科学家有钱以后，不是一般的吓人——论D.E.Shaw的牛逼人生When Scientists Get Rich: The Improbable Life of D.E. Shaw

A Reader's Guide

I hope we get more scientists—and more "filthy-rich" scientists like D.E. Shaw, who use technology to remake society and drive human progress with unstoppable force.

01 History Really Is a Little Girl You Can Dress Up However You Like

In the 1980s, an obscure hedge fund set off a revolution that would go on to utterly reshape and reinvent Wall Street and America's financial industry.

That fund was D.E. Shaw, the quantitative hedge fund founded by the quant pioneer David Shaw.

More than 98% of people have never heard of the firm. No matter—just look at who has worked at D.E. Shaw and you'll know how blindingly brilliant, how formidable, this place is.

1. Jeff Bezos. Yes, that Bezos. After graduating from Princeton in 1986, he joined FITEL, a New York high-tech development company, working on computer systems. In 1988 he joined Wall Street's Bankers Trust Co. as a vice president. From 1990 to 1994 he was at D.E. Shaw & Co., making SVP in 1992. At thirty, having climbed to SVP, he quit to start his own company: Amazon.
2. John Overdeck and David Siegel: two former D.E. Shaw employees—one a statistician, one a computer scientist—who founded another blindingly brilliant hedge fund, Two Sigma Investments, whose AUM is now roughly 1.5 times D.E. Shaw's.
3. Lawrence Summers: American economist, the 71st U.S. Treasury Secretary under Clinton. He won the John Bates Clark Medal for his work in macroeconomics, served as the 27th president of Harvard from 2001 to 2006, and was chief economist of the World Bank.

The company's own website notes: 87 PhDs, 25 employees holding IMO medals, employees who speak 64 languages.

For its 50th anniversary, New York Magazine ran a special feature reconstructing the pivotal historical moments that shaped New York's urban culture. And D.E. Shaw was one of the boldest strokes in that history.

Let's rewind and see how D.E. Shaw managed to master both scholarship and money-making—firm grips on both—and rise to the summit of life.

Shaw is an academic superstar, an idol to PhDs everywhere: a computer science PhD from Stanford, a Columbia professor before he was thirty, specializing in massively parallel computing.

(Figure in original.)

By now you may already be thinking that you haven't worked hard enough to have earned the right to blame it on talent.

Hold on—keep reading.

For all his early success, Shaw wasn't happy at Columbia; he found it rather boring. Columbia sits in New York, a city steeped in money and hard reality, where capital is the only thing you need.

Everywhere you looked there were newly minted hedge-fund guys living it up in wine and debauchery, cavorting among all sorts of models—while an equally brilliant professor could only sit on the cold bench writing computer models. In a scientist's eyes the latter kind of model is arguably even sexier, but that wasn't Shaw's style. So he was done.

In 1986 David Shaw left Columbia, went to Wall Street, and joined Morgan Stanley to do quant trading (loosely, using computers to automatically trade stocks, bonds, and foreign exchange).

Morgan Stanley reportedly lured him with six times his assistant-professor salary, along with a promise of financial resources he could never have gotten in academia (anyone who's slaved over research-grant applications at a university will understand this perfectly).

At Morgan Stanley he floated the novel idea of applying computer technology to investing, but a big shop like Morgan Stanley is a deeply constraining environment for pushing innovation through, and as a geek, David Shaw was nowhere near cut out for political warfare. In a place like Morgan Stanley—big money, big drama, where the political fights are as fierce as the technical ones—it took just two years before he was forced out.

Cut to the summer of 1988. The hedge fund manager Donald Sussman took a call from a former Columbia professor. The man had found a new job on Wall Street and wanted his advice. The caller was David Shaw.

Shaw admitted he was wrestling with a puzzle: his firm's fierce rival, Goldman, had extended him an offer. That's why he wanted Sussman's advice.

Sussman was experienced and had a good eye for talent, but he'd never met a raw newcomer like David Shaw—a computer expert—nor heard anything like his ideas.

Years later, not even Sussman himself had foreseen that his eye for talent that day would make not only his own fortune but also a pioneer in American financial history: David Shaw.

Over the years to come, David Shaw reshaped the whole American financial industry, replacing brilliant traders with a computing revolution. Wall Street's decades of uniform suits and ties were no longer mandatory; jeans and T-shirts were everywhere. Who's the real big shot? I don't know.

