Black Edge

by Sheelah Kolhatkar

  Ross was flabbergasted. How could Martoma possibly know about those details? It was as if Martoma had seen the presentation he had just seen. But he knew that was impossible.

  “And what about the dose effect?” Martoma said. “How can you be so high on it when you look at the dose effect?”

  “I don’t care about the dose effect,” Ross said, sounding irritated. “I just know my patients.” He found himself wondering how he could have gotten mixed up with someone who was so indifferent to the human suffering behind Alzheimer’s disease. Didn’t Martoma care about all those elderly people, relegated to miserable deaths in hospital wards? They needed bapi to work. Ross said goodbye and walked back to his hotel room.

  —

  The hotel ballroom was filled the following afternoon for the formal presentation. All 1,700 seats were taken, with additional people standing in the aisles and along the back wall. It either was—or was not—going to be the setting of one of the great breakthroughs in medicine.

  Gilman was not feeling well. He’d had a series of chemotherapy treatments that week, and he was depleted. He approached the stage and adjusted his eyeglasses, grasping the podium with both hands. The room grew quiet. Then Gilman started talking about the first of his twenty-two slides. “Encouraging Clinical Data from Passive Immunotherapy in Alzheimer’s Disease,” read the title of his presentation. He described the objectives and the design of the Phase II study. Then he moved into the safety characteristics of the drug.

  As he cycled through six slides containing the efficacy analysis of the drug, showing a series of blue and green bar charts indicating that there wasn’t much of a relationship between patients’ cognitive improvement and strength of dose, the auditorium began crackling like a switchboard. Wall Street analysts were furiously typing on their BlackBerrys, sending messages back to their funds to sell Elan and Wyeth even though the stock market had already closed.

  “I can remember gasping,” said one analyst in the audience, who texted out his own Code Red.

  As soon as the presentation ended, Katie Lyndon, an SAC analyst who worked for Martoma, got up from her seat and rushed back to her hotel room. She hadn’t seen Martoma all day. Typically, when they both attended industry events like this, they kept separate schedules, to maximize the number of presentations they could watch, and then compared notes later. She was anxious to see what was going on with their portfolio and logged into Panorama, SAC’s internal system that tracked its positions. She was sure that Elan and Wyeth were going to get killed as soon as the market opened the next day. The only question was how much they would go down.

  As she stared at Panorama, she stopped breathing for a moment. Money was moving around the system in and out of the different portfolios faster than she’d ever seen before. She couldn’t tell what was going on.

  Shortly after midnight, Martoma sent her an email. “What was your impression of data?” he wrote. “Any buzz at dinner?”

  “I thought the data was okay, not as great as I was hoping, but probably also not meriting the euphoria with which it was greeted by the shorts,” she replied. She was dying to know what was going on with their position. “I saw what u did with WYE yesterday,” she continued, “and I’d be interested to hear your thoughts about that going into the data.”

  That same evening, back at SAC’s offices, Tim Jandovitz was sitting at his desk when the press release announcing the bapi results crawled across his Bloomberg monitor.

  “That does not look good,” he thought as he clicked on the headline.

  The market was closed, but the stocks had already dropped significantly in aftermarket trading. Jandovitz looked at Panorama, too, hoping that someone had reduced SAC’s position or at least put some hedges in place. But he didn’t see any Elan or Wyeth sales recorded. He was certain they’d lost at least $100 million. A disaster.

  At the same time, as the news was filtering out of the ICAD meeting, Ben Slate and David Munno started furiously emailing Jason Karp. “The drug failed,” they wrote.

  Karp felt ill.

  Slate and Munno knew better than to gloat. After all, as far as they knew, SAC had just lost a devastating amount of money, possibly enough to throw the future of the firm into doubt. Still, Munno couldn’t help but feel vindicated. Martoma had been wrong all along.

