Infectious Greed

by Frank Partnoy

  Again, Krieger bet against the dollar; again, the bet paid off. He also made money trading other currencies, including British pounds and German marks, in various complex strategies. Krieger described his sense of the market during this time as “preternatural.” He was extremely confident he could make money trading in the panic of the post-crash markets. All totaled, his options trades represented roughly $40 billion of underlying currency positions; these trades were fluctuating in value by as much as $20 million a day. Krieger referred to such changes as “just noise.” He felt $50 million was an amount he “easily could make up in trading.” In December 1987, Krieger finally stopped trading for the year, and left for the British Virgin Islands. At the time, his positions were so large that he wasn’t sure exactly what his profits for the year had been, but he believed he had made perhaps as much as $250 million. By any measure, it had been an incredible year.

  By late 1987, Bankers Trust had become a one-man show, and Krieger’s profits were carrying thousands of other employees. Without Andy Krieger, the bank was having a terrible year. Many of Bankers Trust’s businesses were losing money, and the bank had huge losses on loans to the Third World. The stock-market crash had hurt other parts of Bankers Trust’s business, and from the outside it looked as if Bankers Trust might have its first losing year in the bank’s fifty-year history. As the year closed, shareholders of Bankers Trust nervously awaited news of the bank’s earnings.

  On the morning of January 20, 1988, Bankers Trust sent out a press release reporting its preliminary, unaudited earnings for the prior year. Thanks to Andy Krieger, the bank had squeaked out a tiny $1.2 million profit. Shareholders cheered. Miraculously, the bank’s losses in several areas were offset by $593 million of income from currency trading, including $338 million during the fourth quarter of 1987.36 According to Bankers Trust, more than half of the total profits—about $300 million—was from one person: Krieger. In all, Bankers Trust reported more income from currency trading that year than any other bank in the world. It was a remarkable feat, and to many employees Krieger was a hero. Before he had arrived at Bankers Trust, the most any currency trader had made for the bank was $18 million.

  Shareholders were relieved, but some employees questioned how a 30-year-old trader could have made so much money in such competitive markets. They wondered how much risk he was taking. Stock analysts were skeptical, too. Was it really possible for one person to make $300 million in markets that most economists argued were efficient? Rumors were swirling that Krieger hadn’t really made so much money, and that someone at Bankers Trust had misvalued Krieger’s options.37

  But Bankers Trust’s management dismissed the rumors and began working on the most important task of the year: awarding bonuses. In January 1988, the prevailing worry among managers at Bankers Trust was not the riskiness or legitimacy of Krieger’s trading; instead, it was how much he should receive as a year-end bonus. Bankers typically spend much of January debating bonus-related issues, and it was difficult to do business with—or even speak to—a trader during this time. The only topic most traders will discuss in January is their “number.”

  A common trading commission at the time was in the range of five percent, so that a trader who generated $10 million in profits might receive a $500,000 bonus. In fact, five percent was precisely the promise Bankers Trust had made, according to Krieger. But five percent of Krieger’s profits would have been $15 million, and no Bankers Trust employee had ever received anything approaching such a huge bonus. Charlie Sanford wanted to reward risk-taking, but not excessive risk-taking. Paying Krieger $15 million might set a dangerous precedent.

  Moreover, Krieger’s boss, Jay Pomrenze, did not support awarding Krieger $15 million. Pomrenze and Krieger were close friends (Pomrenze even came to Krieger’s house for Krieger’s daughter’s naming), and Pomrenze said he thought Krieger was “the most powerful trader he’d ever seen.” But Pomrenze was reportedly “really, really nervous” about Krieger’s huge trading positions.38 There was no guarantee Krieger could repeat his trading profits. If anything, other traders had caught on to his strategies. Krieger’s profits might have been due to simple luck, and even if his trading profits were due entirely to skill, the markets would be more competitive next year. Why give Krieger a windfall this year if he wasn’t going to repeat his performance?

