Earlier this week, we told you about a research paper that suggested some hedge fund managers tend to mis-price their publicly-traded long positions. In that paper, researchers compared managers’ 13F filings to publicly-available stock price data to see if they matched up (they sometimes did not). As the authors of that paper noted, previous attempts to catch hedge fund managers “in the act” of misreporting were based on a telltale discontinuity in the return distribution of hedge funds as a whole: the odd dearth of very small negative returns (as in this earlier study we covered by Nicolas Bollen of Vanderbilt and Veronika Pool of Indiana University).
Another new paper posted in the past month uses the Bollen & Pool method to examine what kinds of funds tend to exhibit this suspicious return pattern. While Bollen & Pool asked, “Do hedge fund managers misreport returns?” Petri Jylha of Aalto University in Helsinki asks the question “When do hedge fund managers misreport returns?” (We’ll issue the same caveat here that we did earlier in the week: The possible existence of return-management at some funds does not suggest that all “hedge fund managers misreport returns.”)
Like Bollen & Pool, Jylka finds a statistically improbable dearth of slightly negative returns and an almost equally-improbably surfeit of slight positive returns – suggesting that some managers may have influence over the final results.
But then he wonders what kind of funds might be involved with “nudging” small negative returns into positive territory, and in what kinds of circumstances might they do this. As intuition would suggest, he found that funds facing an exodus of investors are more likely to panic when they post a negative return. He divides his universe of over 400,00 fund-months into two categories: returns experienced by a manager who has a net increase in AUM that month (top 2 charts) and returns experienced by a manager who has a net decrease in AUM that month (bottom 2 charts).
Further, he found that younger funds (hungry for a track record) and funds displaying a historically strong relationship between performance and asset flows tended to have a higher probability of similar statistical anomalies.
But then he wondered how all of these variables interact with each other. For example, do young funds facing an investor-revolt undertake suspected return-smoothing more than older funds facing the same situation?
Rather than listing out reams of probability densities and z-statistics, we summarized the results below into a series of 2×2 matrices. A shaded circle means Jylha founds a massive discontinuity – that this category of managers has a relatively large number of returns-managers. A blank circle denotes the absence of statistical anomalies and the likely dearth of manager involvement in setting the final return figure. The pies were created by simply eyeballing the z-statistics in the paper.
When he compared the response of old and new funds to negative asset flows, he found little difference. Both categories of returns showed anomalous results. But when he compared old and new funds’ response to positive flows, he found that new funds still seemed to fiddle with returns while older funds did not.
You can see why a manager might pull out all the stops when times are tough. But why fiddle with returns when you’re looking at likely inflows? Jylha suggests it’s because smaller funds are keeping an eye on subscription and redemption NAVs…
“…fund managers protect the fund’s shareholders from dilution when the fund receives a large redemption or subscription. Such large flows may result in trading costs that are borne by all investors in the fund, not just the redeeming or subscribing investors. To protect the existing shareholders from this dilution, hedge fund managers refrain from misreporting during large out flows and overstate returns during large inflows.”
Next, Jylha looked at how the historical “flow-performance” relationship impacts funds that face asset inflows and outflows. If a hedge fund has a bunch of jumpy investors (“hot money” as the say in the industry) then you might expect the manager to try and nudge a slightly negative return into positive territory. It turns out that funds with both a strong and a weak flow-performance relationship showed a similar propensity to fiddle with slightly negative returns.
Finally, Jylha looked at the link between the direct of asset flows (into the fund or out of it) and the relative size of those flows – large (over 5% of funds assets) or small (under 5% of assets)…
Curiously, he found that funds facing a small outflow of assets tended to display a larger discontinuity of returns than funds who were about to watch over 5% of their assets walk out the door. This is an example, according to Jylha, of where a manager’s incentive to avoid a negative sign in front of their return is superseded by their desire to report a negative return as a going-away gift to redeeming investors. Far from altruistic, however, Jylha points out that hedge fund managers are often significant investors in their own funds.
This study shows that the statistical anomaly identified by Bollen and Pool isn’t as ubiquitous as once thought. In fact, funds facing a positive flow of assets tend not to obsess about a few small negative returns and appear less likely to goose small losses into small gains. But when assets are flowing out, things appear to be very different.
This makes us wonder about 2008 and 2009. As we all know, a vast swath of hedge funds faced net outflows over 2008 and part of 2009. Does that mean the small monthly loss is becoming an endangered species? Is the hedge fund industry now awash with slightly positive returns now?