There are a number of well-known biases in the reported returns of hedge funds. Among them are survivorship bias (estimated to be about 3 to 4 percent) and backfill, or “instant history,” bias (estimated at more than 5 percent per year).
There’s also the issue of self-selection, or self-reporting, bias. Poorly performing funds may choose not to report, although some funds that have performed well may also choose not to report because they may be closed to new investors and thus don’t need to attract new assets.
Self-reporting also leads to situations in which it’s possible that different models are used to value assets, as well as the potential for earnings smoothing, understating the true volatility of returns and making assets appear to be less correlated with other assets than they truly are. Indeed, researchers have found that smoother returns are associated with managers who possess greater discretion in sourcing the prices used to value the fund’s investment positions.
There’s yet another large and important bias in the data—liquidation bias. This phenomenon occurs because funds that become defunct frequently fail to report their last returns. One study estimated that this effect resulted in a bias that overestimates returns by 3 percent to as much as 6 percent.
Data Revisions Study
If all these problems weren’t already enough, Andrew Patton, Tarun Ramadorai and Michael Streatfield—authors of the study “Change You Can Believe In? Hedge Fund Data Revisions,” which appears in the June 2015 issue of the Journal of Finance—supply investors with another concern related to the reported historical performance of hedge funds.
Their study covered the period July 2007 through May 2011 and more than 12,000 hedge funds. Following is a summary of the authors’ findings: