The Billionaire's Apprentice: The Rise of the Indian-American Elite and the Fall of the Galleon Hedge Fund

by Anita Raghavan

Meanwhile, Gupta’s outside activities, while he was still affiliated with McKinsey, did not go unnoticed. Ironically, though, it was not his fund-raising for NSR that came to McKinsey’s attention but rather his desire to join the board of Goldman Sachs. In 2005, Lloyd Blankfein’s predecessor and former secretary of the Treasury Henry M. “Hank” Paulson Jr. had approached Gupta about joining the Goldman board of directors. Paulson had first met Gupta when the two were working on an assignment for Sara Lee. Many years later, in 1996, John Bryan, Sara Lee’s chief executive, reintroduced the two at his sixtieth-birthday party at his home in Chicago. When Gupta moved to New York, the two became better acquainted. They both served on a Harvard Business School alumni advisory committee, and one year Gupta invited Paulson, a nature lover, on a trip to Indonesia to see the Komodo dragons, a large species of lizard.

  The first time Paulson approached Gupta about joining the Goldman board, he appeared eager but then demurred. McKinsey would not waive the restriction it had prohibiting partners from joining corporate boards. But in November 2006, Gupta retired as a partner and became a consultant to the firm. Almost immediately he stepped onto the Goldman board, taking up the first of many board seats. In 2007, he was named a director on the board of consumer products giant Procter & Gamble. Two years earlier, Gupta helped P&G’s then chairman and CEO A. G. Lafley sketch the company’s $57 billion takeover bid to buy Gillette. And when the talks hit a roadblock, Gupta, who was close to Gillette’s then chief executive James Kilts, helped bring the two back to the negotiating table. He sat for hours with the men in a room hashing out their differences until a deal was struck. Kilts remembers Gupta’s role to this day. At a thorny point in the talks, there was only one person Lafley and he trusted to call: “That person was Rajat.”

  As a former managing director of McKinsey and a consultant to the firm now, Gupta still retained many of his old perks, including his Stamford office, his countless McKinsey phones and BlackBerry, and his access to the company’s email, which he would rely on to solicit investors for NSR.

  Tireless in his fund-raising, Gupta turned to his vast database of contacts to find people to pitch. He fired off dozens and dozens of impersonal, form-like letters to McKinsey partners he had worked with at the firm as well as to corporate chieftains he had come to know over the years and even to neighbors. “I hope you remember me,” he started one letter he emailed to Robert Devlin, the former chief executive of insurer American General. A letter he sent to former Citigroup chief Sandy Weill, whom he knew from being on the board of Weill Cornell Medical College, was emblematic of the marketing push: “Dear Sandy, I am the chairman of a new 1.5 billion dollar private equity fund called New Silk Route (NSR) and I would very much like you to consider investing in it.” Gupta went on to lay out brief profiles of his investment partners, noting that they had put in $125 million of their own money. “We are now in the process of offering the final $200 million to a few friends and family in the range of $1–5 million each and hope to have a final close in the next 30 days.”

  As part of the NSR fund-raising drive, Gupta sent similar letters to Ajit Jain of Berkshire Hathaway, a top lieutenant of Buffett’s and a contender to succeed him; Herb Allison of TIAA-CREF; Goldman board members John Bryan and Bill George; Marcus Wallenberg, a member of the prominent Swedish family; PepsiCo chief executive Indra Nooyi; and dozens of others. The efforts paid off, and NSR succeeded in raising $1.3 billion for its private equity arm; its capital drive for the hedge fund side failed, though, and that business was never launched.

  It was around this time that Gupta’s relationship with Rajaratnam deepened too. He and Kumar started advising Rajaratnam on ways that he could boost his assets at Galleon, and in late 2006, Gupta moved to exercise his option to lift his stake in Voyager, which had racked up strong returns. He borrowed $5 million from Rajaratnam to buy out Trehan’s stake in the investment vehicle, giving him a 20 percent stake in the heavily leveraged entity. He also helped facilitate a high-interest bridge loan from Rajaratnam to his friend Ramesh Vangal, now of Katra Holdings, whose company, Scandent, Gupta had invested in many years ago and had made a lot of money on. Gupta and Vangal were working to buy a bank in South India called Tamilnad Mercantile Bank, which they ultimately succeeded in purchasing with other investors in May 2007. For a while in December 2006, as Vangal dragged his feet on repaying the loan, Gupta’s relationship with Rajaratnam turned thorny.