  There also was concern that Krieger was not a “team player.” When Krieger placed a billion-dollar trade, he moved the market. Some employees felt that Krieger did not keep them informed about these moves. They claimed they lost money when Krieger’s trades caused currencies they were trading to rise or fall. In reality, many of the traders also had been profiting from Krieger’s strategies, because they were involved indirectly in the trades. But their profits had been middling compared to Krieger’s.

  Even if Krieger had made $300 million for the bank, $15 million was simply too much money, given the circumstances. Management decided to pay Krieger $3 million, a very substantial bonus during a very poor year on Wall Street.

  The $3 million bonus was seven times the amount Bankers Trust had guaranteed Krieger the previous year, twenty times his bonus at Salomon three years earlier, and 60,000 times his bonus at O’Connor the year before that. It was even double the bonus of the bank’s chairman, Charlie Sanford. However you sliced it, Krieger would be paid big money in 1987, more than just about anyone working on Wall Street, and much more than he ever had received before. It was decided that Jay Pomrenze would tell Krieger his “number” in late January. Krieger was told, “You’re only 30 years old, you don’t need any more than this.” Krieger didn’t say a word in response.

  It is hard to imagine that the son of an accountant—a man who four years before had received a bonus check for $500—would be disappointed by a $3 million paycheck. But Wall Street traders are hard to imagine; besides, Krieger insisted he had been promised more. On February 23, 1998, he resigned, saying the bonus was too small relative to his profits for the year.39

  Krieger claimed Bankers Trust should have paid him a bonus in the range of $15 million. “I was very, very disappointed with the bonus on principle, rather than the actual amount. It was more money than I ever thought I’d need, but it wasn’t fair.”40 Krieger said he was tired of working 120-hour weeks. He had been ignoring his family and badly needed a vacation. He spent a few weeks “pairing off” his trades, so that Bankers Trust would not be exposed to any currency risk associated with his positions, which already were up more than $50 million for 1988. Then he left for the Caribbean.

  Currency-options traders were shocked, and trading in currency options nearly halted for a few hours after Krieger resigned. Again, rumors swirled about Krieger’s resignation, including word that Bankers Trust had incurred huge losses in currency options, perhaps as much as $100 million. A spokesman denied the rumors, but the denial only fed speculation about what had happened at Bankers Trust.

  Krieger experienced his first fifteen minutes of fame, complete with front-page coverage in the Wall Street Journal. But the markets quickly settled down, and traders soon forgot about him.

  Meanwhile, back at Bankers Trust, the nightmare was just beginning, with the startling revelation that makes Krieger the unwitting “patient zero” of the virus that has spread through the financial markets during the past two decades.

  Bankers Trust had issued its January 20, 1988, press release, announcing a profit for the year, a little too hastily. Bank managers typically receive end-of-the-year reports on the value of the bank’s trading positions, just as individual investors do for their own investments. Like most banks, Bankers Trust initially relied on its traders to evaluate the profits or losses in their own portfolios. Krieger simply used his own computer spreadsheet to do this, because Bankers Trust did not have any systems for traders to use. Every day, financial-control teams—sometimes called the back office—used their own spreadsheets to mark to market the value of the trades. At the end of the year, the back office marked to mark
et all of the traders’ positions, and these marks were the basis for reporting annual results.

  Unfortunately, Bankers Trust had announced its profit before the back office control teams had finished checking the values of Krieger’s trades. There was no reason to think the values Bankers Trust had assigned to Krieger’s trades would be inaccurate. But as the control teams continued their work, it appeared that Krieger had not made as much money as the managers initially had thought.

  In order to mark an options trade, the back office personnel typically input an estimate of what a particular trade was worth, based on computer models, publicly available information, and quotes from competing banks. The goal in marking to market these positions was to reflect as accurately as possible the market value of each trade.

  The key factor in determining the value of options—as the Black-Scholes model showed—was the volatility of the underlying currency. For example, if the New Zealand kiwi had been fluctuating greatly on a daily basis, kiwi options would be very valuable, because there was a good chance the right to buy or sell kiwis at a specified rate would be valuable. But if the kiwi had been stable, kiwi options wouldn’t be worth much.

