In 2003, however, this formula sprang a leak. In its drive to hire the best brains on the planet, Renaissance had discovered one of the achievements of the old Soviet Union: The country recruited the brightest kids from its fifteen republics and transported them to the Institute of Physics and Technology in Moscow. There, separated from their families, these prodigies underwent intensive training, and the ones who survived this assault course found themselves internationally marketable after the Soviet system disintegrated. Renaissance recruited one of these Russians, and he recommended another one, and pretty soon the faculty in East Setauket had a sizable Russian caucus. But Renaissance’s leaders had never paused to ponder the behavioral consequences of a Soviet upbringing. Kids raised in that failing political system were likely to assume that authority is corrupt and that a person’s only obligation is to look out for himself; and kids who had been separated from their parents were all the more likely to mature as hardened individualists. Sure enough, after spending long enough on the Renaissance faculty to master its secrets, two of the Russian researchers presented Simons with an ultimatum. They refused to sign the firm’s noncompete agreement and demanded higher pay. If Simons refused, they would quit and join a rival.31
Simons refused to be blackmailed, and the Russians left to join another hedge fund. On the face of it, this looked like a catastrophic blow. Because of the open structure at Renaissance, the Russians understood a lot about how the system worked. If they started trading on Renaissance’s signals, they would siphon off part of its profits; it would be as though pirates were making generic copies of a pharmaceutical company’s blockbuster therapy. Patents do not protect financial innovations in the way that they protect medical ones, so Simons’s legal remedies were uncertain. And yet the remarkable thing was that Medallion’s performance continued to leave rivals in the dust. Like the magician who drinks poison and survives, Simons emerged looking more mysterious than ever.
How could this survival act be possible? Part of the answer may lie with Renaissance’s lawyers: By suing the Russians and their new employer, Simons may have deterred them from rolling out a rival system at full speed; and in 2006 a settlement laid down that the Russians would cease trading.32 But lawyers are not the whole answer, since the Russians did operate a rival system for two or three years, and during those years Medallion did extremely well—even if its profits would probably have been higher still without the alleged theft of intellectual property. The lesson seems to be that the infrastructure of Renaissance is as important as its research, and that the research itself advances constantly. It takes enormous amounts of time and money to set up systems that absorb a trillion bytes of data daily, that make these data accessible and malleable to researchers, and that turn the research findings into hundreds of thousands of automated trades that go off without a glitch in markets from Spain to Singapore.33 And while the Russians were struggling to create a halfway comparable platform, the faculty at Renaissance was moving on. Each Tuesday in East Setauket brought another Big Meeting and another set of fresh ideas. The Russians were running to catch up with a fast moving target.
JAMES SIMONS HAD A PARADOXICAL EFFECT ON THE REST of the hedge-fund industry. The Medallion Fund was like the Formula One race car that an auto firm might build: Most customers never got the chance to climb inside, but the existence of this mouth-watering machine encouraged them to buy ordinary vehicles. A fund that trades short-term signals cannot afford to get too large, since liquidity and time constraints prevent it from putting too much on each trade. Medallion therefore closed to new outside investors in 1993, and by the 2000s the $6 billion or so in the fund consisted almost entirely of employees’ money.34 But the very existence of Medallion had a halo effect on the rest of the industry, offsetting the blow to the reputation of black-box trading administered by the collapse of Long-Term Capital.
Each time Simons’s picture appeared on the cover of a financial magazine, more eager institutional money flooded into quantitative trading systems. Simons himself capitalized on this phenomenon. In 2005 he launched a new venture, the Renaissance Institutional Equities Fund, which was designed to absorb an eye-popping $100 billion in institutional savings. The only way this huge amount could be manageable was to branch out from short-term trading into more liquid longer-term strategies—and since pure pattern recognition works best for short-term trades, it followed that Simons was offering a fund that would rely on different sorts of signal—ones that might already have been mined by D. E. Shaw and other rivals. By the summer of 2007, the new Simons venture had raked in more than $25 billion, making it one of the largest hedge funds in the world. But then the financial crisis hit. Like almost everybody else, Simons felt the consequences.
