Infectious Greed

by Frank Partnoy

  The losses at Kidder Peabody were a prime example of how legal rules did not prevent or deter financial manipulation. They did not create incentives for managers to control their employees at the outset, and they did not punish them for failure to supervise them after damage was done. As Kidder Peabody showed, the only remaining penalty was the destruction of the corporation, a penalty that was imposed more on shareholders than on the perpetrators who caused the loss.

  The $350 million of losses at Kidder were even more troubling than the hundreds of millions of dollars in unreconciled balances at Bankers Trust and Salomon Brothers, for two reasons. First, they involved much more straightforward financial instruments than Bankers Trust and Salomon had traded, instruments for which there was a liquid market and there were even daily quotes in the newspaper. Second, unlike Bankers Trust and Salomon Brothers, whose shareholders understood that those firms’ focus was on financial markets, Kidder Peabody was owned by General Electric, whose shareholders considered that firm to be primarily industrial in focus. GE shareholders never imagined the company was taking speculative bond positions in the range of $40 billion, twenty times the biggest trades Andy Krieger had made at Bankers Trust.

  The news about Kidder was a black mark on the record of John F. Welch Jr., the longtime chairman of General Electric. Welch had a reputation for building top businesses, slashing costs, and producing consistent profits. When he began running General Electric, Welch fired one in four employees, even as he spent $25 million on a new guest house and conference center at GE’s corporate headquarters.99 He became known as “Neutron Jack” for his practice of buying firms and leaving the buildings intact while eliminating all the people.100 His strategy was to be first or second in a business, or abandon it. He set firm quarterly and annual targets for his businesses, and managers met them—or else. The strategy seemed to work: for fifty-one straight quarters, GE’s earnings were higher than those of the previous year.

  Welch was also legendary for his brusque temper and management style; not surprisingly, he was livid about the losses at Kidder. General Electric had purchased Kidder in 1986, when the investment bank was suffering through a series of 1980s insider-trading scandals. (Martin Siegel, Kidder’s investment-banking superstar, did time in prison.)

  After GE bought Kidder, Welch provided special support for Kidder’s personnel, including Edward Cerullo, the head of bond trading. Welch was reportedly “intense” in his determination to fix Kidder, although analysts questioned his obsession with bond trading, and wondered why it was so “critical” or “vital” to a company like General Electric.101 But Welch knew of the profits at Bankers Trust, First Boston, and Salomon Brothers. Although shareholders still thought of General Electric for its lightbulbs, in reality GE under Welch already had become much more like an investment bank. It increasingly depended on GE Capital, the subsidiary that had issued billions of dollars of structured notes. By 1993, GE Capital was responsible for more than a third of GE’s earnings. Although Welch lacked experience in modern finance, he had great confidence in his employees, and he gave them free rein to meet his performance targets.

  At first, GE seemed to have turned Kidder around, cutting costs and even acquiring Drexel Burnham Lambert’s trading floor when that firm went under in 1990.102 Kidder’s employees were well paid and morale was high. Ed Cerullo stressed that Kidder would reward people based purely on their performance. Two of Cerullo’s top hires fit this performance focus: Melvin Mullin, a mathematics Ph.D. who built Kidder’s structured notes and options businesses into profitable lines that made $58 million in 1993,103 and Michael Vranos, a shy mathematics undergraduate who turned Kidder into one of the top mortgage-trading firms on Wall Street. By 1993, Kidder would succeed Salomon Brothers as the top mortgage-trading firm; it was the firm that sold Worth Bruntjen and David Askin many of their inverse IOs.

  Jack Welch didn’t meddle in the details of Kidder’s business, but he provided unwavering support, giving the firm more than a billion dollars of capital, and introducing Kidder bankers to GE’s top clients. Throughout the late 1980s and early 1990s, Jack Welch sent one simple message to Kidder employees: “stretch goals.”104 Welch wanted GE to be the dominant firm in all of its businesses. With Kidder, Welch could add a top Wall Street trading operation to GE’s diverse mix of businesses.

