GFIA has released a study on the usefulness of some of the more commonly used risk metrics in the hedge fund industry. The study was conducted on a universe of 316 funds with at least 2 years of track record, sourced from the Asiahedge database.

Summary findings include:
- Statistical risk measures did not persist during the financial crisis, suggesting that historical statistics are not a precise decision tool for allocators.
- However almost half funds in the universe demonstrated high persistency in the relative “riskiness” of a fund with its peer group.

Peter Douglas CAIA, principal of GFIA, commented: “We have always focussed our practice on qualitative assessments of funds and their managers, and have always been nervous of investors’ reliance on statistical measures as a decision tool. Our research confirms that it’s a mistake to consider statistical risk measures such as standard deviation (volatility), downside deviation, etc., to be accurate indicators of future risks. These measures, however, are quite likely to remain relatively constant between peer funds, especially for funds at the extreme ends of the risk spectrum.”