Monday, Apr. 11, 1994

Attack of the Data Miners

By John Skow

BACK IN THE PALEOCAPITALISTIC ERA, as long ago as 10 years, anthropologists studying the breeding, feeding and plumage patterns of Wall Street concentrated on carnivores -- called gunslingers -- grownup frat boys in yellow ties and red suspenders who peddled junk bonds, drove BMWs and bought $2 million co-ops on Manhattan's East Side. Forget them: today they're fuddled old greedsters sitting around in their East Hampton beach houses wondering what happened.

What did? Well, yellow ties are out, but avarice remains popular, and the financial universe -- Is anyone surprised? -- still has masters. These new barbarians at the gates of international commerce may have the geeky, high- water-pants look of the typical math grad student, and they may caress their Sun Microsystems workstations rather than I-got-mine mobiles. But nearly everyone agrees that they are even scarier than the gunslingers. They are "math jockeys," "nerds," "pop eyes," "quarks," "techies." Call them quants, for quantitative analysts. They are odd birds indeed, the field biologist discovers, and . . . Hark, here's one now!

". . . European portfolio MITTS in a market-index, targeted-term security. It's an equity-link note. This one had 90% principal protection, plus equity upside in its European portfolio." The specimen is Jamie Greenwald, 30, managing director of global equity derivatives for Merrill Lynch. He reports, with satisfaction, that the Japan index "provides upside in the market in Japan in a domestic instrument, U.S. dollar-based, no currency risk, no downside risk: worst case you've got about a . . . ((pause)) . . . 2.34% yield. That was very applicable to pension funds, to insurance companies, to mutual funds."

There is a perilous impulse here to say, "Sure, if you say so." And not just outsiders are baffled by such impressively technical discourse. Older managers weren't taught this stuff. "The guy who heads your group, a head trader, he's never solved a quantitative problem before," says a 30-year-old Ph.D. at a leading brokerage house. "An important problem that could take you three weeks to solve properly, he'll want it done in two days. It's very difficult for a quant who thinks of himself as a Ph.D. from a top-notch school and comes to Wall Street, and a high school dropout screams at him and calls him an idiot." His colleague, an engineer, agrees: "Many times, what your boss is saying is just hilarious. It's wrong, you know; mathematically it makes no sense. You can't even say, 'Look, you don't know what you're talking about.' "

But Wall Street and the quants are stuck with each other. Stanley Diller, 58, an early quant who is managing director of fixed-income research at Paine Webber, left a job as an economics professor at Columbia in the mid-'70s to join Goldman, Sachs & Co.'s equity-research department. In those days, he says, "research was largely an image builder. It was something that brought in the customers." Now quantitative research "is the whole deal." If you don't have it, says Diller, you can't produce the new financial instruments, " 'cause you get crushed trying to hedge them." Meaning that Wall Street's new products are so complicated and interdependent that only the advanced number crunching of the quants can untangle the risks involved; without it, the market crushes you.

The result is that techies in large numbers -- engineers who lost jobs at the superconducting supercollider, doctoral students bored with their computer-science dissertations -- are heading for Wall Street. Says Diller: "The people who study science but are not themselves weirdos -- a small subset, people who can adapt to the real world -- become aware that they can make a lot more money."

Wall Street as the real world is a concept that could raise eyebrows. But some financial experts wonder whether the quants are weakening whatever contact with reality the street may have had. Steve Barnett, an anthropologist who is a principal of Global Business Network, a think tank, says that in the pre-quant days, Wall Street was patriarchal, intuitive, much more related to the world as it was. But hard-core quants, he complains across the generation gap, "are almost idiots savants with numbers . . . There is an almost prayerful communion with the computer. They're intense and operate to a rhythm. If you ask them a question, they turn and their eyes are glazed, coming out of whatever cyberspace they are in." In this trance, he says, "they're not really in a world of other people. They think they're in a world of pure technical manipulation, like a chemist creating a molecule. It's as though there are no social consequences."

Quants, says Barnett, who has a Ph.D. in anthropology from the University of Chicago, tend to be bachelors (few are women) who live in apartments as messy as the room they left in grade school. Many of them drink hard after hours, mostly with fellow workers. They mate, if that's the word, mostly in one-night stands. When they air their lives out, it's with ski holidays and ecotourism, not yachting or casino crawling. With salaries for researchers that start at about $90,000 and can climb well over $500,000 for those who excel, they could afford to dress with the flash of yesterday's gunslingers. Most don't. An atypical Merrill Lynch computer jock keeps a 360-hp speedboat in Westport, Connecticut. This appears to embarrass him, and he blusters, "That's not who I am, and if you don't tell me right now that you're not going to put it in the article, I'm going to have to get serious and call our public relations people and have them call TIME."

It takes three to six months, says one quant who has made the transition, to change a shy, bookish type into a ruthless money-making machine. What's required, says this alumnus of the system, is "to lose your sense of decency. You have to be rude, brash, you have to be selfish. Also you have to start ignoring 90% of what you are told." He describes, perhaps admiringly, a vulnerable Ph.D. from Princeton University. This fellow wore $50 suits and thick glasses. He was painfully polite. Transformed, he became the quant from Hell. "He's got this personality suddenly. He could eat these guys alive," says the quant. For someone like this, academia loses reality, and from Wall Street's viewpoint, a professor with a scholarly paper is "like a two-year- old coming with something he drew."

Anthropologist Barnett, reflecting on the brokerage business, says, "You can't do it intuitively anymore." He adds, "The next generation of computer architecture, be it massively parallel programming or 64-bit addressing or hyper- or meta-computing, essentially is going to be data mining where the data will be searched in even finer granularity to discover patterns that even this generation can't get at."

His somewhat glum conclusion: "So wait till the Generation Y quant people hit Wall Street." When this occurs, at least one Generation X precept is unlikely to be disproved. As one quant said last week, "If you make money, ) nobody calls you a geek."

With reporting by Massimo Calabresi and Sribala Subramanian/New York