Black Edge

by Sheelah Kolhatkar

  Karp and his two analysts, David Munno and Benjamin Slate, had just had three hugely profitable years in a row and were regarded as the best healthcare team at the firm. None of them particularly liked Martoma. All three were startled to see the Wyeth shares appear in SAC’s account, and they started to ask around.

  “Why is the firm making a billion-dollar bet on Wyeth?” Karp asked Cohen.

  “It’s Wayne’s position,” Cohen told him. That was supposed to be the end of it.

  The name aroused awe on SAC’s trading floor. Wayne Holman was a former SAC portfolio manager, a graduate of Yale and the NYU School of Medicine who had worked as a pharmaceutical analyst at Merrill Lynch before coming to work at SAC. Cohen had tried to hire him multiple times before the compensation finally won him over; once at SAC, he quickly started to make more money than almost anyone else. Martoma referred to Holman as a “healthcare god.”

  When Holman left SAC to start his own hedge fund, called Ridgeback Capital Management, in 2006, Cohen contributed $800 million, on much more flexible terms than he typically gave to the new ventures of former traders. Cohen was so unhappy about the prospect of investing in healthcare stocks without Holman’s input that he asked Holman to continue giving him advice. Holman signed a consulting agreement codifying his willingness to talk to Cohen about Wyeth even though he no longer worked for SAC. In exchange, SAC agreed to pay Holman an “advisor’s fee” of 20 to 30 percent of SAC’s returns on the investment. Cohen didn’t usually offer deals like that, but Holman was worth it.

  Holman believed that the more rational way to make a bet on bapi was by investing in Wyeth, which was bigger and had more products than Elan, meaning that if bapi didn’t work out, it was less likely to suffer. Elan, on the other hand, had its future tied up in the Alzheimer’s trial. It had only one other viable drug in development, Tysabri, which was designed to treat multiple sclerosis. But Tysabri had proven problematic, causing brain infections in some patients. Holman predicted that if bapi didn’t work out, Elan shares would get crushed. It was, given all that, a nakedly risky investment. Even though Holman was his healthcare god, Cohen seemed to ignore him when it came to Elan and told everyone that Martoma was the one he trusted. Martoma was “tagged” for the position in Cohen’s account, which meant that he was considered responsible for the Elan investment and would share in the profits it generated. He was tagged for Wyeth, too.

  Munno and Slate couldn’t understand why Cohen was risking so much on two volatile drug companies. SAC made aggressive investments all the time, but Cohen prided himself on his risk management skills. Carefully assessing whether the money you might make on a trade offset the potential losses was crucial to a hedge fund’s survival. Making a bet that was so big that a single drug trial result could wipe out all of the fund’s profits for the year was not prudent. Munno and Slate approached Martoma several times, asking if he’d be open to discussing his views on the stocks and the bapi trial.

  “Let’s share notes,” Munno would say.

  “It’s okay,” Martoma would reply, infuriating Munno, who felt that SAC shouldn’t be invested in Elan at all.

  At first, Munno and Slate tried to figure out where Martoma’s confidence might be coming from, by trying to retrace his likely steps in the research process. They were certain, given their expertise in the field—Munno had a PhD in neuroscience—that they could replicate whatever work he’d done that led him to his conclusions. But over time they became convinced he was just wrong. The only other explanation was that there was something shady going on. They took their concerns to Karp.

  “Can you figure out what this guy is doing?” they asked.

  Karp started to get genuinely worried. Munno and Slate were the brightest healthcare analysts at the firm, in his opinion. If they thought that the position was a potentially catastrophic risk, he wanted to protect the company. His bonus was at stake, too, after all. When he approached Cohen to talk about Martoma, though, he got nowhere.

  “Just let him be,” Cohen said. “Don’t talk to him—he’s not your employee, just stop going down that hole.”

  “Do you think they know something, or do they have a very strong feeling?” Munno asked, referring to Martoma and Holman.

  “Tough one,” Cohen said. “I think Mat is closest to it.”