  Even though he had gotten into the habit of working out of the New York office a couple of days a week, Jandovitz made a point of reporting to SAC’s main office in Stamford the next morning. He was fairly certain that Martoma would be fired. There was a decent chance he would be fired, too. He couldn’t remember witnessing such a catastrophic loss; it was multiples greater than anything he’d ever been involved in. He briefly wondered what else he could do with his life, whether another firm would ever hire someone associated with a loss of this scale.

  When he got to his desk, he checked Panorama again. It reflected no positions in Elan and Wyeth at all, but the market hadn’t opened yet. Something wasn’t making sense. He finally reached Martoma on his cellphone. “What happened to our Elan?” Jandovitz asked.

  “We no longer own the stock,” Martoma told him curtly.

  Jandovitz felt a sense of relief. Maybe he still had a job after all.

  Just before the stocks opened for trading, Jandovitz got an instant message from a friend, a salesman at JPMorgan whom he’d known since they were five years old.

  “TELL ME MARTOMA GOT OUT OF ELAN,” his friend wrote.

  “W/out getting into detail,” Jandovitz wrote back, “Wednesday and this week has been GREAT for us.”

  “I LOVE IT…I LOVE IT…I LOVE IT…GREAT STUFF, MY MAN, GREAT STUFF,” his friend replied.

  “Stuff that legends are made of,” Jandovitz answered. He imagined that the decision to sell all of the firm’s Elan—and then short it before the disastrous trial announcement—would someday end up on a list of the most celebrated trades in Wall Street history. “We’ll catch up over a beer, and I’ll tell you a tad bit more.”

  Jandovitz was still bewildered as to how they’d avoided the Elan-Wyeth disaster. He recalled all the times Martoma had promoted the stocks, claiming a conviction rating of 9 in the idea. Something had happened to cause Martoma and Cohen to suddenly change their minds, and it had happened in secret.

  Elan shares had closed at just under $33 before the ICAD presentation. The scientists who had looked at the bapi results were divided about what they meant. Some felt that the drug still had potential, while others interpreted the data negatively. The results, Alzheimer’s researchers agreed, were confusing. Wall Street, however, was clear about how it felt: Elan opened at $21.74 that morning and fell below $10 within two days. Wyeth fared a little better. The massive losses investors suffered in the two stocks became the talk of Wall Street, as traders tried to figure out who might have been put out of business and who made millions taking the other sides of those trades. Anyone who had purchased shares in recent weeks and held on to them without the benefit of knowing the secret trial results lost money.

  —

  Martoma arrived for lunch with Gilman at the Peninsula Hotel’s Shanghai Terrace in Chicago the day the market was reacting to the news about bapi. It was considered to have the best Peking duck in the city, served by waitresses wearing cheongsams. “Did you hear about what happened to Elan’s stock?” Martoma asked Gilman once they’d sat down.

  Gilman was annoyed at the question. “No,” he said. “That’s not what I do.”

  “Well, it dropped like a rock,” Martoma told him. More than 30 percent, in fact.

  Gilman was surprised. He thought bapineuzumab still had promise and that it might come out of Phase III testing with some commercial potential. “Did I do something wrong in my presentation?”

  “The market doesn’t like a drug that only helps half the population,” Martoma said.

  A few days later, when Martoma was back in Stamford, he called Jandovitz into his office. Jandovitz was still fuming about not being included in t
he decision about the Elan trade.

  Martoma apologized for what had happened. “Steve told me not to tell you about our decision to sell,” Martoma said.

  “Why?” Jandovitz said. He was offended. He was the healthcare trader, after all, the one who had built the position up over months. Why was he suddenly excluded?

  “Steve told me to keep it between me and Villhauer,” Martoma said. After feeling confident all those months, Martoma had changed his mind about the drug trial, he added, and thank god for that. “I reviewed my notes over the last couple of weeks and just didn’t feel confident anymore.”

  Gilman slowly recovered from his lymphoma treatment over the course of the summer. He was still weak, but his hair was starting to grow back.

