The grim figures for August losses among some of the world’s biggest hedge funds prompts the FT to examine what went wrong – and right – among the computer-driven hedge funds, or “quants”. The results show that some funds recovered fully, while others are stuck with most of the losses from the appalling first 10 days of the month.
James Claus, who founded the fund after leaving London-based GSA Capital, reported to investors that his GMN Fund had $600m of assets in August, down from $1bn at the start of this year, after losing 19.1 per cent in the past month.
“There’s no doubt that the sector as a whole will suffer just as the convertibles sector suffered a few years ago,” one quantitative manager told the FT. “Investors will get out at just the wrong moment.”
The big unanswered question, says the FT, is whether investors will reject all quantitative equity funds, which use “black box” computer models to select stocks, even though many managed to claw back all their losses. In the flight from convertibles, the managers who stemmed losses during the rout of 2004-05 were hit with withdrawals almost as hard as poor performers, as investors in hedge funds shifted money to other strategies.
Paul Trickett, European head of investment consulting at Watson Wyatt, told the FT the boom in computer-driven hedge funds raised the risk of many using the same basic models, and so selecting the same stocks.
But some of the computer models bounced back strongly in August, including big names such as Jim Simons, the former mathematics professor who runs Renaissance Technologies; 32 Capital from Barclays Global Investors, part of the London bank; DE Shaw, part-owned by Lehman Brothers; and Highbridge Capital, owned by JPMorgan. Both Renaissance and DE Shaw are now raising new money.
A handful of other quant funds, including Sushil Wadhwani’s London-based Wadhwani Asset Management, which uses both macro and equity strategies, avoided the crisis altogether. Wadhwani ended the month down just 0.5 per cent, hurt by the move in the yen.