  On Thursday, December 21, 2006, after not hearing a peep from Vangal, Rajaratnam fired off an angry email to Gupta: “We still have not received total payment for Katra Finance, nor has Mr. Vangal contacted me as he indicated to our CFO yesterday. Under the circumstances I am not able to meet you for lunch today and I am cancelling all further meetings for TAJ Capital.” The email was signed formally, “Regards, Raj Rajaratnam,” and not “Raj,” as he normally signed missives to Gupta.

  The chasing of the money came down to the wire. On December 28, George Lau, a Galleon executive, sent an email to Gupta from his BlackBerry saying, “Our team has been chasing Ramesh’s team on a daily basis and we still have not received a 100% signal that the wire has been sent out. Tomorrow is the last day of the month and the funds have to be in our account. Please advise what else I can do??”

  Gupta replied, “Rest assured please tell Raj that he should also believe I have been working nonstop the past 7 days to make this happen.” When Vangal repaid the loan not long after, Gupta and Rajaratnam got back to lunching together. But the balance of power in the relationship had changed.

  Chapter Twenty-Three

  The Million-Dollar Man

  “Anil, this can wait until you return from London. We trust you!” emailed Peg O’Malley, the executive assistant to AMD’s chief administrative officer and top lawyer, Tom McCoy, on September 28, 2005. O’Malley had just sent Kumar a confidentiality agreement for a top secret assignment called Project Super Nova that AMD was in the early stages of hatching. The company was looking to hire McKinsey as an adviser. As was often the case, Kumar was traveling overseas and was not in a position to sign the nondisclosure agreement immediately, but O’Malley was not worried. Kumar was one of the most trusted and discreet advisers AMD retained, and no one had any reason to doubt his loyalty to the company.

  The explosive growth of video games and other graphics-driven computer applications laid bare a gaping hole in AMD’s panoply of products. Both AMD and its archrival, Intel, were in the business of making the “brains,” or the microprocessor, of a computer. Both specialized in chips that did the computational functions of a personal computer, the “left brain,” if you will. Neither owned the market in graphics chips, or the so-called right brain functions, which had exploded with the growing use of computers as game consoles.

  Intel had a slight edge over AMD. It already sold integrated chip sets with built-in graphics, allowing it to offer hardware companies a package of chips that worked well together instead of having its customers buy one chip from one company and another chip from a different firm. AMD, by contrast, relied on graphics chips from the two biggest manufacturers, Nvidia, a Santa Clara, California–based company, and ATI Technologies, an Ontario, Canada–based company. Besides developing graphics chips for personal computers, ATI designed the chips used in Nintendo’s GameCube and Microsoft’s popular Xbox.

  A looming strategic challenge for AMD if it was to close the gap between it and Intel was to enlarge its graphic chips footprint by buying or partnering with either Nvidia or ATI Technologies.

  Few people within AMD knew about Project Supernova, the code name for a possible merger of AMD with either Nvidia or ATI Technologies. The plan was so closely guarded that Kumar, who had already signed a general agreement with AMD not to disclose confidential information about the company, was required to enter into another nondisclosure agreement that would cover his work on Project Supernova. Even the confidentiality agreement was confidential. Sharing it or copying it was not permitted without the prior permission of Bharath
Rangarajan, the AMD strategy czar.

  The reason for the high level of paranoia was simple. In any merger deal, secrecy is paramount. Companies abhor leaks because they can kill a deal. When word spreads that one firm is looking to buy another, the stock price of the target company will jump, making the acquisition more expensive, possibly prohibitively so. At AMD, executives were so nervous that they even changed the code name from Project Supernova to Go Big.

  It was not long after Kumar signed the confidentiality agreement that he started divulging details of AMD’s supersecret plan to buy a graphics chip maker to Rajaratnam. On some days, he did not even wait to return to his office at McKinsey before calling Rajaratnam. Instead, he would slip into a conference room where the McKinsey team decamped. There, away from his colleagues, he picked up the phone and called Rajaratnam.