Back then they were also planning to launch an early email system, and using it to survey the prospects of online retail. In the 1990s, David Shaw once chatted with colleagues about how a mature internet would one day change people's lives.

Even then, he had the "singular foresight" to reckon that in the future people wouldn't just buy things online—while buying, they'd also comment: "this pipe is pretty good," "that pipe is worse," and then they'd post those reviews online.

He spoke without design, but a listener took it to heart. Bezos happened to be at D.E. Shaw working as a coder (you read that right—he wasn't a financier, he was a coder), and hearing Shaw say this, his mind started to wander.

Granted, when Shaw talked about reviews he was surely picturing an online store to replace Home Depot, not the bookstore Amazon first launched with. But it was from this moment on that Bezos's resolve to enter e-commerce hardened. Because that remark laid bare one of the essences of e-commerce: it's not simply about moving goods from brick-and-mortar stores online and selling them. More importantly, in a virtual economy, information—especially UGC (User Generated Content) information—plays an extraordinarily important role.

Many people probably assume Bezos was simply tired of Wall Street life and so chose to start a company on the West Coast. But if we view his choice through the lens of two or three decades, Shaw's remark may in fact have given him the spark: a kind of ebb-and-flow—East Coast finance versus West Coast high tech. Back then, though, history wasn't nearly so black and white.

What happened afterward we all know. Yet almost no one back in that era could have foreseen today's picture.

History really is a little girl you can dress up however you like.

02 Early Glory: A Thug Who Knows Kung Fu

Back to David Shaw.

Sussman thought Goldman's offer wasn't good enough for him: "If you believe your idea is viable, and you're confident in it, then you should come work for me."

Three days later, David Shaw accepted Sussman's offer.

Sussman's investment firm Paloma Partners put a hefty $30 million into David Shaw's new hedge fund, D.E. Shaw. The fund focused on quant trading, using high-speed computer networks and momentary inefficiencies in the market to run high-frequency statistical arbitrage. It hired six employees and set up in Greenwich Village, the New York artists' enclave.

Unlike today, when high-frequency trading is crawling with people, back then computer hardware and software were nowhere near as advanced. So the people who could master high-speed network programming and large-scale parallel computing—the ones capable of computing ballistic trajectories and simulating nuclear detonations—could also become the first generation to do high-frequency trading. What they did could basically be considered risk-free arbitrage: simply exploiting market inefficiencies, fleecing the market's wool, with the speed of the money depending on how fast you could shear.

For David Shaw, a genius who specialized in massively parallel computing, this was a thug who knows kung fu—nobody could stop him, his wool-shearing speed world-class. His life quickly entered a new peak. The firm went on to grow into a $47 billion giant, making its investors over $25 billion. By 2015, his personal net worth had reached $4.1 billion, putting him in the top 500 of the Forbes global wealth ranking.

The quant revolution D.E. Shaw led has become the most powerful trend in today's hedge fund industry: of the industry's $3 trillion total, quant funds account for $500 billion. Quant-driven hedge funds reign supreme, their returns towering over the rest.

(Figure in original.)

By D.E. Shaw's own account, seven of the ten biggest behemoth hedge funds are quant shops. And one of those, the well-regarded Two Sigma, was founded by former D.E. Shaw employees. This is one of D.E. Shaw's contributions.

But the change D.E. Shaw brought was hardly confined to hedge funds. That's where his true brilliance—part two—lies. Keep reading.

03 Once a Scientist Gets Rich, It's Genuinely Terrifying

As Shaw's financial empire grew, he gradually found life empty and tedious. David Shaw achieved financial freedom at around forty, and by the thinking of most of us today, he could have stopped writing models and dived headlong into New York's high life—gone dating real models. But, as the famous online essay on Newton's life puts it:

