In today’s Radio Derb, I pass some remarks about “quants”–i.e. quantitative analysts, the guys who do the heavy number-crunching in Wall Street’s back offices. These guys generally (though by no means always–it’s pure meritocracy on the quant floor) have a Ph.D. or two in math and allied subjects. They are very smart:
IQs up around three sigmas from the mean for their population group.
(Above three sigmas, you’re in the top one or two per thousand.)
Well, here is something on the recent market ructions from an actual quant–an exceptionally knowledgeable & accomplished one, who’s been doing this stuff for 20 years. I’ve doctored it slightly to remove identifying traces, and to improve the English. (These guys tend to think too fast to be bothered with formal grammar, and are anyway often recent immigrants with a first language other than English.)
“On the off chance that you or the folks at NRO are interested, I thought I’d relate a few of the goings on from the quantitative space of the financial markets. …
“The program trading systems known as ‘statistical arbitrage’ have been running for years now and have become very popular. Some of them have made billions of dollars by exploiting small statistical abnormalities in the relationships between various stocks. But the problem is that the vast majority of these programs work based on an assumption of eventual convergence, so as more people do the same thing, the spreads between instruments begin to close and more leverage is required to get the same level of absolute return on the investment. These days, many of these strategies run from 3 to 10 times leverage (I’ve even heard of one which runs at 12 times leverage) meaning for every dollar they have invested, they have borrowed an additional 3 to 10 dollars and invested that as well.
That causes you to accumulate something called ‘liquidity risk’, meaning it will take you something like 3 to 10 times as long to sell the position in an emergency.
“When the credit market had its crisis, it caused just such an emergency. Several large players needed capital to meet their margin, so they sold in the most liquid market they have which was the US equity stat-arb space. The problem is, there are only so many assets available in the US market, and only so many statistical abnormalities, so since many of these guys all had the same mathematical training, and were using the same data, they were all finding the same alpha. It’s like 10 guys each digging a gold mine only to discover a mile below the earth that they were all mining the same vein, and now the ground above them is too weak to support the roof. The selling by the big players caused stop-loss limits to be triggered at other firms, so there was more selling which caused more triggers to be hit and so on, and so on.
“Ironically, it’s the best companies (those with the best forward prospects) that are most aggressively being sold off, and the less healthy companies which are being bought up, so this liquidity crisis is actually causing a fairly large value inefficiency in the market, and if someone can figure out when to try to catch the falling knife, they are going to make a lot of money. The economy is still in pretty good shape in spite of this market issue, but their assets are being systematically mis-priced in the market anyway. That can only happen temporarily since the market is so naturally self-correcting, and on that correction someone is going to make a killing.
“As for my personal quantitative strategy, since it identifies value from a different basic operating paradigm than statistical arbitrage, I’m having a small loss (less than 2%) and a rise in volatility, but nothing outside the expectations of the model. In point of fact, my year is still going quite well, but that isn’t the case with most of the industry. I was on a conference call yesterday with [name of a renowned quant], and of the [three-digit number] people on the call, I think I was the only one who was still at a healthy profit on the year.”