Friday, December 14, 2007

Stat Arb + Economic Stress = Trouble?




As part of a continuing series on surviving as a quant , I would like to focus on how investors need to know the economic rationale behind a quant strategy.

The statistical arbitrage hedge fund strategy, or “stat arb”, is a case in point. Classic stat arb can be simplified as buying oversold stocks and shorting overbought stocks, along with some risk control layered on top of the stock selection process.

The economic rationale behind this type of strategy is that the stat arb practitioner is being paid to provide liquidity to the market. In normal times, this approach can be quite profitable but it can backfire badly during periods of economic stress. If you use a short-term investment strategy of buying oversold stocks and shorting overbought stocks during a recession, you will ride the big losers (e.g. Adelphia, Enron, etc.) all the down to the bottom.

The accompanying chart shows the investment results of an overbought/oversold model. It ranks US large cap stocks on a short-term overbought/oversold measure and buys the bottom 20% most oversold and shorts the top 20% most overbought stocks. I do not pretend for the moment that this is an actual stat arb strategy as it has no risk control. However, it does serve as a proxy for the performance for these types of strategies as I have discussed elsewhere. This model had a drawdown of over 20% in 2001, as many stat arb strategies did at the time, and has been having some difficulty currently.

The signs of economic stress in are everywhere, particularly in the US. Investors should be wary of too much exposure to stat arb strategies under these economic conditions.

No comments: