The Physics of Wall Street: a brief History of Predicting the Unpredictable

The Prediction Company 
•
153
Griffin Street office in July 1991. the time for being choosy was com-
ing to an end. the partners knew they didn’t want to sell the company 
so soon into the adventure, but the idea of being someone else’s hedge 
fund was starting to look appealing. At least they’d have capital, and 
they would be (more or less) independent. they had spent months 
interviewing possible partners, and at this point it was hard to imagine 
a better solution.
And then, in early March 1992, a miracle happened. farmer had 
been invited to give a presentation at an annual computer conference. 
He had reluctantly agreed to attend, on the basis that Silicon valley 
investors would be there and they might be willing to offer some no-
strings-attached financing. He gave a talk on the role of computers in 
prediction, which generated a lot of questions. Afterward, as he was 
packing up his slides, a man in a suit approached him. He introduced 
himself as craig Heimark, a partner at o’connor and Associates —
the firm that had made its first fortune by successfully modifying the 
Black-Scholes equation to account for fat-tailed distributions, under 
the guidance of Michael Greenbaum and clay Struve. By 1991, it was 
one of the biggest players in the chicago commodities markets, with a 
focus on high-tech derivatives trading. the company had six hundred 
employees and billions of dollars under management. o’connor wasn’t 
using nonlinear forecasting, and the Prediction company wasn’t inter-
ested in derivatives. But nonetheless, o’connor and Associates were 
the Predictors’ kind of people. In fact, one of o’connor’s recent hires 
had been a friend and fellow researcher back in farmer’s and Packard’s 
academic days.
Shortly after farmer and Heimark met, farmer received a phone 
call from another o’connor partner, named david Weinberger. Wein-
berger had been one of the very first quants, leaving a teaching job in 
operations research (essentially, a branch of applied mathematics) at 
Yale to work for Goldman Sachs in 1976, even before Black arrived. 
He’d moved to o’connor in 1983, to help that company come up with 
new strategies as more and more companies got on the Black-Scholes 
bandwagon. He was one of the few people in the industry, even in 1991, 
who both was high powered enough to make a deal and also spoke the 

---

language of the scientists running the Prediction company. He called 
on a friday afternoon, from chicago. on Saturday morning, he was 
sitting in the Griffin Street office.
o’connor turned out to be just the kind of firm that the Predic-
tion company wanted to work with — in large part because the people 
working at o’connor were able to understand what farmer and Pack-
ard were doing well enough to evaluate it themselves. Under the deal 
they ultimately negotiated, the Prediction company maintained its in-
dependence. o’connor put up the investment capital, in exchange for 
the majority of the proceeds; it also fronted the Prediction company 
the funds it so desperately needed in order to pay salaries and buy 
equipment in the meantime.
the deal with o’connor seemed perfect at the time. But it turned out 
to be even better than the Prediction company founders had hoped. 
When o’connor came knocking on the Prediction company’s door, it 
already had a long-running partnership with Swiss Bank corporation 
(SBc), a nearly century-and-a-half-old Swiss bank. And then, in 1992, 
before the ink was dry on o’connor’s deal with the Prediction com-
pany, SBc announced its intention to buy o’connor outright. the 
Prediction company found itself in a partnership negotiated with its 
kindred spirits at o’connor but funded by the much deeper pockets 
of SBc. Weinberger was given a top management position at SBc and 
continued as the principal liaison for the Prediction company. It was 
an ideal arrangement. the Predictors had hit the big time.
In 1998, SBc merged with the still-larger Union Bank of Switzer-
land to form UBS, one of the largest banks in the world. despite the 
size difference, however, most of the senior positions at UBS went to 
former SBc managers and the relationship with the Prediction com-
pany was maintained.
the Prediction company, following the o’connor tradition as 
a secretive high-tech firm, never released any metrics of its success 
publicly — and none of the former principals or board members with 
whom I spoke were authorized to share any concrete information. 
this might seem suspicious. After all, if you’re successful, why hide it? 
Here, though, the opposite is the case: on Wall Street, success breeds 
154 
•
t h e p h y s i c s o f wa l l s t r e e t

---