  A few days later, Slate asked Cohen the same thing again. “Seems like Mat has a lot of good relationships in this area,” Cohen said vaguely, waving him off. He accused Munno and Slate of “pissing on the parade” with their negativity. He was starting to get annoyed.

  With all the hyperaggressive hedge fund traders in the market, investing successfully in stocks had become harder than winning at poker. Not only did you have to know what cards the other players had, you had to know what they thought of their cards. Sometimes, companies announced positive news but their stocks went down, because traders were already expecting it. Guessing what everyone else was anticipating became as important as understanding a company’s products. Knowing this, Munno started trying to map out just how good bapi’s results would have to be in order to register as positive for all the investors betting on the drug trial.

  Over and over, Munno, Slate, and Karp nagged Cohen: Does Martoma think he knows what the results will be, or does he know? Cohen refused to answer.

  SAC was one of the most powerful hedge funds in the world, and its most talented employees had been reduced to acting like teenagers, bickering and scheming behind one another’s backs.

  Though Martoma kept up a hard front, he felt increasingly vulnerable. He was starting to feel like Munno was trying to sabotage his career by going behind his back and sowing doubts about him with Cohen. He tried to double-check some of the information he was getting from Gilman with another doctor working on the trial, whom he found through a different expert network firm, Wall Street Access. Then in April, Martoma begged Gilman, his prized contact, to talk with Munno and Slate and encourage them to calm down a little.

  The debate continued inside SAC’s office, but Cohen didn’t change his position. Finally he told Munno and Slate to back off for good. He didn’t want to hear any more from them on the subject of Elan.

  —

  As part of his role as an advisor to the more junior traders, Jason Karp developed a system for categorizing information that he taught to all of his analysts, a way to understand what was safe and what might be illegal.

  There was “white edge,” which was obvious, readily available information that anyone could find in a research report or a public document, information that wasn’t worth much, frankly, but wasn’t going to get anyone in trouble.

  Then there was “gray edge,” which was trickier. Any analyst doing his job well would come across this sort of information all the time. For example, an investor-relations person at a company might say something like, “Yeah, things are trending a little lower than we thought….” Was that material nonpublic information? The only way to be sure was to talk to SAC’s legal counsel, Peter Nussbaum, who’d been with Cohen since 2000. Nussbaum was not an imposing figure, but he had a drawing of a shark lurching out of the water on the wall outside his office, which some employees interpreted as a bit of art direction intended to make himself seem intimidating.

  If Nussbaum decided that a particular bit of information could get SAC in trouble, the stock in question was put on the restricted list and they couldn’t trade it. Of course, for that very reason, traders only went to him if they absolutely had to. Nussbaum was the firm’s equivalent of Internal Affairs.

  Karp’s third category of information was “black edge,” information that was obviously illegal. If traders came into possession of this sort of information, the stock should be restricted immediately—at least in theory. In the course of doing their work, analysts inevitably came across this type of information—a company’s specific earnings numbers before they were released, say, or knowledge that the company was about to get a big investment—although the vast majority o
f what the traders trafficked in was gray. Karp found the color-coded euphemisms to be helpful for the guys working for him; it enabled them to talk more openly about what they were doing. “If you do one thing wrong, you’re in jail and your life is ruined,” Karp told them. “There is no trade that’s ever worth it.”

  In reality, of course, they were all playing a game, trying to get the most valuable information they could without getting into trouble. Edge was the water, and they were swimming in it. And Cohen prided himself on hiring the most determined swimmers.

  In Karp’s view, Martoma was disregarding simple rules for assessing the quality and legality of his information. The thought crossed Karp’s mind that Martoma might also be bluffing. He’d seen it before, portfolio managers faking conviction to prompt Cohen to build a big position, essentially a gamble that would earn them a huge bonus if it happened to turn out well. Or maybe there was something illegal going on.

  Munno suspected something similar when he typed out another aggravated email to his colleagues, complaining about the fact that Cohen was listening to Martoma instead of them. “We’re idiots for getting involved in this to begin with,” he wrote to Slate. “Stupid on so many levels.” There was no way for them to win, Munno said, when Martoma was telling people that he had “black edge.”

  CHAPTER 7

  STUFF THAT LEGENDS ARE MADE OF

  As SAC’s traders struggled not to lose money through the spring, as the financial crisis deepened, the government’s insider trading investigation was finally beginning to gather momentum. In early 2008, Bear Stearns was sold to JPMorgan Chase in an emergency rescue negotiated by the government. From that point on, investors found themselves in an environment they had never seen before. Nothing they tried seemed to work. Even when traders were able to gather valuable information, it turned out to be useless for trading purposes. Companies would report earnings that were better than expected, and their stocks would plunge anyway. The market was headed in one direction, and that was down.

  Just as the market worsened, an application to wiretap Raj Rajaratnam’s cellphone was submitted in secret to a federal court, accompanied by an affidavit signed by B. J. Kang. Permission for the government to listen in on private phone conversations is not something courts give lightly. The application is first reviewed by the Justice Department in Washington, D.C., before being submitted to the court with jurisdiction over the area where the crime is alleged to be taking place. For the request to be granted, the FBI must show evidence that a crime is in progress over the specific phone line they are interested in tapping and that other investigative methods—such as reviewing documents and finding cooperators—have been exhausted or are not likely to work. They must show that a wiretap, known as a “Title III,” is the only way for the government to put a stop to the criminal activity.

  On March 7, 2008, after twelve months of digging by the SEC and the FBI, prosecutors at the Southern District of New York persuaded Judge Gerard Lynch to authorize a thirty-day tap on Rajaratnam. For almost the first time, the FBI was eavesdropping on Wall Street professionals doing their jobs.

  Cohen, meanwhile, was growing increasingly pessimistic about the economy, signaling to his employees that he saw a market crash coming. “Unless oil trades down dramatically, I don’t see how the market holds these levels,” he wrote in one firm-wide email. “I would use rallies to sell.”

  Strangely, this pessimism did not extend to SAC’s huge, unhedged positions in Elan and Wyeth, positions so large that they violated the firm’s risk management protocols. Rather, Cohen’s confidence in Martoma’s views on bapi was buttressed when Martoma arranged a private dinner at Cohen’s Greenwich mansion on June 4, 2008, with Elan’s CEO, a former Merrill Lynch investment banker named Kelly Martin. Most investors would never have the opportunity to have a private conversation with the chief executive of a multibillion-dollar company, let alone a dinner, but Cohen was different from other investors. He probably could have bought the company if he wanted to, and most chief executives would likely agree to give him some personal time if he asked for it.

  Martoma hoped that meeting Martin might give his boss additional confidence about their investment, especially in the face of David Munno and Ben Slate’s ongoing campaign to convince Cohen to sell. Elan had a history of worrisome accounting problems, and Cohen agreed that some face-to-face interaction would help him feel more comfortable about owning so much stock. He had always felt good about his intuition about people and wanted to get to know Martin a little, to get a sense of him as a person.

  Martoma and Cohen strategized ahead of time about the best way to elicit hints about how the drug trial was going. Martoma drew up a list of carefully crafted questions he planned to ask. The plan was to pay close attention to Martin’s body language in addition to what came out of his mouth, looking for nonverbal cues as to what was going on. “More recently, there has been increased emphasis on AAB-001 in your communications,” read the first point Martoma planned to raise. “How important is the success of AAB in this phase II trial towards establishing your lead/dominance in this space?”

  During the meal, Cohen watched Martin as he spoke. Martin struck him as subdued, almost downcast, not focused and energetic, the way you would expect someone leading a company that was about to solve one of the world’s great medical crises to behave. Cohen’s gut told him that things weren’t going to work out with bapi.

  After Martin left, Cohen turned to Martoma. “That doesn’t sound like a guy who just solved Alzheimer’s,” he said.

