From www.hedgeworld.com - login required [...]No market has been isolated and Mr. Rothman said that the credit crisis has had an effect on the models because it dried up liquidity. The starting point, he said, was most likely that some large quant managers experienced losses in their credit or fixed-income portfolios. In an attempt to de-lever their illiquid credit funds, or to avoid having to mark those portfolios to market, those managers sought to raise cash, he said. They may have sold long positions that they would have held longer in normal times. In addition, these quant managers may have covered their short bets by buying those securities in an attempt to exit those positions. "Short names started to rally and long names started to fall as these trades started to hit the market," Mr. Rothman wrote. As most quant managers use similar models, those trades were not confined to a few funds. As a result, he wrote, models are working in a "perverse manner. . . . The names that are short are outperforming often notably while the names that are long are underperforming, although less severely." Speaking on a conference call, Mr. Rothman said that the performance of quantitative strategies improved on Friday, and reversed around 70% of the losses from the first four days of last week, Reuters reported. But he said that the market will continue to be volatile and that models need to be adjusted as a result. [...] Based on a study of $204 billion worth of long stock positions owned by quantitative-driven hedge funds, Mr. Kostin ( Goldman Sachs ) determined that lack of liquidity was an important factor behind underperformance of these funds. In fact, illiquidity of a given stock was more of a determinant of poor performance than either volatility or equity capitalization, he wrote. He measured liquidity as an aggregate quant fund's ownership of a stock divided by its average daily volume. The difficulty in selling shares in the current market drove the bad performance during the July 19-Aug. 8 period studied by Mr. Kostin in his research. Within quant funds, stock positions with the least liquidity underperformed positions with the most liquidity by a significant amount, he noted. In terms of performance numbers, Mr. Kostin found that based on his stock sample, the least liquid stocks posted a 9.2% loss from July 19 to Aug. 8, while the losses for the most liquid positions were limited to 3.6%.[...] In a research noted titled "A Challenging Environment for Quant Strategies," Keith Miller, head of global quantitative research at Citigroup, pointed to correlation as a factor driving bad performance. Quant models are built based on several factors or variables. Lately, correlation among those factors, such as valuation, price reversal, earnings momentum and profitability, has been on the rise, Mr. Miller wrote, which is problematic because higher correlation leads to more risk.