• Sir Isaac was a very powerful figure in the world of science, ennobled by the Queen and made a peer—"Sir Newton," they called him—rising to Master of the Royal Mint and President of the Royal Society. Only Albert Einstein, had he not turned down the presidency of Israel, would have been a match for him.
• To say he was powerful isn't just about high office; mainly it's about the size of his contributions. If the Nobel Prize had existed in the 17th century, Sir Newton could have monopolized at least four straight physics prizes (the spectrometer; the construction of the mechanical system; the reflecting telescope; universal gravitation), and to honor his attainments in alchemy, we'd throw in a chemistry prize too. And this rascal cooked up the method of fluxions, so the Fields Medal has to go to him as well. Bear in mind, he made basically all these discoveries before the age of 26. So you can imagine how, in the Europe of his day, a single call from him would summon a hundred responses; the only ones who dared cross him were Leibniz and Bishop Berkeley. When Sir Newton died, the nobility of all England vied for the honor of bearing his coffin, and the notables of all Europe flocked to London. The French fool of a liberal-arts student, Voltaire, was so shaken at the state funeral that he went home and wrote a poem, his envy dripping off every line.
• Sir Newton's life was a life of genius, a life of struggle, and a lonely life too. Of course he surely wasn't lonely, because the world of science is boundless in its delights, one thrill after another. Contrary to what conventional folk imagine, science is in fact far more seductive than any woman—playing science is far more of a rush than playing women. The high from arriving at a result is intense and lasting; not only is there pleasure, there's a vast sense of self-affirmation, far surpassing the boundless emptiness and desolation that follow those few seconds of shudder. Which is why Chen Jingrun was actually so besotted with beauty he couldn't pull himself free—his frail body couldn't take the excess of climaxes, and he was pleasured to death.

In an interview (original at http://queue.acm.org/detail.cfm?id=1614441), Shaw said that in D.E. Shaw & Co.'s early days he was deeply immersed, doing a great deal of research on using quantitative computation to improve investment management. But as the firm grew, he had to spend more and more of his time on management work.

He felt himself getting dumber year after year (in his words: "I could feel myself getting stupider with each passing year."). He strongly disliked the feeling, and to find some joy, he began working on some theoretical scientific problems. Because he couldn't have a team doing research together the way he had at Columbia, he could only do it alone, at night. He found himself enjoying it more and more, and as time went on he began to miss the days of doing research full-time.

The geek gene stirred and was reborn in the depths of his body. Once the old man Shaw had made his name and fortune, the nerve of emptiness began to act on him.

Around this time a friend of his—Rich Friesner, a computational chemistry professor at Columbia—complained to him that some research problems couldn't be solved because the algorithmic models and computers were too slow, and asked whether he had any ideas. Shaw started taking some of his friend's problems home to work on himself. Once he sped up an entire computation by a factor of 100, and it began to dawn on him: his computing expertise might be able to solve some of the hard problems in computational chemistry.

As he neared his 50th birthday, he began to think about how to spend the rest of his life. There's a broader backdrop here: computational chemistry had been developing for many years, always in a lukewarm sort of way. To put it plainly, computers were still too weak; computational chemistry applied to real problems couldn't compute accurately, and its precision actually fell short of just running an experiment.

So, in both chemistry and biology, anyone doing computational chemistry—professors and PhDs alike—had to either collaborate with the experimentalists and live at the bottom of the pecking order, or hide in a corner making modest, self-satisfied circles.

So he pulled a sharp about-face back into the world of science, took off the suit, and let his true geek colors show. He turned his gaze to the withering field of computational chemistry.

He recognized that molecular dynamics could greatly accelerate the understanding of biological processes at the molecular level—advancing not only basic science but also the development of drugs to treat major diseases like cancer. So in 2001 he founded D.E. Shaw Research, a private research institute, and served as its chief scientist doing research full-time.

He wanted to build a supercomputer purpose-built for computational chemistry, thousands—tens of thousands—of times more powerful than existing supercomputers.

Many might ask: today's supercomputers are already immensely powerful, basically packing hundreds of thousands of CPUs and the like—why build a special computational-chemistry supercomputer?

Because: a general-purpose computer is very smart, but not well suited to the computational-chemistry line of work.

Here's an analogy. Today's computer is a jack-of-all-trades; you can have it do anything: harvest wheat, cook, sweep the floor, and so on. It can do it all, always more efficiently than a human, and yes, it really is smart. This is what's called general-purpose computing: one machine, different software, all kinds of functions.

But on a task like harvesting wheat, this all-rounder robot can hardly beat a professional large combine harvester. The big combine harvester is dumb, but it's built entirely for harvesting wheat—so its hardware is bespoke and extremely optimized. This is what's called special-purpose computing: a specially built machine, with specially built software too, delivering only one function, but doing it ferociously and efficiently.

