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

particles bumping around in the sky. And yet, when one steps back, 
these mindless particles somehow organize themselves into hurri-
canes. Similar phenomena occur in biology. Individual ants seem to 
behave in pretty simple ways, foraging for food, following pheromone 
trails, building nests. And yet, when one takes these simple actions 

---

and interactions in aggregate, they form a colony, something that ap-
pears to be more than the sum of its parts. As a whole, an ant colony 
even appears to be able to adapt to changes in its environment, or the 
deaths of individual ants. once these ideas were in the air at Santa 
fe, it was a natural leap to ask if the economies of nations and the be-
havior of markets could also be understood as the collective action of 
individual people.
the Santa fe Institute hosted its first conference on economics, en-
titled “International finance as a complex System,” in 1986. farmer, 
who at this point was the head of the complex Systems research group 
at Los Alamos, was one of a small handful of scientists who was asked 
to speak. It was his first exposure to economics. the other speakers 
were from various banks and business schools. these bankers stood 
up and explained their models to a group of stunned scientists who 
found the financial models almost childishly simple. the bankers, 
meanwhile, walked away thinking that they had heard the siren call 
of the future, though they had virtually no understanding of what was 
being said. excited, they urged the institute to host a follow-up con-
ference and invite various luminaries from economics departments at 
top universities.
the idea behind the second conference was that even if the finan-
ciers couldn’t follow the latest advances in physics and computer sci-
ence, surely the professional economists would be able to. Unfortu-
nately, things didn’t go as planned. farmer and Packard both spoke, 
as did various other Santa fe Institute researchers. the economists, 
likewise, made their presentations. But there wasn’t much communi-
cation. the two groups were coming from two radically different cul-
tures and taking too many different things for granted. the physicists 
thought the economists were making everything much too simple. 
the economists thought the physicists were talking nonsense. the 
great synthesis of disciplines never occurred.
Undeterred, the institute tried a third time in february 1991. this 
time, though, the economists stayed at home. Instead, the institute in-
vited practitioners from the banks and investment houses that actu-
ally ran the world’s financial markets. the tone of the conference was 
much more practical and focused on how to create models, test them, 
146 
•
t h e p h y s i c s o f wa l l s t r e e t

---

The Prediction Company 
•
147
and use them to develop trading strategies. the traders proved much 
less defensive than the economists, and by the end of the conference 
each group had gained an appreciation of what the other had to offer. 
farmer and Packard, in particular, left with a clearer sense of how 
practical trading strategies worked. they also left with the conviction 
that they could do better. A month later, they gave notice to their re-
spective employers. It was time to enter the fray.
Building a company is different from building a radio or a motorcycle 
engine, or even a computer to beat roulette. But many of the same 
skills prove useful: the vision to see how to pull the pieces together 
in a new way; a tolerance for tinkering with something until you can 
make it work; unflagging persistence. Making something new is ad-
dictive, which might be why so many entrepreneurs are engineers and 
scientists.
farmer and Packard were also motivated by a strong antiestablish-
mentarian bent stretching back to their days as eudaemons. the new 
company wasn’t designed as a first step into the financial world — it 
was part of a plan to upend it, to take Wall Street for all it was worth 
by being a little smarter, a little more conniving, than the suits. It was 
a company founded in much the same spirit as the roulette project, 
a Yippie adventure and a return to a culture of pure research and no 
rules. farmer wore an eat the rich t-shirt to the new company’s 
first formal meeting, in March of 1991.
But there was more at stake here than in roulette. farmer and Pack-
ard wanted the project to work, and they were willing to consider the 
possibility that real business acumen could be useful to them. So they 
brought in Jim McGill, a former physicist turned entrepreneur, as a 
third partner. In 1978, McGill had founded a company called digital 
Sound corp., which specialized in the kinds of microchips necessary 
to process data from electric musical instruments and microphones, 
and then later branched out into voice-mail devices. McGill was, at 
least nominally, the ceo of the Prediction company, the business face 
of their Birkenstock-and-blue-jean outfit. farmer and Packard were 
perfectly adept at imagining what they would do with, say, a hundred 
million in capital. McGill’s job was to find someone to give it to them. 

---

McGill would be the difference between the Prediction company and 
a rerun of eudaemonic enterprises.
It quickly turned out, however, that finding would-be investors 
wasn’t as difficult as the founders imagined it would be. farmer and 
Packard had earned reputations during the days of the Santa fe Insti-
tute’s economics conferences. When rumors began flying that farmer 
and Packard were leaving academia to take on Wall Street, some in-
fluential people took notice. farmer had to buy a new suit to look 
presentable for meetings at places like Bank of America and Salomon 
Brothers. things got even better after the New York Times Magazine 
ran a cover article called “defining the new Plowshares those old 
Swords Will Make,” on how physicists, who had largely been absorbed 
into the military-industrial complex in the wake of World War II, were 
branching out as the cold War came to an end. the article led with the 
Prediction company — a perfect tie-in, given farmer’s history with 
Los Alamos. After it appeared in print, hundreds of suitors began to 
call, from rich oil men to Wall Street banks.
the trouble wasn’t getting money. It was what the would-be inves-
tors wanted in exchange. Some of the Wall Street outfits were thrilled 
with the idea of starting a hedge fund based on the Prediction com-
pany’s ideas. But farmer and Packard didn’t like the idea of traveling 
the country trying to raise capital, which they would need to do if 
they were managing a hedge fund. Ideally, they wanted seed money so 
they could focus on developing the science. other companies wanted 
to buy the Prediction company outright — equally unappealing to a 
group of men who had just made up their minds to break out of the rat 
race and start their own business. Some companies were willing to put 
up capital in exchange for a portion of the proceeds, but they wanted 
more than just a return on their investment. for instance, david Shaw, 
a former computer science professor at columbia who had started his 
own hedge fund, d. e. Shaw & co., in 1988, wanted to own the com-
pany’s intellectual property in exchange for a few years’ worth of seed 
money.
Many of these offers were appealing. But farmer and Packard con-
tinued to balk. nothing felt right. Unfortunately, they couldn’t run an 
148 
•
t h e p h y s i c s o f wa l l s t r e e t

---