"Quant funds"

Strategic artificial intelligence can do more than trounce Russian chess players; it can also beat Wall Street traders at their own game.  Recent reports of multi-billion dollar losses at several large so-called 'quantitative' hedge funds are making investors question the value of relying on machine over man. But making a robot the patsy for 'quant fund' woes is specious. 

Computer based models

Quant funds use pattern recognition software to weigh the attractiveness of certain securities and then select them based on quantitative analysis. Computer models can be advantageous because they are emotionless, and react to market trends much faster than a human ever could. In pure "quant shops", the final decision to buy or sell stocks is made by the model. Other funds use human judgment in addition to a quantitative model.

One could argue that a computer model is only as good as its inventor. But the quant funds generally adhere to a sound strategy of statistical arbitrage -- or StatArb -- that identifies and profits from the relative mis-pricing of related stocks. A hedge fund identifies stock pairs where a historical relationship in relative stock price moves has temporarily broken down. One stock in the pair is bought long, the other is sold short. This hedges risk from whole-market movements and puts on a trade that will turn a profit when the relationship re-establishes. Quant funds use systems that automatically put on hundreds or thousands of these kinds of trades on a daily basis, each one hopefully generating a tiny profit that together generates a significant return.

Rats and sinking ships

As concerns over the US sub-prime crisis continue to buffet markets, hedge fund managers have disclosed significant losses in their quantitative funds. Nevertheless, the sharp declines experienced by many quant funds seem a result of the credit crunch and human frailty rather than flubs written into a software programme:

• In late July, funds facing huge losses in credit portfolios needed to raise cash. 
• Selling of the troubled StatArb positions was undesirable because it would have forced funds to realise a distressed market value for these investments instead of a more forgiving model-derived valuation.
• Instead, the funds decided to cash in the most liquid part of their investments -- US equities.

All this had a big impact on equity pricing, prompting long positions held by quant funds to drop significantly in value, and short positions to increase in value. Faced with loss-making positions in their portfolios, quant funds rushed for the exits.

Fears that these troubles will escalate into a systemic problem are set to recede. Nevertheless, the episode illustrates how the interconnected nature of modern markets and the proliferation of hedge funds can result in erratic dynamics.