  He waited for weeks to hear from Martoma, but there was no word. He was surprised by it. He couldn’t believe his friend wouldn’t at least check in on him, given his cancer and everything else—in the past, Martoma always showed such concern for Gilman and his health. Finally, in September, Gilman broke down and typed out an email. “Hi Mat, I haven’t heard from you in a while and hope that all is well with you,” he wrote, trying to sound cheerful. “I hope that you have not been too seriously set back by the great turmoil in the markets, plus the disappointing drop in Elan stock.” As he watched the stock market plunging day after day and saw headlines about bank rescues and financial panic, Gilman had been worried about his friend and his young family. “Anyway,” he added, “no need to call, I have nothing new. I just wonder how you are faring. Best Regards, Sid.”

  Martoma scheduled an appointment through GLG to meet with Gilman a few months later at a medical conference in Seattle, but Martoma canceled the meeting. They never saw each other again.

  —

  There was more drama to come that summer. While Elan stock was still going down as a result of the drug trial results, the SAC technology analyst Jon Horvath got a piece of information that he was sure would make him and his boss, Mike Steinberg, a lot of money. Jesse Tortora, Horvath’s friend and the ringleader of their stock-tip-sharing circle, had learned that Dell’s earnings were going to be a disappointment. Tortora had a connection, a former colleague named Sandeep Goyal, who had a friend who worked at Dell. The hedge fund Tortora worked for, Diamondback Capital, was paying Goyal for his Dell tips by wiring $75,000 a year to Goyal’s wife because the firm Goyal worked for wouldn’t let him do outside consulting. Apparently Dell’s gross margins were looking like they would be lower than what people were expecting, in contrast to the relatively strong performance of the stock.*2

  Typically, Tortora sent this kind of information first to Todd Newman, his boss at Diamondback; he was like a cat, whose job it was to bring the dead mouse to the people paying his own bonus first. Then Tortora would forward it to Horvath, Sam Adondakis, their friend at Level Global, and the others in the email group. “Dell checks,” Tortora wrote them on August 5, three weeks before earnings were going to be publicly announced. “gm looking 17.5% vs street 18.3%. Doesn’t sound good, but still very prelim and could change.” This was where figuring out what others in the market were expecting became critical to betting the right way. Horvath figured that the 17.5 percent gross margin number, which gave a sense of the strength of the company’s sales numbers compared with its costs, was likely to send the stock down. Steinberg started shorting Dell in anticipation of the company’s earnings announcement at the end of the month.

  Horvath was desperate to prove his value to Steinberg and Cohen. He was still doing legitimate research into his computer companies, building spreadsheets and cultivating the investor relations staff and pressing them for information. But he was also feeding Steinberg the updates he was getting from Tortora. Typically, in the last month of each quarter, before Dell announced its earnings, Tortora would send Horvath a flurry of details about what was coming. He never specified where he was getting his information from, but it was clear that it was coming from someone fairly senior inside the company.

  On August 18, 2008, Horvath left for a long-planned trip to Cabo San Lucas, where he had rented a condo near the beach. While there, he tried his best to stay in touch about the Dell position. Tortora called him with another update that confirmed what he’d been saying for the last couple of months: The earnings were going to be more than disappointing, they were horrendous. “Please be extra sensitive with these numbers,” Tortora told him.

  Horvath immediately called Steinberg and passed along what Tortora had conveyed to him. After he hung up, he grew worried that he hadn’t been clear enough about keeping it quiet, so he followed up with an email. “Pls keep the Dell stuff especially on the down low,” he wrote. “Just mentioning because JT asked me specifically to be extra sensitive with the info.”

  By August 25, three days before the earnings release date, Steinberg had a short position on Dell of over $3 million. That day, Horvath received an email from Cohen’s “steveideas” account. “Cohen Sector Position Alert,” read the subject line. “Please reply with any comments or updates you have on the Cohen Account positions below.” The accompanying chart showed that Cohen’s personal SAC trading account was long Dell. Horvath experienced a clenching sensation in his stomach. He and Steinberg were betting Dell would go down, and Cohen was betting it would go up.

  “Steve didn’t like losing money,” Horvath said later—a mild understatement. “You were kind of in the bad books if you lost him money.”