  For options traded on an exchange, the value of the option was set by supply and demand, and was easy to determine. The value of such options was whatever people would pay for them, and those amounts were published every day. But Krieger’s options were traded in the over-the-counter markets, where there were no published prices and not as many buyers. Consequently, the value of these options was more open to question.

  No one had seriously questioned the values of Krieger’s trading positions when Bankers Trust issued its press release. But now the control teams were concluding that the volatility estimates used in marking Krieger’s trades were higher than they should have been. According to one report, the volatility measurements for Krieger’s portfolio were overstated by as much as 25 percent.41 That meant that the value of Krieger’s positions was overstated by—in aggregate—$80 million.42 In other words, Krieger had only made $220 million, not $300 million, as everyone at the bank originally had anticipated. Krieger was no longer as much of a hero. And Bankers Trust no longer had a profit for 1987.

  By late February, the managers at Bankers Trust were in a serious bind. Their January press release had included the $80 million of phantom profits. It had been inaccurate, and now they knew it. They were approaching the deadline for the bank’s 1987 annual report and Form 10-K, the annual filing public companies were required to make with the Securities and Exchange Commission. If they now admitted that the January press release had contained an $80 million mistake, shareholders almost certainly would sue, and the media would vilify them for their recklessness in issuing the press release. Even worse, without the $80 million, the bank’s tiny profits from 1987 would be wiped out, and Bankers Trust would record its first loss since the 1930s.

  But if they didn’t include the mistake in their Form 10-K, they would be knowingly committing securities fraud. In the late 1980s, federal securities prosecutors were aggressively pursuing criminal securities-fraud cases. And every banker knew the scene from Oliver Stone’s 1987 film Wall Street, in which federal agents lead Bud Fox from the floor in hand-cuffs. That scene was based on a real case, and no one at Bankers Trust wanted to be in the sequel.

  Now the rumors were really swirling, and Bankers Trust’s managers continued to deny them, saying they believed any impact on earnings related to Krieger’s departure would be “immaterial.” They admitted—in a private phone call to a group of stock analysts—that someone at the bank had mispriced several one-year currency-options contracts, based on Japanese yen, West German marks, and New Zealand kiwi.43 They pleaded with the analysts that such options were extremely difficult to value; it could happen to anyone. Shareholders were not privy to this call, and as far as they knew Bankers Trust was still profitable.

  Krieger said he was out of the country when the options were valued.44 He also said he rarely acted alone at Bankers Trust: “I sat next to the worldwide manager for forex [foreign exchange], Jay Pomrenze, and he was my boss. We had loss limits, and everything was worked out.”45 The most likely explanation for the mistake was that someone at Bankers Trust had tapped into the wrong data source for the volatility numbers used to value Krieger’s trades. Other traders at Bankers Trust had experienced similar problems, when the computer system occasionally applied volatility numbers from a simulation, instead of from a live data feed. As Krieger put it, “I don’t know exactly what happened, but they certainly didn’t have Victor Haghani from Salomon writing their systems.” Whatever the truth was, by the time Bankers Trust’s managers were trying to decide whether to disclose the fact that $80 million of income had “disappeared,” Andy Krieger had resigned and was no longer involved in the decisions, and no one at Bankers Trust was claiming responsibility.

  What could Bankers Trust do? The choices were bleak. They consulted their outside accountants, from Arthur Young & Co., now part of Ernst & Young. Arthur Young was conducting its annual audit of Bankers Trust and was preparing to issue its opinion that Bankers Trust’s financial statements were “fairly presented.” This opinion was a crucial part of the bank’s Form 10-K filing.

  The question was: how could the bank account for the missing $80 million? Was there a way Bankers Trust nevertheless could claim to have earned a $1.7 million profit, even though it now knew it had $80 million less in trading profits?

