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

The Prediction Company 
•
157
information concerning the expected future performance of a stock
— is not necessarily in conflict with the Prediction company’s success. 
It sounds like a paradox. But think about the basis for the efficient 
market hypothesis. the standard argument goes something like this: 
Suppose that there was some way to game the markets; that is, sup-
pose that there was some reliable way to predict how prices are going 
to change over time. then investors would quickly try to capitalize on 
that information. If markets are always at a local high in the last week 
of May, or if they always drop on the Monday following a Giants vic-
tory, then as soon as the pattern gets noticed, sophisticated investors 
will start selling stocks at the end of May and buying them as soon as 
the Giants win — with the result that prices will drop at the end of May 
and rise on Mondays after Giants victories, essentially washing out 
the pattern. Sure enough, every time an economist appears to find an 
anomalous pattern in market behavior, it seems to correct itself before 
the next study can be done to confirm it.
fair enough. this kind of reasoning might make you think that even 
if markets somehow got out of whack, there are internal processes that 
would quickly push them back into shape. (of course, one of the major 
reasons to think that the efficient markets hypothesis is deeply flawed 
is the apparent presence of speculative bubbles and market crashes. 
Whether these kinds of large-scale anomalies, where prices seem to 
become unmoored, are predictable is the subject of the next chapter. 
Here I am thinking of smaller-scale deviations from perfect efficiency, 
supposing that such a thing exists.) But what are these internal pro-
cesses? Well, they involve the actions of so-called sophisticated inves-
tors, people who are quick to identify certain patterns and then adopt 
trading strategies designed to exploit those patterns. these sophisti-
cated investors are what make the markets random, at least according 
to the standard line. But they do so by correctly identifying predictive 
patterns when they arise. Such patterns might disappear quickly. But 
if you’re the first person to notice such a pattern, the argument about 
self-correcting markets doesn’t apply.
What does this mean? It means that even if you take the standard 
line on efficient markets seriously, there is still a place for sophisticated 
investors to profit. You just need to be the most sophisticated investor, 

---

the one most carefully attuned to market patterns, and the one best 
equipped to find ways to turn patterns into profit. And for this task, a 
few decades of experience in extracting information from chaotic sys-
tems plus a room full of supercomputers could be a big help. In other 
words, the Prediction company succeeded by figuring out how to be 
the most sophisticated investor as often as possible.
of course, not everyone buys the idea that markets are efficient. 
farmer, for one, has often criticized the idea that markets are unpre-
dictable — and with good reason, since he made his fortune by pre-
dicting them. Likewise, wild randomness can be a sign of underlying 
chaos — which, perhaps counterintuitively, indicates that there is often 
enough structure present to make useful predictions. And so what-
ever your views on markets, there’s a place for the Predictors. It’s no 
surprise, then, that droves of investors have followed in farmer’s and 
Packard’s pioneering footsteps. In the twenty years since the first com-
puters arrived at the door of 123 Griffin Street, black box models have 
taken hold on Wall Street. they are the principal tool of the quant 
hedge funds, from d. e. Shaw to citadel. the prediction business has 
become an industry.
158 
•
t h e p h y s i c s o f wa l l s t r e e t

---

D
idier sornette looked at the data again. He rubbed his 
forehead thoughtfully. the pattern was unmistakable. Some-
thing was about to happen — something big. He was sure of 
it, even though predicting such things was notoriously difficult. He 
leaned back and looked out the window of his office at the University 
of california, Los Angeles geophysics institute. Such a tremor could 
have substantial consequences. the question, though, was what to do 
about it. Should he issue a warning? Would anyone believe him? And 
even if anyone did, what could be done?
It was late summer 1997. Sornette had been working on this theory 
for years now, though the idea of applying it in the present context 
was new. Still, he had had ample time to test it with historical data. In 
each instance, before a major event, he had seen this same characteris-
tic pattern. It looked like a wavy line, but with the oscillations getting 
faster and faster over time, the peaks becoming closer and closer to 
one another as though they were all trying to bunch up around the 
same point. the critical point. Sornette had found, both theoretically 
and experimentally, that these patterns should be robust enough to 
make predictions, to project when the critical point would occur. the 
Tyranny of the Dragon King
c H A P t e r 7

---

same pattern appeared all over the place: before earthquakes, before 
avalanches, before certain kinds of materials exploded. But this time it 
was different. this time, Sornette actually saw the pattern in advance. 
It was the difference between realizing a prediction was possible — a 
risk-free endeavor — and actually making it. But Sornette was confi-
dent. He would be willing to bet on this.
He picked up the phone and called his colleague olivier Ledoit. Le-
doit was a young faculty member at the Anderson Graduate School of 
Management at UcLA. Sornette told his friend what he had found. 
the data showed that a critical event was coming. earth-shattering, 
perhaps, but not geological: this event would be a potentially dramatic 
crash of the world’s financial markets. Sornette could even say when it 
would happen. His calculations put it at the end of october, just a few 
months away.
Sornette had been working his way into finance for several years, 
but even so he was still a physicist. Ledoit knew the financial indus-
try and could help him figure out the next steps. the two settled on 
a plan. first, they would file their warning with the authorities. Sor-
nette and his postdoctoral researcher at UcLA, another geophysicist-
cum-economist named Anders Johansen, wrote a notice and sent it to 
the french patent office. no one would believe them now, of course
— none of the traditional methods of analyzing markets pointed to in-
stability. And if they waited until after the crash, no one would believe 
them either, though for a different reason: their voices would be lost 
among the thousands of economists and investors who would insist 
they had seen this coming. the patent filing would be their insurance 
policy, their proof that they really had made the prediction, over a 
month before the crash. the notice was filed on September 17, 1997. It 
predicted a market crash in late october of that same year.
the second step? Profit. It’s easy to make money when markets are 
rising. But in many ways, a market crash is an even more dramatic 
profit opportunity, if you can see it coming. there are several ways to 
make money off a crash, but the simplest way is buying put options. 
the options I discussed earlier are known as call options. You buy the 
right to purchase a stock at some fixed price, called the strike price, at 
some time in the future. If the market value of the stock goes above 
160 
•
t h e p h y s i c s o f wa l l s t r e e t

---