14
PREMONITIONS OF A CRISIS
By the middle of the 2000s, the scale and persistence of hedge funds’ success was transforming the structure of the industry. The first generation of hedge-fund titans had been seen as freakish geniuses, whose eye-popping returns were possibly lucky and certainly not reproducible. But by 2005 nobody could argue that hedge funds were exceptional in any way: More than eight thousand had sprouted, and the long track records of the established funds made it hard to dismiss their enviable returns as the products of good fortune. Bit by bit, the old talk of luck and genius faded and the new lingo took its place—at hedge-fund conferences from Phoenix to Monaco, a host of consultants and gurus held forth about the scientific product they called alpha. The great thing about alpha was that it could be explained: Strategies such as Tom Steyer’s merger arbitrage or D. E. Shaw’s statistical arbitrage delivered uncorrelated, market-beating profits in a way that could be understood, replicated, and manufactured by professionals. And so the era of the manufacturer arrived. Innovation and inspiration gave way to a new sort of alpha factory.
You could see this transformation all over the hedge-fund industry. By the early 2000s, there was no longer much doubt that long/short equity stock picking, as practiced by Julian Robertson’s Tiger, could deliver market-beating returns. The challenge was not so much to invent the strategy; it was to implement it successfully. Dozens of Tiger look-alikes sprang up to do the job, many of them run by men who had themselves worked for Tiger; and an eager industry of hedge-fund consultants and funds of funds emerged to allocate capital to the most promising among them. The biggest sponsor of Robertson clones was none other than Robertson himself. After shuttering Tiger in 2000, he turned his offices into an incubator for “Tiger seeds,” which managed his money, benefited from his coaching, and used the prestige of association with the great man to raise more capital from outsiders. Under the old Tiger model, Robertson had maintained personal control of all the big investment calls, but now he let his protégés run their own shows: He had switched from inventing an investment technique to franchising it. The switch provided Robertson with a lucrative final chapter to his illustrious career. By 2006 the reinvented Tiger complex was managing $16 billion. The premises on Park Avenue grew bigger than ever.
The purest expression of the new factory chic was the so-called multistrategy hedge fund. Rather than claiming an edge in a particular investment style, the multistrategy funds began from the principle that you could develop an edge in whatever style you liked: You just had to hire the people. Like a pharmaceutical giant that vacuums up ideas from university researchers and biotech start-ups, the multistrategy factories collected multiple alpha-generating strategies under one roof, blending them together so as to diversify away risk, then shifting capital among the various styles according to market conditions. The factories talked little about invention and a lot about process; they viewed hedge funds less as vehicles for financial creativity than as financial products. A Chicago-based hedge fund called Citadel emerged as a prime exponent of the multistrategy mind-set; its goal, an executive explained, was “to see if we can turn the investment process into widget making.”1 Ken Griffin, Citadel’s thirtysomething boss, was a keen consumer of management texts. His
staff sneaked glances at the tomes on his desk so that they could brace themselves for the next six-step plan, and he pushed people out of his company with a mechanical determination. Griffin liked to compare hedge funds to buses. People get on. People get off. The bus keeps rolling forward.
The new multistrategy funds grew from babies to behemoths in the blink of an eye. Again, Citadel was a case in point. Griffin had started out trading convertible bonds from his dorm room at Harvard, and at the start of 2000, when he was still just thirty-one, he was running about $2 billion. Then the age of the manufacturer arrived and Citadel took off, so that its assets swelled to $13 billion by 2007. The firm found it could charge clients almost anything it pleased: It billed them for expenses amounting to more than 5 percent of their capital before slapping on the 20 percent performance fee.2 Griffin’s personal earnings were said to be the second-highest in the industry, just behind James Simons, and he let it be known that Citadel would one day compete with Goldman Sachs and Morgan Stanley.3 Meanwhile, Eton Park Capital Management, launched in 2004 by an ex-Goldman merger arbitrageur named Eric Mindich, offered another example of multistrategy growth. Mindich raised $3 billion in assets before even opening his doors; four years on, he was managing $11 billion. During the 1990s, all the top hedge funds had struggled with the burden of bigness, and many had returned capital to investors. But by 2007 alpha factories managing $5 billion plus accounted for 60 percent of the assets in the industry.4 A magazine published a list of all the hedge funds in the “Billion-Dollar Club.” If you were not on the list you were a nobody.
There was a powerful logic in this rush to bigness. Small companies may excel at generating ideas, but big companies excel at implementation. Once the hedge-fund industry had progressed through its garage-workshop phase, it took sleek professional outfits to bring its inventions to market. The successful alpha factories boasted state-of-the-art computers that executed lightning trades, legal departments that understood the rules in multiple countries, treasury departments that negotiated the best terms from brokers, and marketing departments that churned out glossy monthly reports to satisfy high-maintenance institutional investors. Since their edge lay in the efficiency of their platforms rather than the originality of their ideas, it was natural to use the platforms to support multiple alpha-generating strategies—and multiple strategies meant that the new funds could manage huge amounts of money. The multistrategy format responded to customer pressure too. The fund-of-funds industry, which collected money from endowments and pension funds and allocated it to hedge funds, had amassed almost $400 billion in assets by 2005, partly by promising to shift capital nimbly among different hedge-fund strategies as market conditions altered. The way MBA-minded hedge funds saw it, they could cut out the middleman. If endowments were looking for a product that would shift flexibly among strategies, multistrategy hedge funds would build the widget that the clients wanted.
And yet, for all its logic, the sudden growth of alpha factories made wise observers feel uneasy. Too many people were making too much money too fast. Opportunistic consultants staged workshops on how to open a hedge fund; a book called Hedge Funds for Dummies appeared in the stores; and grandees with no known background in asset management, such as Madeleine Albright, the former secretary of state, jumped into the industry. The frenzy recalled the extremes of the leveraged-buyout boom in the 1980s or the dot-com mania in the 1990s. Surely this bubble could not last? Wasn’t it bound to end painfully?
IN THE MID-2000S, AS THE HEDGE-FUND BUBBLE WAS growing, an outfit named Amaranth emerged as the very model of the modern alpha factory. Its founder, Nick Maounis, was a convertible-arbitrage specialist by background, but he had hired experts in merger arbitrage, long/short equity investing, credit arbitrage, and statistical arbitrage; and in 2002, following the collapse of the corrupt energy company Enron, Maounis had snapped up several stranded employees to open an energy-trading operation. Maounis made the standard arguments for this mission creep: A blend of alpha-generating strategies would diversify away risk, and Amaranth would move capital aggressively among strategies as market conditions shifted. The fund’s energetic shape-shifting was a point of pride. In the first months after Amaranth’s launch in September 2000, nearly half of its capital had been focused on merger arbitrage. A year later, that strategy had been cut to practically zero, and more than half of Amaranth’s capital was focused on convertible arbitrage. Scroll forward another year, and the portfolio began to shift into bond trades, and then into statistical arbitrage and energy. There seemed no good reason for a pension plan to hire a fund of funds when it could go directly to Amaranth, bypassing the middleman’s fees, particularly since Amaranth’s results were excellent. In its first three full years of operation, Amaranth returned 22 percent, 11 percent, and 17 percent—this at a time when the S&P 500 was mostly heading downward.
Yet for all Amaranth’s glittering appearance, there was a certain hollowness about it. Contrary to his marketing patter, Maounis had no clear edge in deciding which strategy to shift into. He upped his allocation to investment styles that had worked well recently and cut back on those that fared poorly; but there was no sure way to identify which strategies would succeed in the near future.5 Moreover, precisely because alpha had become a commodity, dozens of rival factories were driving down returns by manufacturing the same thing: Amaranth’s shape-shifting was less about cleverly timing market cycles than about desperately searching for the next trick to keep profits from tanking.6 And because Maounis was allocating capital to specialist traders whose books were difficult to understand, his decisions were necessarily affected by instinct. Gut feelings about the various traders on his team could sway decisions dangerously.
These dangers came together in the shape of a young Canadian named Brian Hunter. Standing six feet five inches tall, occasionally donning a jersey of the National Hockey League’s Calgary Flames, and equipped with a graduate degree in math, Hunter was imposing physically as well as intellectually. He was earnest, soft-spoken, and unfailingly calm, and from the moment he landed at Amaranth in 2004, his returns from trading natural gas stood out conspicuously.7 He had spotted an anomaly in winter gas prices. Unlike oil, which is shipped around in tankers, natural gas is delivered mainly in pipelines; supply routes cannot easily be changed to fill unexpected local shortages. As a result, gas prices are volatile: Time and again, a blast of cold weather would cause demand for household heating to spike, and in the face of rigid supply, prices would leap upward. Hunter’s discovery was that options whose value would shoot up in a shortage were strangely cheap—they represented bargain weather insurance. Hunter loaded up on these options, figuring he had found a classic asymmetrical trade: The most he could lose was the small cost of buying the options, but if a shock hit the market and the gas price spiked, he could earn many times more than that. Another way of cashing in on the same insight was to buy a pair of futures contracts: Hunter would go short a summer contract and long a winter contract, betting that the narrow spread between the two would widen if winter prices leaped upward. The strategy had worked in recent winters, and in November 2004 it came good again. The price of natural gas jumped to around 80 percent above its low point in the summer, and Amaranth cashed in handsomely.
In the spring of 2005, Maounis confronted an unpleasant dilemma. Many of Amaranth’s strategies were faring poorly. Convertible arbitrage had hit a wall and showed no sign of recovering. Maounis had invested heavily in statistical arbitrage, telling colleagues he wanted a piece of James Simons’s action, but he had little to show for it. The one star act in the Amaranth lineup was Brian Hunter and his winter gas; but in April, Maounis learned that Hunter had been offered a $1 million bonus to sign on with a rival firm, SAC Capital. Feeling his back against the wall, Maounis took a chance. Rather than lose his star player, he promoted him. Hunter became cohead of Amaranth’s energy desk, gaining the authority to place his own trades; meanwhile, Maounis pumped up the share of his firm’s capital allocated to the
energy desk from 2 percent the previous spring to around 30 percent. With these two decisions, Maounis effectively bet his company on a thirty-two-year-old trader who had been with him for barely one year. In the go-go atmosphere of the mid-2000s, this was the sort of thing that happened.
Hunter’s promotion was all the more remarkable given his background. He had come to Amaranth from Deutsche Bank, where he had supervised gas trading. In December 2003, his trading group had lost $51 million in a single week, but Hunter’s response was nothing if not brazen. In a suit he later brought in New York state court, Hunter ascribed the loss to “an unprecedented and unforeseeable run-up in gas prices,” as though his failure to foresee the market’s behavior rendered him blameless. He pointed to the “well-documented and widely known problems” with Deutsche Bank’s trading and risk-management software, which made it hard to exit losing trades—as though his taking of excessive risks could be laid at the door of Deutsche’s managers. Hunter also argued that even though his group had registered a loss, he himself had earned $40 million for the bank during 2003: Therefore, he deserved a bonus. By February 2004, Deutsche Bank’s managers had concluded that there was no place for Hunter on their trading team. This was the man whom Amaranth was now promoting.8
More Money Than God_Hedge Funds and the Making of a New Elite Page 36