  AMD had been a favorite company of Rajaratnam’s for years. He’d tracked AMD when he started covering semiconductor firms twenty years before at Needham & Co. Rajaratnam used to keep a poster board in the office with a list of companies. Next to the companies were Galleon employees’ names. Rajaratnam expected these inner-circle Galleon analysts to cultivate corporate insiders to get revenue, guidance, and margin information for the companies they covered. If a trusted analyst didn’t call a company or his information was wrong, he was fired. Next to Rajaratnam’s name on the poster board was AMD. He made no secret of the fact that he was wired into AMD. In Galleon’s lexicon, he was “the axe,” or the authority on the stock. Often when speaking to Ali Far, who worked for Galleon in Santa Clara, California, Rajaratnam would refer to a source at the company, saying, “my guy from AMD.”

  Few at Galleon would have ever imagined his “guy” was a straitlaced McKinsey consultant as plugged into AMD as any high-level company insider. Kumar’s gleanings on AMD gave Rajaratnam a unique window into the company at a pivotal time in its history. He learned of AMD’s game-changing strategic steps and new products long before most investors were aware of the shifts, giving him an edge on how to reevaluate the stock. Most investors knew Opteron, AMD’s new chip, was getting positive feedback in the market, but few knew the impact it would have on AMD’s profitability. Thanks to Kumar, Rajaratnam kept tabs on Opteron every step of the way.

  “Are you absolutely sure?” Rajaratnam asked.

  He was rarely surprised by anything in the technology space. He was too familiar with the ever-changing dynamics of the industry and its various players to be caught flatfooted. But the potential merger Kumar told him about in the spring of 2006, a tie-up between giant semiconductor company Advanced Micro Devices and graphics maker ATI Technologies, appeared like a bolt out of the blue. In his mind, it defied logic. If AMD was going to buy a graphics company, it should be the stronger of the two, Nvidia.

  Yet Kumar was convinced the company AMD was lusting after was none other than ATI, the Canadian tech firm founded in the early eighties by a tourist who happened to be visiting from Hong Kong.

  “C’mon…I’m in the inner circle over here,” Kumar boasted. “I know everything about this company. And they’re very very keen to do ATI.”

  Over the years, the relationship between Rajaratnam and Kumar had turned rather sadomasochistic. In a savvy move, Rajaratnam taunted Kumar from time to time, chiding him that he, a hedge fund trader, knew more about AMD than Kumar did. Undercutting Kumar’s confidence in his indispensable role at AMD made him want to perform for Rajaratnam—not just for a monetary payoff, but almost for a paternalistic pat on the back.

  Now, for the first time, Kumar could show Rajaratnam just how key a player he was. Kumar was in the loop about an important development at AMD that no one was expecting—not even Rajaratnam. If AMD pulled it off, it could be a winning strategy for the company. Kumar was excited.

  He told Rajaratnam that AMD was looking at both ATI and Nvidia. If it was impossible to strike a deal with Nvidia because it was too expensive, AMD would consider buying ATI. Rajaratnam was very intrigued because the information was counter to the prevailing market view. At Galleon, Rajaratnam honed the strategy of “arbitraging consensus,” or developing a view on a stock that varied from the mainstream thinking on Wall Street. Galleon’s aim was to profit when the company’s stock price rose or fell as investors moved away from Wall Street’s herd mentality and came closer to Galleon’s way of thinking. In efficient markets, the easiest way to “arbitrage consensus” is to have inside information—a tip, for instance, that a large semiconductor company is poised to make an unexpected acquisition or that a firm is expected to report an earnings shortfall.

  For a couple of months, AMD considered both Nvidia and ATI, but by mid-December 2005, it was clear that ATI was the target. Nvidia was simply too expensive. To keep the talks tightly under wraps, the companies were assigned code names. AMD was Los Angeles and ATI was San Antonio. Sometimes code names in deals refer to the companies’ headquarters cities. But AMD and ATI were so intent on keeping the talks confidential that they picked code names designed to throw people off the trail. ATI was actually based in Markham, Canada, and the only link AMD had to Los Angeles was that its founder, Sanders, had an office in Beverly Hills.

