Unlike mutual funds, the reporting requirements of hedge funds, particularly offshore hedge funds, mean that managers often have a significant amount of discretion with regard to valuing their portfolios. So do some hedge fund managers take advantage of this and fudge returns? Researchers have uncovered several clues that suggest they do. Regular readers will recall this study, for example, showing that there is an odd lack of slightly-negative returns in the performance history of hedge funds. It’s almost as if managers exercised some “flexibility” in valuing their portfolios in order to nudge their returns above zero. For those managers, the benefits of not having a minus sign in front of their return outweighs the costs.
But what is the mechanism that leads to this anomaly and others like it? Do managers just goose the returns they report to databases and clients or it is more complicated? A new study explores how exactly “misreporting” occurs. Instead of looking for statistical anomalies in reported fund returns, Gjergji Cici of the College of William and Mary and Alexander Kempf & Alexander Puetz of the University of Cologne examine 13F filings to look for discrepancies between a manager’s (self-reported) 13F valuations and valuations from other sources. Specifically, the trio compared the stock prices reported by managers in their 13F filings with stock prices for the same securities reported by the Center for Research in Security Prices (CRSP).
They gathered 13F data and painstakingly linked it to the CISDM hedge fund database to isolate the actual hedge funds (they also used other sources to identify funds that did not report the CISDM). The process yielded over 2 million quarterly instances of hedge fund stock holdings at 1,025 funds between 1999 and 2008. Across all 2 million observations, the mean absolute difference between reported stock prices (via the funds’ 13F’s) and the CRSP stock prices for the exact same stocks was 0.22%. (See chart below created with data from Table II of the paper).
The trio concludes that “hedge fund advisors intentionally manipulate their equity valuations…” But is it fair to tar all hedge funds with the brush? Are all hedge fund holdings fudged? No. In fact, the study finds that in 2008, for example, over 95% of all instances of hedge fund stock holdings don’t have any stock price mismatch at all. The chart below created with data from Table II of the paper shows the proportion of hedge fund stock holdings that had such a mismatch. In fact, the proportion of mismatches seems to be falling. (So we appeal to the mainstream media covering this study: Please do not use the headline “Hedge Funds Manipulate Returns!” even if it sells more papers.)
Nonetheless, any misreporting is problematic and needs to be explored further. Cici, Kempf & Puetz wonder if part of this could be the result of managers exercising their (legal) discretion in self-valuing stocks that did not trade on the last day of the quarter. As you might expect the mean absolute stock price mismatch for this group was way higher (1.76% vs. 0.22% for the full population) and the proportion of holdings with a perfect match between 13F filings and CRSP data was a paltry 30% (i.e. 70% has non-zero deviations). The trio points out that these instances may all be totally legitimate and legal, but that since they only account for a tiny proportion of the overall population (2.6%), they don’t account for the “vast majority of observed valuation deviations.” (We do note, however, that 2.6% x 70% = 1.82% of the population; and since only 6.9% of the population had non-zero deviations, these legitimate deviations seem to still represent about a quarter of all non-zero deviations.)
But what about stocks that do trade on the last day of the quarter but are relatively illiquid. In these cases, write the authors, the hedge fund manager is “more likely to apply pricing discretion.” As you might guess, the less liquid the stock, the greater the absolute difference between 13F self-reported prices and CRSP prices. (See chart below showing liquidity deciles created with data from Table II of the paper).
What kinds of funds are likely to have non-zero deviations between their 13F filings and CRSP prices? Not surprising, the researchers find that funds domiciled offshore, with infrequent audits and that report to clients and databases more frequently seem to have more of these pricing discrepancies. Basically, conclude the authors, funds that “are less likely to get caught” and “that inform their investors more frequently” tend to mismark their positions more. And although this mis-marking may represent a very small proportion of the typical portfolio, it may be just enough, as the authors suggest, to impact returns (either by smoothing them or by nudging them above zero, for example).
The irony here is that greater regulated transparency wouldn’t seem to remedy this situation. After all, hedge funds already need to report large (long) positions in their 13F filings and the numbers still don’t seem to match publicly-reported prices. Recent changes to accounting standards (FAS 157, in particular) may help but as the authors note, are only now coming into effect. So it will be interesting to see if this, coupled with Madoff, greater due diligence, and the “institutionalization” of hedge funds will make a difference.