Tuesday, August 14, 2007

"Ticking" Sound in a Black Box

In August 10th post “In the Black Box” I recommended some “light” reading on quantitative investing strategies, in particular the so-called “stat arb” trading that has become so popular with the hedge funds.

The fingers have been pointed at the programmed trading schemes that exacerbated flights to quality and liquidations by holders of securitized sub-prime mortgages. However, I have not been as quick as some to blame the hedge fund mathematicians and software programmers for the volatility we have seen in the U.S. equity markets.

It is clear that investor reactions over the last month have not been uniform and that some stocks have moved opposite to the logical direction based on economic fact. It is also clear that some stocks gyrated up and down in wide moves that were not related to company-specific, sector fundamentals or even the credit issues that dominate the financial news.

Ok, so maybe the PhDs and programmers were not prepared for such a meltdown in sentiment. After all, quantitative strategies rely heavily on historic trading patterns. Perhaps they were less prepared than they thought for the disruptive effects of unprecedented exogenous factors.

First, sub-prime mortgages have not been used to the extent that we have seen in the last five years or so. The securitization activity was at record levels. Hedge funds bought up those instruments at record levels…and perhaps used record leverage. The week of July 16-20, when Bear Stearns announced that two of its hedge funds were in trouble, might have been the first that some outside….or even inside…the hedge funds’ computer rooms gave any thought to how well the effects of exogenous events are incorporated into their programmed trading strategies.

Second, there is another overlooked factor in the mix that may have had an effect on how programmed trading strategies were executed. Here is where I am prepared to concede that quantitative strategies could create stock price volatility.

On July 3, 2007, the SEC’s plan to eliminate the Uptick Rule became effective. The long-standing Uptick Rule prevented short-sale trades on a downtick. Since short-sales are key positions for quantitative trading strategies, we suspect the permanent elimination of the rule impacted the execution of programmed trades in the fevered days following Bear’s “bearish” news.

Unfortunately, I have no data to prove anything one way or the other. Nonetheless, it would make an interesting study to determine whether the combination of unprecedented fundamental developments, an unexpected event, and a changed trading rule-set combined to render programmed trading schemes invalid.

Listen carefully...is that a ticking sound in the Black Box?

No comments: