mulberry purses Burying the Myth of Survivorship Bias
If not the coffin, we have at least a few nails toward dispelling the conventional wisdom about survivorship bias in hedge fund indexes. Our empirical research will demonstrate that hedge funds more often stop reporting returns following periods of positive performance, not when they are likely dying off.
What is survivorship bias? Survivorship bias is the distortion of an index that results from not including certain members in the average. If the funds with poor results dissolve, the remaining positive funds cause the average to overstate the industry average. This issue has been assessed and quantified numerous times across many stock and mutual fund indexes. Since some of these indexes do not include members that have failed, the index is distorted by the exclusion of negative events.
As mutual funds are registered securities, active funds must report their results. Funds with good performance continue to existnd as registered securities must report so the average always includes positive members. Therefore, the exclusion of deceased funds skews the average by excluding those members that had poor results. This creates a bias toward the results of surviving funds. The resulting overstatement of performance is known as survivorship bias.
There are several critical elements contributing to survivorship bias in stock
market indexes. Foremost, the loss of a member is due to below average performance (thus distorting the average upward); ongoing positive performers can and don stop reporting their results. Second, reporting is mandatory, thus the index represents the universe of stock or mutual funds alternatives. We have no dispute that survivorship bias is alive and well in stock and mutual fund indexes. However, this issue has also been expropriated from the public realm and misapplied to the private world of hedge funds and their indexes.
oth studies indicate that survivorship bias in hedge fund indexesmay actually cause an understatement of returns available from hedge fund investing./p>
These distorting elements do not exist in the world of hedge funds. There may be other issues clouding the picture, but not survivorship bias. On the other side of the looking glass, hedge funds cease to report results for two reasons: strong performance leading to a losingof the fund to new investors and poor performance losuresleading to new careers for the partners. As well, unlike their publicly registered counterparts,
hedge funds report voluntarily to one or more industry databases that compile indexes from those funds that choose to be included. As a result, hedge fund indexes represent only a general proxy for the industry rather than a comprehensive measure of it.
Most hedge fund databases generally exist to provide funds and investors access to each other. Hedge funds that are closed to new investors have less incentive to provide return information. As a result, hedge fund indexes are not comprehensive. In addition, funds that do report may later choose to stop reporting not only due to poor performance, but also due to strong performance. This creates a distribution of events that could either overstate or understate the industry averages and leads to a key question: arethere more poor performers leaving than strong performers bowing out?
Why do we even raise this question when, of course, everyone knows that hedge funds are risky investments and are constantly lowing up only to bury their investorscapital balances. Not so fast.
In two separate studies, we have attempted to quantify the impact of survivorship bias among hedge fund indexes. Both studies produced similar results. In the first, we assessed a series of sixty funds that had ceased to report results to a recognized hedge fund database. The returns from each fund were reviewed for their performance during the prior three months and twelve months. Funds with negative performance were marked as possible losures,whereas those with positive results were considered losingsto new investors. We found that almost two thirds of funds that ceased to report had positive results preceding the end of their reporting.
The second and more comprehensive study included 462 hedge funds that had stopped reporting over a number of years. The funds were assessed for their performance over the three, six, and twelve month periods prior to their last report. For the entire group, the results were substantially similar to the first study: only about one third of funds had poor returns through the point of voluntarily ceasing to report.
Both studies indicate that survivorship bias in hedge funds is skewed heavily toward losings(to new investors) rather than losures(going out of business). Given the lack of reporting requirements and the potential for both positive and negative drivers within the hedge fund industry, survivorship bias may actually cause an understatement of returns available from hedge fund investing.
Neither of our studies purports to qualify for the financial journals. Wee seeking to provide balance to what has been a onesided case. The studies represent empirical and anecdotal evidence that compellingly dispels the tenet of survivorship bias. Though the results of our studies are not conclusive, they clearly indicate that further work is needed to quantify the survivorship bias in hedge fund indexesnd the amount by which their returns actually may be understated.