  —

  Two weeks later, there was positive news: On June 17, 2008, Elan and Wyeth announced “encouraging top-line results” from Phase II of the trial. It was only a preliminary indication of how things were going, but it was hopeful. Both companies publicly reaffirmed their decision to launch a Phase III study. Martoma was elated. He sent Cohen an email about it before the market opened.

  “Yee-haw,” Cohen wrote back. “Well done.”

  Munno, the most vocal Elan dissenter in the office, was frustrated. He still believed that Elan was a bigger risk than it was worth. When he expressed as much to Cohen, again, Cohen gloated.

  “Here we go again, round 2,” Cohen said. “And round 1 to Martoma.”

  When the market opened, both companies’ stock prices popped up a few dollars. Cohen and Martoma bought hundreds of thousands more shares of each. By the end of that month, CR Intrinsic owned $233 million worth of Elan and $80 million of Wyeth. Cohen also had a large position in his own portfolio, with another $400 million invested in the two stocks—altogether, more than $700 million.

  Jason Karp, Munno and Slate’s boss, was still annoyed about how much time Munno and Slate were wasting on their rivalry with Martoma, and he admonished them. “So now he looks smart and you guys look stupid,” he said. “And now I look stupid because you spent so much time trying to make an example out of him. Just let it go.”

  Martoma, meanwhile, kept buying more. As the position ballooned, Cohen’s personal assistant, who listened to nearly every conversation her boss had all day long via the “Steve-cam,” applied to senior management at SAC for permission to open a personal trading account. There were only two stocks she wanted to buy: Elan and Wyeth.

  —

  The two drug companies were scheduled to announce the final results of the bapi trial at the annual International Conference on Alzheimer’s Disease, on July 28, 2008, in Chicago. It was one of the most hotly anticipated medical research announcements in recent memory. By the standards of the science world, the conference was a glitzy affair, held over five days at the Hyatt McCormick Place Hotel. Hundreds of Wall Street analysts were signed up to attend, along with scientists and researchers from around the world.

  Three weeks before the conference, Elan notified Gilman that he had been selected to present the final data from the bapi study. It was a huge honor, the sort of opportunity that Gilman would normally have jumped at, but his health was failing. He had ju
st been diagnosed with lymphoma and was undergoing chemotherapy treatment. “The hair on my head is just about all gone,” he wrote despairingly to Elan’s medical director. “With my current appearance, I could fill in the role of an evil scientist in an Indiana Jones movie. I am looking for a diamond stud for my shirt so that I can look like Daddy Warbucks.”

  Executives at the drug companies assured him that he was the person they wanted to do it, bald or not. Was he well enough? Gilman said yes. Though his role was supposed to be secret, he immediately told Martoma. As the keynote presenter at the conference, Gilman would be one of the first people to have access to the full, unblinded bapi test results. Elan and Wyeth worked to ensure that there wouldn’t be any leaks.

  While Gilman was getting ready to present the drug trial results to the rest of the world, the stock market continued to crash. On July 15, the SEC issued an emergency ban on short selling of financial stocks in an attempt to calm the market, a desperate measure that only served to scare investors even more. Watching SAC’s portfolios lose money every day was an unusual experience for Cohen, and one he did not enjoy. He had lost all confidence that the situation could be brought under control through government intervention or anything else. He sent a company-wide email, warning of further financial violence. “I want to be clear that any rally that comes will be a 1–2 month affair followed by additional weakness,” he wrote. “Let me reiterate that the indices could see new lows over the next couple of weeks.”

  On July 15, Gilman flew by chartered plane to Elan’s medical facility in San Francisco, where he huddled with the teams who had overseen the Phase II trial. They were joined by a group of statisticians whose job it was to assist the doctors in interpreting the data. It was a defining moment in the quest to find a cure for Alzheimer’s disease. This was when the data would be “unblinded”—all of the parts that had been kept secret would be revealed for the first time in their entirety. Before they began, Dr. Allison Hulme, who led the bapi study for Elan, reminded Gilman again that everything he was about to see had to remain secret.