  Remember, this was 1988, more than a decade before dozens of companies—from Cendant to Enron to WorldCom—were accused of inflating revenues and reducing expenses to meet quarterly earnings targets, and two decades before the collapse of major financial institutions from subprime mortgage-related risks they hadn’t disclosed. Some senior managers back then were willing to whisper information about the upcoming quarter to the securities analysts covering their companies, but few would consider manipulating their accounting statements to fit expectations. That was fraud, after all, and reputable managers simply didn’t do it.

  Someone—it remains unclear who—suggested that Bankers Trust could resolve its problems and avoid the need to correct its January press release by simply reducing its compensation expenses by $80 million, to offset the loss exactly. Was it possible? Could the managers really argue that bonuses would be lower because of the losses, that the $80 million of phantom profits was offset by exactly $80 million less in bonus payments? That position would provide cover for Bankers Trust managers, who could argue they were not knowingly making any misstatements. Instead, they were relying on their accountants’ advice that it was reasonable to reduce the bank’s compensation expense.

  And they did it. The accountants at Arthur Young made an entry reducing by $80 million an account that recorded Bankers Trust’s obligation to pay employees’ future bonuses. The bank issued its Form 10-K, reporting the same profit it had announced in its January press release. Arthur Young issued a signed opinion that this profit was fairly presented. And the bank’s shareholders didn’t know a thing about any of it. Bankers Trust was gambling that no one would discover the $80 million reduction that exactly matched the $80 million of missing profits.

  The bank might have gotten away with this accounting sleight of hand if a few sharp-eyed banking regulators at the Federal Reserve hadn’t spotted the $80 million discrepancy in a banking call report Bankers Trust filed privately with the Fed. They forced the bank to file an amended call report. In their view, the accounting entry was dubious at best. Hidden accounts for future reserves were illegal, and the $80 million exact match was too much of a coincidence. Besides, Bankers Trust’s bonuses couldn’t be cut, because they had already been paid, and one person whose bonus arguably should have been reduced—Andy Krieger—was no longer an employee. An $80 million compensation reduction was absurd.

  With the cat out of the bag, Bankers Trust was forced to adjust its Form 10-K filing after all. The Securities and Exchange Commission—and the bank’s s
hareholders—were about to learn the truth.

  On July 20, 1988, six months after Krieger’s profits were first reported to investors—and five months after Krieger resigned—Bankers Trust announced $80 million of “adjustments” to its income from the previous year. The Bankers Trust press release stated that the company and its auditors believed the adjustments were “not material.”46 The managers put on the best spin they could. Chairman Charlie Sanford said, “What’s the big deal? There’s nothing sinister here. There’s no effect on earnings. The company is worth the same amount. It’s just an accounting thing.”47

  Securities analysts appeared to be furious, even though many of them had known about the discrepancy for months. They called the bank’s valuation process “loosey-goosey,” something “you could drive a truck through,” and threatened that the bank “is going to pay for this.”48 One analyst said, “I don’t want to invest in a casino.”49 Another summed up the sentiment: “Everybody in the world would like to see Bankers take a hiding.”50

  Other banks distanced themselves from Bankers Trust’s risky trades. A spokesman for Chase Manhattan said, “The majority of our business is customer driven. We don’t take the same large bets that other banks take.”51

  But, surprisingly, the Securities and Exchange Commission and the Department of Justice did not take any action. They didn’t bring charges against Bankers Trust, Andy Krieger, or any other current or former bank employee. It wasn’t an unwillingness to prosecute securities fraud. They were bringing dozens of insider-trading cases, and Michael Milken was about to go on trial. So why didn’t the federal prosecutors bring charges in this case?

  One reason was that the charges related to options valuation were extraordinarily complex and novel; most prosecutors had never heard of the Black-Scholes model. Another was that it was an election year, and it might not have been such a good idea to bring a high-profile case against a top investment bank and a top accounting firm when those two industries were also top campaign contributors. The federal government took the position that Bankers Trust and Arthur Young had corrected the mistake; that was enough.