Shaw wanted to build a "special-purpose harvester" for computational chemistry. This harvester is called Anton.

It was very, very expensive, but Shaw didn't have to worry about the budget at all—he had more money than he could spend and nowhere to spend it (and even if he did spend it all, earning it back was a matter of minutes for him). Besides, researchers are actually cheap, so he hired a crowd of talent across biology, chemistry, computer hardware, software, and other fields to develop the Anton supercomputer.

Here's something both funny and bittersweet: in America, even though the average IQ is high, most people with pure science PhDs can't find jobs after graduating, and those with research dreams can barely scrape by on a postdoc salary.

David Shaw said, this is a small matter, let me help you out. He recruited "a whole bunch" of jobless PhDs—people who were, financially, pure have-nots—and offered them $100,000 a year.

A hundred grand a year is the entry pay-plus-bonus of a 21- or 22-year-old junior analyst at an investment bank—on Wall Street it's actually the bottom rung: dressed sharp on the outside, threadbare underwear beneath. But to this bunch of geek have-nots it was the equivalent of being kept in style, two to three times the pay they could find for a research post. Amazing. And so the very best computational chemistry, biophysics, and electrical engineering PhDs came flocking.

Starting around 2004 (corrections welcome from anyone in the know), the D.E. Shaw Research that Shaw founded formally began operations. Under David Shaw's "meticulous keeping," 30-some PhDs, wanting for nothing and deaf to the world outside, read papers for a solid year and a half in an exceedingly elegant environment, and produced a draft blueprint for Anton.

In 2007, nearly a year ahead of schedule, the have-nots and geeks who'd come from all corners released the first generation of Anton. The greatest piece of black-tech in computational chemistry was born: about 10,000 times faster than an ordinary supercomputer. And 1,000 times faster than the best supercomputer.

That's right. A monstrous 10,000 times—four zeroes, four orders of magnitude. Roughly speaking, a molecular microscope.

What does 10,000 times mean? In computational chemistry, the length of time over which you can simulate a molecule's trajectory of motion is hugely important. Use a tennis match as an analogy: before, a "supercomputer" would compute for a month and we could only simulate the one-second instant of the ball being struck; now, with Anton, we spend the same month and can simulate the trajectory of the tennis ball across the entire match.

This was unprecedented supercomputing power, a monstrous "large combine harvester"—you could say it opened up a God's-eye view.

(Figure in original.)

From 2007 on, D.E. Shaw's team shot to fame. D.E. Shaw Research's results in protein folding and molecular dynamics were published in top journals like Nature, Science, and Cell, becoming landmark achievements in those fields and sending its academic reputation soaring. Later they went on to build Anton 2, which continued to trounce the supercomputers that had "made great strides over the past eight years."

David Shaw himself was elected to the American Academy of Arts and Sciences in 2007, became a member of the Obama administration's Council of Advisors on Science and Technology in 2009, and was elected to the National Academy of Engineering in 2012 and the National Academy of Sciences in 2014.

04 So Many People Want to Live as Themselves—but Do You Really Know Who You Are?

David Shaw's life is the stuff of legend. To this day he keeps up his low-key, mysterious, spare-of-words style. He is intensely private and rarely grants media interviews. He lives simply and without extravagance; in a 1996 Fortune interview he mentioned that one of his favorite restaurants was a very ordinary Brazilian buffet across from his office.

They say money makes you free to do as you please. His idea of doing as he pleased may lie precisely in being able to follow his heart, to do the thing he loves most—which to onlookers means a string of achievements, using computers to cross over into and continually push forward finance, chemistry, biology, and social progress.

I hope we get more scientists—and more filthy-rich scientists like D.E. Shaw, using black-tech to drive science forward with unstoppable force.

05 A Closing Word

Having read through David Shaw's story, how do you feel?

How would you tell your child the story of this man: a study genius? A scientist? A financial titan? What of his traits would you help your child notice: curiosity? The boundless joy he took in studying problems? His attitude toward money? His knack for using computer technology to solve problems in other disciplines?

Whatever your reading, I think it's a good story worth savoring for yourself, and worth explaining to your kids.

References

• "The Life of the Mighty Newton"
• "What black-tech is out there in the field of computational chemistry"