Friday, August 10, 2007

In the Black Box

Unless you were in a coma this last couple of weeks, you have not missed the roller coaster stock market in early August 2007. The seemingly bottomless pit dug by the sub-prime mortgage lenders has the hedge funds - who supplied lenders with capital in the first place by buying their paper - running for cover. Apparently, there have been margin calls and liquidations aplenty.

The stock market has whipsawed up and down daily for the last two weeks, and some point to so-called “black box” investing as another cause of the volatility. The finger of blame is pointed in particular at the statistical arbitrage funds.

To me it is a bit counter intuitive that quantitative investing strategies lead to increased volatility. The “stat arb” folks, many of whom have PhDs in mathematics and computer science, use time series statistical methods to identify mispricings in stock prices. Didn’t we learn in economics 101 that arbitrage activity eventually brings prices into equilibrium and shouldn’t that mean less volatility?

Arbitrage strategies have become increasingly popular, especially at hedge funds because such strategies are by definition riskless and self-financing. These funds employ software developers to create sophisticated programs for finding and then triggering “riskless” positions in securities, all of which are aimed at finding the money left under the sofa cushions by fundamental and even other technical investors.

Quantitative hedge funds
Sowood Capital and Renaissance Technologies both reportedly have moderate to serious problems associated with exposure to defaulting mortgages as well as shrinking liquidity. It will likely take months to find out what has been happening inside those black boxes.

As a fundamental analyst, I find myself a bit uncertain on the impact of sophisticated arbitrage strategies on market volatility. Arbitrage strategies have been around since market makers were gathering under the tree down on Wall Street. Benjamin Graham discussed price discrepancies created by special situations such as mergers and buyouts in his seminal financial text, Security Analysis. Graham did not envision that computer technologies could take simple arbitrage principles into new uncharted territory such as index arbitrage and volatility arbitrage.

The Complete Arbitrage Deskbook is comprehensive by listing nearly all arbitrage strategy categories, making it a good primer for the beginner. However, the author, Stephane Reverre, skimps on the math that is the lifeblood of statistical arbitrage. Understanding Arbitrage: An Intuitive Approach to Financial Analysis by Randall Billingsley is not a how-to book either, but it does illuminate the principals of arbitrage that run through all financial markets. With its context in the modern computer world, it makes a good follow-up reading after Graham.

For those who want to get to the heart of “mean reverting pairs” statistical arbitrage strategies, Pairs Trading: Quantitative Methods and Analysis by Ganaphathy Vidyamurthy is a must read. It is also a tough read, but of all the books mentioned here, it is most likely to provide insight into how statistical arbitrage could contribute to market volatility.

No comments: