22. Mark Landler, “Year of Living Dangerously for a Tycoon in Indonesia,” New York Times, May 16, 1999.
23. Dorinda Elliott, “The Fall of Uncle Liem,” Newsweek, June 15, 1998.
24. Shoeb Kagda, “Stanchart, M’sian Plantations Among Shortlisted to Buy BCA,” Business Times (Singapore), November 29, 2001.
25. Andrew Spokes, interview with the author, July 25, 2008.
26. Meridee Moore recalls of Spokes, “He sat in our office in San Francisco for eight months. The women here were just falling all over themselves. He was the most desirable bachelor in town.” Moore interview.
27. CalPERS announced it would stay out of Indonesia in February 2002. Craig Karmin and Sarah McBride, “Calpers Pulls Out of 4 Countries, Dealing Blow to Southeast Asia,” Wall Street Journal, February 22, 2002.
28. Spokes interview.
29. In a further pleasing detail, the government bonds paid a floating rate, so that BCA’s owners would be hedged against changes in interest rates on bank deposits. Spokes interview.
30. Data are from Bank Indonesia’s Web site, for years ending March 31, 2002, 2003, and 2004.
31. As Indonesia sought to reduce its debt burden with the Western donors of the Paris Club in April 2002, many, including U.S. executive director of the IMF Randal Quarles, cited the BCA deal as evidence that Indonesia was worthy of fresh IMF support. Andrew Spokes recalls: “We were really our own catalyst. It was event driven, and we were our own event, because that transaction pretty much turned around that entire market.” Spokes interview.
32. Deborah Frazier, “Underground Water Plan Has a Friend in an Old Foe,” Rocky Mountain News, October 4, 1996.
33. Gary Boyce, interview with the author, July 23, 2008.
34. Steyer interview.
35. Mark Wehrly recalls, “We were successful in basically polarizing every single constituency against us. So we got the politics wrong. I think we got the science right, but the world wasn’t ready for it, and we were doing a horrible job persuading them that this was a good idea. So we retrenched.” Wehrly interview.
36. “[Yale president Richard Levin] was misled, and I think that the school was misled by Farallon.” Joe Light, “Ranch Deal Prompts Donation, Reevaluation,” Yale Herald, February 1, 2002.
37. Andrea Johnson, who acted the role of the transparency fairy, recalls, “Obviously it was goofy, but you do these things for the photo op.” Andrea Johnson, interview with the author, June 30, 2008.
38. Steve Bruce, Farallon’s public-relations adviser, emphasizes the efforts to protect the salamanders. “They hired an environmental engineering firm to come in and do a study on salamanders: where they hatch their eggs, where they move them, how do they get to the beach, what sort of pesticides do you have to use, how do you keep the course in place without screwing up their breeding facilities. So by the time the critics brought it up, this was a red herring disguised as a salamander.” Steve Bruce, interview with the author, June 25, 2008.
39. “Farallon Founder Hits Back at Critics,” Financial News (Daily), March 28, 2004.
40. “I just remember David Swensen being really angry. It was very clear to me that he found our campaign extremely upsetting. It was personal to him, because he had received so many accolades even then, and it has only gotten more since then for his incredible management of the investment of the endowment. And I think more than that, there is a sense of pride in that endowment office that they are managing a nonprofit institution’s money and that they have standards. I got the sense that he didn’t feel like he invests in just whatever.” Johnson interview.
41. Swensen’s altercation with the students is recalled by Andrea Johnson and is captured in a news photograph and story. See Tom Sullivan, “Yale Defends Record Privacy,” Yale Daily News, April 5, 2004; Johnson interview.
42. Meridee Moore recalls, “You have to get out there and figure out what the potato farmers are going to do. We weren’t on the ground that much. That turned out to be much more important than we expected.” (Moore interview.) Mark Wehrly, the Farallon general counsel, says, “Once in a while you end up with the wrong partner, and we did there, and it cost us.” (Wehrly interview.)
43. Swensen himself argued that illiquid markets offered bargains. “Success matters, not liquidity. If private, illiquid investments succeed, liquidity follows as investors clamor for shares of the hot initial public offering. In public markets, as once-illiquid stocks produce strong results, liquidity increases as Wall Street recognizes progress. In contrast, if public, liquid investments fail, illiquidity follows as investor interest wanes. Portfolio managers should fear failure, not illiquidity.” Swensen, Pioneering Portfolio Management, p. 89.
CHAPTER THIRTEEN: THE CODE BREAKERS
1. This figure is net of fees, which were considerable. Rather than charging clients a management fee of 1 percent or 2 percent and keeping 20 percent of the investment gains, Medallion charged a management fee of 5 percent. Sandor Straus, a mathematician who was a partner at Renaissance Technologies and its antecedents between 1980 and 1996, recalls that the 5 percent fee was chosen in 1988 because that was what was needed to cover technology expenses. In addition, the Medallion Fund charges a performance fee that has increased over the years from 20 percent to 44 percent. Sandor Straus, interview with the author, July 25, 2008.
2. Elwyn Berlekamp, interview with the author, July 24, 2008.
3. Sandor Straus regards Henry Laufer as the most important contributor to Medallion’s success, notably because of the work he did starting in 1989. Laufer did his breakthrough work on short-term patterns between 1983 and 1985, according to Straus. Laufer then went back to academia for a while before reengaging with Medallion in 1989. Some public accounts erroneously state that Laufer’s involvement began in the 1990s. Straus interview.
4. Robert Frey, a Renaissance alumnus, describes the firm’s pattern recognition as neither mean reverting nor trend following. Rather, in response to a shock, the market moves around in multiple ways: “If I think of an electrical circuit or any sort of physical system, if I put an input in, the initial output may be negatively correlated to the input, and then it may become positively correlated. It depends on how the thing resonates through the system.” Robert Frey, interview with the author, July 28, 2008.
5. Straus interview.
6. Mark Silber, interview with the author, July 30, 2008. Silber is the chief financial officer of Renaissance Technologies.
7. Eric Wepsic, interview with the author, January 28, 2009. Wepsic is a member of D. E. Shaw’s six-person Executive Committee.
8. Richard Bookstaber, A Demon of Our Own Design: Markets, Hedge Funds, and the Perils of Financial Innovation (Hoboken, NJ: John Wiley & Sons, 2007), p. 187.
9. Ibid., p. 189.
10. Trey Beck notes, “At Morgan Stanley, they had a whole bank of IBM mainframes. When we started we had one Sun workstation. We did not need NASA technology because we did not expect other people to be making the same trades.” Trey Beck, interview with the author, August 31, 2009. Beck is a managing director at D. E. Shaw.
11. Michael Peltz, “Computational Finance with David Shaw,” Institutional Investor 28, no. 3 (March 1994): pp. 92–94.
12. The economists’ idealized models created different versions of this vulnerability in different parts of the world. Trey Beck of D. E. Shaw cites an example from emerging markets: Two apparently equivalent bonds issued by the same government might trade at different levels, tempting an arbitrage-minded economist to bet on convergence. But the difference might reflect a factor omitted from the economist’s model: Perhaps the more expensive bond was substantially owned by a well-connected oligarch, with the result that its default risk was far lower because the government would not wish to alienate him. Beck interview.
13. Wepsic interview.
14. Trey Beck, interview with the author, September 2, 2009.
15. Wadhwani recalls, “Like a lot of ex-academic economists, I was very drive
n by value. And I guess the key thing I learned from observing great traders was actually that value is a great medium-term factor, but tactical trading is about many, many, many more things other than value. On average, be in the direction of value, but you want to pay attention to all these other things too.” Sushil Wadhwani, interview with the author, July 28, 2009. See also Steven Drobny, Inside the House of Money: Top Hedge Fund Traders on Profiting in the Global Markets, (Hoboken, NJ: John Wiley & Sons, 2006), p. 174.
16. Wadhwani recalls, “Often it was the case that you were already using the input variables these guys were talking about, but you were perhaps using these input variables in a more naive way in your statistical model than the way they were actually using it.” Wadhwani interview.
17. Mahmood Pradhan, who worked with Wadhwani at Tudor, elaborates: “There are times when particular variables explain certain asset prices, and there are times when other things determine the price. So you need to understand when your model is working and when it isn’t. For example, sometimes current account deficits have a strong bearing on exchange rates. But other times people are quite willing to tolerate very large current account deficits because of some new preoccupation that is not in your model. Sovereign wealth funds may have emerged. Or the Asians have more capital. Or something else is going on that you may not be capturing.” Mahmood Pradhan, interview with the author, April 29, 2008.
18. Mark Dalton, interview with the author, September 29, 2008. Dalton is the president of Tudor.
19. Eric Wepsic of D. E. Shaw confirms, “Our staff started on average a little younger, a little more right out of school, a lot of people who had just got their PhDs, or people like me who didn’t even have a PhD.” (Wepsic interview.) One of the few exceptions to the Renaissance rule of not hiring from Wall Street was Robert Frey, a mathematician who had been at Morgan Stanley.
20. Wadhwani interview.
21. See, for example, Peter F. Brown, Stephen A. Della Pietra, Vincent J. Della Pietra, and Robert L. Mercer, “The Mathematics of Statistical Machine Translation: Parameter Estimation,” Computational Linguistics 19, no. 2 (1993). As noted below, the Della Pietra brothers followed Brown and Mercer from IBM to Renaissance Technologies.
22. As far back as 1949, code breakers had wondered about the application of their technique to translation. But they lacked computing power; statistical translation depended on feeding a vast number of pairs of sentences into a computer, so that the computer had enough data from which to extract meaningful patterns. But by around 1990, statistical translation was possible on a well-equipped workstation.
23. Mercer had spent a couple of summers at IDA working with Nick Patterson, a British mathematician who went on to join Simons. Simons’s connection to Baum also helped him persuade Brown and Mercer to join up. Peter Brown recalls: “When Bob and I were contacted by Jim Simons we hadn’t heard of him. But when we heard he had worked with Lenny Baum we started to take the offer seriously.” Peter Brown, interview with the author, July 28, 2008.
24. An account of the reaction to the Brown-Mercer work is given in Andy Way “A Critique of Statistical Machine Translation.” In W. Daelemans and V. Hoste (eds.), Journal of Translation and Interpreting Studies: Special Issue on Evaluation of Translation Technology, Linguistica Antverpiensia, 2009, pp. 17–41.
25. See, for example, Pius Ten Hacken, “Has There Been a Revolution in Machine Translation?” Machine Translation 16, no. 1 (March 2001): pp. 1–19. This source erroneously attributes the quote on firing linguists to Peter Brown.
26. The initial versions of the IBM program included no linguistic rules at all. Later versions did use some, but they played a far smaller role than in traditional translation programs.
27. Wepsic interview.
28. John Magee, a leading technician of the 1950s, made a point of reading the newspapers two weeks late in order to be sure that knowledge of the economy would not cloud his judgment.
29. Mercer says, “We will contemplate any proposed signal. But if somebody comes with a theory that does not make intuitive sense, we would examine it especially carefully.” (Robert Mercer, interview with the author, July 28, 2008.) The same willingness to trade on statistical evidence was shared by earlier contributors to Medallion’s success. For example, Elwyn Berlekamp recalls, “Mostly we looked at statistics at Medallion. We found that attempts to look at fundamentals did not get us very far.” Elwyn Berlekamp, interview with the author, July 24, 2008. It is also interesting that Brown and Mercer’s coauthors who followed them to Renaissance, Stephen and Vincent Della Pietra, explicitly presented their experience with statistical machine translation as relevant to finding order in other types of data, including financial data. See Adam L. Berger, Stephen A. Della Pietra, and Vincent J. Della Pietra, “A Maximum Entropy Approach to Natural Language Processing,” Computational Linguistics 22, no. 1 (March 1996): pp. 39–71.
30. To manage the potential linguistic chaos resulting from this permissiveness, neologisms had to be submitted to a review. Mercer interview.
31. The Russian employees were Pavel Volfbeyn and Alexander Belopolsky. The firm that they defected to was Millennium. They argued through their lawyer that their new system was not based on proprietary secrets from Renaissance. See Thomas Maier, “Long Island’s Richest Man from Math to Money,” Newsday, July 5, 2006, p. A04.
32. Silber interview.
33. Robert Frey explains, “Those researchers were sort of like hothouse flowers. They sit there. If they need data, the data are provided. They have no clue of the hoops you have to jump through to make sure that the data are available and clean and ready. There are tens of terabytes of data available at the touch of a button. Someone going out, who left the greenhouse, so to speak, and went out into the cold, cruel world, I think would quickly find out that even if you could produce these simulations and do all of this stuff, which isn’t trivial, you wouldn’t have access to the historical data. You wouldn’t really know how to call up somebody and execute a trade. If you said to me, Robert you don’t have a noncompete agreement and we want you to recreate Renaissance, it would probably be four or five years before you could get to a point where you could actually trade.” (Frey interview.) It should be said, however, that Medallion defectors who join a rival hedge fund that has research and trading infrastructure already in place could damage Medallion in well under five years.
34. The $6 billion number is for 2007 and is given in Richard Teitelbaum, “Simons at Renaissance Cracks Code, Doubling Assets,” Bloomberg.com, November 27, 2007.
CHAPTER FOURTEEN: PREMONITIONS OF A CRISIS
1. Hal Lux, “Boy Wonder,” Institutional Investor, January 2001.
2. The expense ratio for Citadel investors averaged just under 9 percent in the three years 2005 to 2007. Expenses covered costs such as brokerage, legal fees, tax and audit fees, and the building out of Citadel’s computer infrastructure, which partially supported trading businesses whose profits flowed entirely to Citadel, not to outside investors. Meanwhile, other hedge funds found they could raise fees too. In November 2002, Bruce Kovner’s Caxton announced that it would be hiking its fees to 3 percent of the principal plus 30 percent of the performance. Caxton said it needed to raise fees because of the competitive atmosphere in luring trading talent. See “Caxton to Hike Fees, Merge Funds,” Private Asset Management, November 24, 2002.
3. Marcia Vickers, “Ken the Conqueror,” Fortune, April 16, 2007, p. 80.
4. Gregory Zuckerman, “Shake-Out Roils Hedge Fund World,” Wall Street Journal, June 17, 2008.
5. An Amaranth veteran recalls, “Typically at Amaranth when traders made money, they were allowed to keep that money in their portfolio, rather than saying, ‘Oh, great, you just made a billion dollars for the firm, now we’re going to take that and give it to the guys in converts.’ That would not have been the way to motivate people.”
6. A senior Amaranth executive recalls, “In 2003, 2004, 2005, multistrategy arbitrage returns were getting s
maller. The business was getting saturated; the trades were getting crowded.”
7. An Amaranth veteran recalls of Hunter: “He was incredibly intelligent. Just incredibly intelligent. Brilliant in terms of his analysis of, ‘Okay, we think this is going to happen, and here’s how we can use the various instruments out there to take advantage of that.’ And just finding very interesting little market movements, submarket movements, and things going on and how to profit from those. And also how to construct what people like to describe as asymmetrical risk profiles. And people had a tremendous amount of respect for him because he could sort of make those arguments, and then when he implemented them, they were actually incredibly profitable.”
8. One newspaper account reported, “Mr. Maounis says the firm knew of Mr. Hunter’s history at Deutsche Bank but did extensive checks and found ‘nothing that made us uncomfortable.’” See Ann Davis, “Private Money: The New Financial Order—Blue Flameout: How Giant Bets on Natural Gas Sank Brash Hedge-Fund Trader,” Wall Street Journal, September 19, 2006.
9. According to one insider, Hunter’s compensation for 2005 consisted of $75 million in cash and $50 million in deferred compensation.
10. An Amaranth veteran recalls Maounis saying of Hunter, “Don’t you think he is a genius?” Another Amaranth veteran says of Maounis, “What I believe is, he must have said, ‘Brian’s book is like a zero-premium convertible book.’ So even though notionally it looked large, it’s really not that risky. So with hindsight everyone’s saying, ‘What in the fuck, are you crazy? Look at the size of this thing!’ But the risk guys must have convinced Nick that even though notionally it was very, very large, from a risk perspective it was very, very small. Because that was Nick’s upbringing. That was how a convertible bond portfolio could be.”
11. A senior Amaranth executive recalls, “Nick was always very jealous of, envious, as we all are, of Jim Simons’s ability to manufacture money with the Medallion fund. We spent a lot of money building stat-arb systems, hiring stat-arb people. Didn’t even get in the same universe as that, but he kept trying and trying, looking for the holy grail. Nick had the true belief that there were certain people who were truly special at what they did. And he thought that Brian Hunter was truly special.”
More Money Than God_Hedge Funds and the Making of a New Elite Page 57
