The mystery of “slightly positive results” in hedge funds and college basketball

There has been some talk recently of hedge fund managers misreporting returns to the various voluntary databases that collect such information.  According to a study by Nicolas Bollen of Vanderbilt and Veronika Pool of Indiana University, hedge fund managers show a curious propensity to have more slightly positive months than they do to have flat or slightly negative months.  In fact, the study found that up to 10% of fund returns in one major database are statistically “distorted”. 

You don’t have to be a statistician to see it (chart below).

 

According to the study, hedge fund managers also seem to have smoother returns when they are performing poorly.  Researchers conjecture that this is because managers have an ability to “manage earnings” and are more likely to do so when their returns stink.  That way, at least they can reduce volatility and goose their Sharpe ratio even if returns are lackluster.

 

It appears that such shenanigans are more likely in strategies where the manager has a larger degree of influence over valuations (e.g. distressed) than in strategies where valuations are marked to market (e.g. market neutral equity).  This phenomenon was also found to be prevalent among managers with more than one fund – suggesting to the authors that some managers could take advantage of crossing trades between accounts to give one fund a small, but much needed boost into positive territory. 

This mysterious lack of “slightly negative” returns is particularly pronounced among “defunct funds”.  It appears as if funds that eventually failed to make it where nonetheless uncommonly “lucky” when it came to squeaking out slightly positive results (chart below).

The authors site studies in other fields that point to the effect of a binary “hurdle” such as the coveted “up-month” for a hedge fund manager.  This article from Wharton, for example, analyses “point-shaving” in NCAA Basketball games (coming in just under the number needed to meet the spread).  Like the study by Bollen and Pool, it concluded there may be some funny business going on there too.  Says the article: “… around 1 percent of all games involve gambling related corruption.” 

This begs the question, “should there be tighter regulation on hedge fund reporting?”.  While they avoid any specific analysis of regulation, the authors suggest that there may be a greater need for SEC oversight than currently meets the eye:

“Regulators who argue that the low number of fraud cases prosecuted by the SEC means that additional oversight is unwarranted may find the robust discontinuity we document indicative of a more widespread violations.” 

When it comes to hedge funds, the bottom line seems to be clear: don’t rely on the “percent positive months” metric when buying a hedge fund (it’s a pretty lame metric anyway).

TOMORROW: Another study explores some of the possible systemic reasons behind this apparent misreporting of hedge fund returns.   

Be Sociable, Share!

4 Comments

  1. miami
    November 21, 2007 at 2:36 pm

    The Penn article was laughably wrong. Vegas does not set unbiased lines. Vegas does not try to attract ‘equal action,’ known as the Myth of Equal Action among gamblers.

    Certain teams, like Duke, UCLA, Yankees, ND, etc and large favorites attract more money.

    The book sets their lines ‘higher’ to force favorite bettors to pay a ‘tax.’

    Since the line is purposefully higher than a line set merely to attract equal action, favorites don’t cover as much. This is all well-known and established in gambling markets. Ask any LV bookmaker, they freely admit it because Duke fans don’t stop betting Duke because the line is -22 instead of -20.5.

    In addition, for really big spreads, it is likely the coach has put the scrubs in. 1st team of the other team can usually narrow the lead on the scrubs.


Leave A Reply

← Do individual hedge funds really contain so much beta? The Gross Truth About Hedge Fund Performance and Risk: The Impact of Incentive Fees →