  In anticipation of a deal early in 2006, McKinsey proposed putting together a team that would help AMD prepare for the acquisition and would give it the behind-the-scenes support it needed in its negotiations with ATI. For a full team—a partner, devoting 50 percent of his time to the AMD assignment, a full-time engagement manager, an associate, and one of the firm’s valuation experts providing support—McKinsey said it would charge $400,000 a month, prorated. AMD would also be required to cover 18 percent of the billed amount in expenses.

  On December 20, 2005, McKinsey’s Paul Roche sent Kumar an email with an update on the early talks between AMD and ATI. “Their mtg with San Antonio went well,” he wrote. Roche said the companies agreed on a timeline, proposing to reach terms by early February and unveiling the deal by the end of the first quarter. Rangarajan, the AMD vice president in charge of strategy, was eager to get McKinsey signed on to work on the project immediately. But Bob Rivet—the company’s chief financial officer, who had a reputation for never seeing a cost he liked—balked. “Think we should play it out but if he’s still squishy should talk to Bob,” wrote Roche.

  Soon after, Kumar called Rajaratnam and told him that AMD’s conversations with Nvidia were not going anywhere. However, early discussions had already started with its rival, ATI Technologies.

  This information is “red-hot,” Kumar told Rajaratnam. “Please do keep this to yourself. Don’t let anyone else know.”

  “No one is going to know about this,” Rajaratnam reassured him.

  After more than two years of feeding him information, Kumar seemed to have little sense of exactly how Rajaratnam made his money. He looked at Rajaratnam’s profession as a consultant would, thinking he was obsessed with big-picture machinations, and, much like Warren Buffett, bought and held stocks for years. If the company that Kumar tipped him on did well, Rajaratnam would make some money. For someone so smart, Kumar had little understanding of how event-driven hedge funds like Galleon really worked.

  As the talks progressed, Kumar briefed Rajaratnam regularly, telephoning him about once a month to keep him posted. At times he was almost breathless in his excitement. “ATI looks like the candidate,” Kumar gushed. “This is going to shock the industry.” From the work the McKinsey team was doing, Kumar knew that Advanced Micro Devices would be required to pay a big premium on ATI’s current share price if it was to consummate the marriage. At the time, ATI’s stock was trading at $16 to $17; it had not risen above $21 a share in the previous five years, but ATI felt the market undervalued it. To get ATI to take a bid seriously but still give AMD some maneuvering room so it could sweeten its offer if needed, the company proposed an initial offer price of $20 a share.

  When Kumar told Rajaratnam the bid, he was stunned.

  “This is completely ridiculous. It makes no sense whatsoev
er. I can’t believe AMD is going to do it,” he said. Rajaratnam was amazed that AMD would pay such a huge price for a graphics company that was not even the industry leader.

  Again he asked Kumar, “Are you absolutely sure AMD management is going to do it?”

  Ever knowledgeable on such matters, Kumar explained that ATI was a strategic must-have for AMD and the company was willing to open up its wallet to acquire it.

  “Wow, this is very useful,” said Rajaratnam. Then he went to work feeding his network of informers. Soon after Kumar briefed him on the serious nature of the talks between AMD and ATI, Rajaratnam started spreading the word to his inner circle, ignoring Kumar’s explicit request to keep the information to himself.

  “Buy some atyt,” rajatgalleon instant-messaged quintussf on April 19, 2006. “I will tell u why on fon.”

  “Quintussf” was the instant-message handle for Frank “Quint” Slattery, who ran an unregistered hedge fund in Connecticut called Symmetry Peak in which Rajaratnam invested some money. Slattery was a fervid instant messager who had been in trouble five years earlier for an IM he sent. In 2001, the software firm PeopleSoft publicly blamed Slattery, then twenty-eight years old, for suggesting that the SEC was probing PeopleSoft for accounting irregularities. Even a whiff of a regulatory investigation into accounting missteps can pummel a stock, which is exactly what happened with PeopleSoft. The company’s shares plunged 27 percent even though Slattery retracted the rumor in a subsequent instant message.

  The public dress-down did not diminish his appetite for swapping information. Several times a day, Rajaratnam and Slattery kibitzed about investment ideas—many flowing out of insider tips that either Rajaratnam or Slattery picked up from their vast array of contacts. They chatted about stocks like Google, quizzing each other on who had a “good read” on the company. Rajaratnam was the undisputed king on AMD. “U are the best on it,” Slattery acknowledged once.