlation measures between reported performance and style factors, (2) serial correlations of returns, and (3) distribution of returns—in particular, “kinks” in the distribution for negative returns. Flags based on patterns include the number of negative returns, returns equal to zero, unique returns, recurring pairs of returns, and a test of the distribution of the last reported digit of returns. Using a combination of statistical analysis and simulation, the authors calculate trigger points for each flag. For example, a fund triggers the percent negative flag when the probability of generating the observed number of negative returns is less than 10%.
They create a dataset of reporting and trading violations based on SEC violations and investor lawsuits. Reporting violations are classified separately from such trading violations as illegal short selling or trading on insider information, which do not necessarily lead to fraudulent reporting. For the selected time period, the authors extract hedge fund returns from the CISDM and TASS databases. Funds are classified as being nonproblem funds or problem funds based on their regulatory record. Problem funds are further divided by reporting violation and trading violation.
The authors then compare the rejection rate of problem funds with that of nonproblem funds. For almost all flags, the percentage of problem funds that triggered flags is higher than the percentage of nonproblem funds that did so. For many, the difference in percentages is statistically significant. For problem funds, the most common flag triggered is the percentage of repeats (the number of times a reported return is repeated). The authors create a simple measure of the risk of fraud that incorporates all of the flags identified. Using this measure, they show that almost 50% of funds with a reporting violation had a risk score in excess of the 90th percentile score for nonproblem funds.
Although the authors concede the risk of false positives (the risk of misidentifying problem funds), the tools provide a useful screen for further investigation of suspicious funds.
Conducting due diligence on hedge funds is difficult. To maintain their perceived competitive advantage, hedge fund managers tend to limit the amount of proprietary information that they reveal to investors. This lack of transparency can make it difficult to assess reported performance. The performance flags introduced by the authors appear to be a strong differentiator of potentially fraudulent activity and are a useful addition to an investor’s toolkit. Of course, as the authors highlight, it is often a truism of finance that as soon as an anomaly is identified and reported, it quickly loses its information content. There is a risk that devious hedge fund managers will alter their deceptive practices to remain one step ahead of investors. But because many of the performance flags used by the authors mirror the incentives of hedge fund managers, it is likely that they will remain relevant.