Burton Malkiel of Princeton University and Atanu Saha of the Analysis Group, in their paper entitled ‘ Hedge Fund: Risk and Return ‘, are asserting that the returns of hedge funds may be lower than commonly advertised, possibly making them a riskier choice for investors.
The study analyzed the TASS database, which is run by hedge fund group Tremont Capital Management. The two academics claimed that because of backfill and survivorship biases within hedge fund indices, investors should look at such an index with a skeptical eye.
Backfill bias, the authors assert, occurs because hedge funds report performance on a voluntary basis. Managers quite often will retroactively report returns if they are positive, while often failing to do so if returns are negative. Looking at incidences of such practices, Malkiel and Saha found that backfilled returns were more than 5% higher annually than those that were not backfilled.
Survivorship bias also warps the actual performance of hedge fund investments, according to the authors. While most hedge fund databases reflect the returns of hedge funds currently in existence, most do not include hedge funds that have failed (the study also asserts that only a quarter of funds in existence in 1996 are still around today). It is common, the authors assert, for funds that are failing to stop reporting their returns for the months before they go under This makes it so only funds who are currently enjoying a successful run are included in an index, biasing the results so as to not include the entire hedge fund universe.
The study cites the ubiquitous Long-Term Capital Management, a massive Greenwich-based hedge fund that infamously tanked in 1998, as a prime example of such action. Although the fund lost 92% from October 1997 to October 1998, it failed to report any of the losses to hedge fund databases, skewing the returns of the industry as a whole. Although it was still losing money, hedge fund databases would not pick up on a fund such as Long-Term; so any hedge fund index that purported to speak for the entire universe would be incorrect in its performance reporting.
Critics contend that every index has this problem, according to USA Today. They also point out that not all funds with positive returns report their performance either, since they may not be seeking further capital or may just strictly abide to the industry standard of intense secrecy.
Overall, the study contends that funds that survived outperformed a combination of live and dead funds by nearly 4% over a period stretching from 1996 to 2003. Earlier studies pinned this number at anywhere between 0.6% and 3.6%, according to USA Today.