  Horvath forwarded the email to Steinberg with the note, “steve is long DELL….”

  Steinberg, who was at his weekend house in the Hamptons, was feeling uneasy. What if Horvath was wrong? By this point, they were short almost $10 million worth of stock, which was a big position for him. Short positions were especially risky. If things didn’t go the way they were expecting, it was going to hurt. “Interesting…I have not mentioned anything to him yet,” Steinberg replied. “I would like to express our view to him, but we need to properly weigh the r.r. [risk/reward] of doing so. How high is your conviction here, scale of 1–10, 10 being maximal conviction?”

  They spent the next twenty-four hours debating how, and what, to tell Cohen. They emailed back and forth with another SAC portfolio manager named Gabe Plotkin, who had a $60 million long investment in Dell, to try to figure out how good his information was.

  It was clear that within the star system at SAC, Plotkin was on the rise. He had developed a formula for studying credit card data and shopping mall traffic that he claimed helped him make winning trades in consumer stocks. Whatever he was doing, it seemed to be working. He was making huge profits. Like Cohen, he could have millions at risk in the market and still go home and have the best night’s sleep of his life. And Cohen had been listening to him more and more. Plotkin was the one who had convinced him to buy Dell in the first place.

  The following day, August 26, Steinberg emailed Plotkin and Horvath, asking them to share their thoughts on Dell, as they appeared to have opposing views. “Gabe,” Steinberg wrote at 12:37 P.M., “we think GMs are at risk this qtr.” He went on to add, in stock market dialect, that he thought Dell investors were expecting the numbers to be better than they actually were. “Any thoughts on this—or related points?”

  “I do think that is the biggest risk,” Plotkin replied. He had a more optimistic opinion than Horvath evidently did, however. He listed the factors that he thought would influence the company’s gross margin number. “Where are you modeling gm%? What are your insights…?”

  Plotkin then called Cohen. They spoke for several minutes.

  Horvath had been reluctant to put a conviction number on Dell’s earnings. But at 1:09 P.M., he wrote out another email message from Mexico. He needed to indicate just how solid he felt his information was.

  “I have a 2nd hand read from someone at the company—this is 3rd quarter I have gotten this read from them and it has been very good in the last two quarters,” Horvath wrote. He added that his source was predicting that gross margins and earni
ngs would miss the estimates most Dell analysts were using. Then he pressed the send button.

  Plotkin took a few minutes to absorb what Horvath was saying. “Well—if your checks are right, that is certainly a negative,” Plotkin replied. “I will say however that it seems like recently (more in consumer) everytime someone hits me with a check, it ends up being off….So we will have to see.”

  Steinberg asked Plotkin to be careful with the information Horvath had just shared.

  “I will,” Plotkin answered.

  Plotkin viewed Horvath’s “check” with some skepticism. Still, he forwarded the “2nd hand read” email to Anthony Vaccarino, an SAC trader who had been instructed to keep Cohen informed about Dell. Plotkin then sold 300,000 shares from his portfolio, leaving him with a 2.1-million-share long position—still a large amount of exposure.

  Vaccarino’s official job was “research trader.” He was responsible for monitoring what all of the portfolio managers who traded retail and consumer stocks were doing and making sure Cohen knew about it. Privately, Cohen called Vaccarino his “conduit.” Cohen knew that his portfolio managers sometimes misled him, talking about how much they loved a stock and then selling it in their own portfolios, possibly so that his own, usually larger, volume of trading wouldn’t interfere with theirs. He had five research traders, and he instructed them to follow the trading activity inside SAC like hawks. Cohen wanted to know immediately if someone made a trade of even one share against whatever position he was holding. “Watch what they do, not what they say,” Cohen told Vaccarino on more than one occasion.

  Vaccarino immediately forwarded the “2nd hand read” email to Cohen, who was working out of his house in East Hampton. Then he placed a call to Cohen’s cellphone.

  Minutes after that phone call, Cohen started selling Dell. Over the next two hours, he sold his entire long position of 500,000 shares.