  Several experts questioned whether Welch was asking for trouble by combining a hands-off approach with a relentless focus on producing profits. Samuel Hayes, a professor at Harvard Business School, said Welch “has built a culture of individual fiefdoms, and that decentralized responsibility leaves a firm like GE vulnerable.”105 Edward Lawler, director of the Center for Effective Organizations at the University of Southern California, said, “Welch is intimidating, tough. GE’s culture is results oriented, and that’s the reason they do well and also break rules. It’s the opposite side of the same coin.”106

  Welch began learning about some of these broken rules in early 1994, even before the Fed raised rates. In January, Kidder fired Clifford Kaplan, a 28-year-old derivatives vice president who had made a laughingstock of the firm, not only by costing Kidder almost $2 million in cost overruns on a botched derivatives deal involving Italian government bonds, but also by showing that he could keep a job at Kidder while also being employed by the U.S. unit of La Compagnie Financière Edmond de Rothschild Banque, of Paris.107 Kidder had paid Kaplan a $500,000 bonus the previous year, even though he did not even have a securities license while he was marketing the Italian deal.

  Then Kidder discovered that one of its swaps traders had hidden $11 million in losses on a bond derivatives deal with NationsBank, and that one of its options traders had hidden $6 million in losses on French and Spanish government-bond options. It looked like no one was minding the store, and Jack Welch seemed concerned. He told Fortune magazine, in a March 7, 1994, interview about Kidder, “Things tend to grow to the sky, get momentum. ‘Let’s make it a little higher, a little higher.’ I think we’ve learned a lot about that.”

  Some employees blamed the problems on 46-year-old Melvin Mullin, the math Ph.D. who had developed Kidder’s derivatives businesses and then ran its government-bond trading desk. But in many instances, Mullin appeared to be simply following his marching orders from Jack Welch. When Clifford Kaplan said, “Mel was totally ‘hands off.’ Mel was purely driven by profit—profit, always profit,” it wasn’t clear if the comment was praise or criticism.108

  Was it a negative when Mullin permitted Kaplan to work on a derivatives deal without a proper securities license? Was it a negative when Mullin ignored warnings about a deal that violated Japanese banking laws? Was it a negative when he hired and supervised his wife, helped her determine her own bonus of $900,000, and then inadvertently double-hedged her options portfolio, costing Kidder $2 million?109 Okay, the last one was obviously a negative, but the others were arguably consistent with the culture Jack Welch was trying to create, and Mullin’s “hands-off” approach seemed to be generating profits.

  In 1991, Mullin took the step that would destroy his career, as well as the career of his boss, Edward Cerullo, and would bring down Kidder Peabody. He hired a young bond trader named Orlando Joseph Jett. When Melvin Mullin interviewed Joseph Jett, he said Jett “seemed like a hard worker with quantitative skills.” Kidder wasn’t bothered by the fact that First Boston had just fired Jett, or that he had lasted only a short time before that at Morgan Stanley, his first job after graduating from Harvard Business School. Jett began working at Kidder in July 1991.

  Mullin assigned Jett to trade long-maturity U.S. government bonds called “STRIPS.”110 STRIPS were zero-coupon Treasury bonds, meaning obligations of the U.S. government that repaid principal at maturity but did not pay any coupons in the interim. STRIPS were created from the Treasury bonds traders like Paul Mozer, at Salomon Brothers, bought through a federal-government program called “Separate Trading of Registered Interest and Principal of Securities” (hence, the acronym STRIPS).


  The “stripping” resembled the separating of interest and principal on mortgages, except that the Federal Reserve Bank of New York did all the hard work. As a trader at Kidder, Jett could simply buy a Treasury bond and present it to the Fed. In return, he would receive a collection of STRIPS: one small STRIPS for each semi-annual interest payment, and one big STRIPS representing the repayment of principal. Each one gave him the right to receive a single payment in the future, but no interim coupons. For example, if Jett presented a $1 million, 30-year bond that paid 10 percent interest on a semiannual basis, he would receive sixty coupon STRIPS representing the right to receive $50,000 every six months ($100,000 per year), and one principal STRIPS representing the right to receive $1 million in 30 years.

  Jett could also reverse the transaction. If he gathered up the sixty-one STRIPS necessary to “reconstitute” the Treasury bond, he could present all of those STRIPS to the Fed, and receive the actual Treasury bond in exchange.

  Because STRIPS represented rights to receive just a single payment in the future, they always traded for less than their face value (the amount the holder would receive at maturity). For example, a STRIPS representing the right to receive $1 million in 30 years might be worth $200,000 today. A STRIPS with a maturity of three years might be worth $800,000 today.

  Traders of STRIPS were very active, and the business was extremely competitive. In theory, a trader could buy a bunch of cheap STRIPS, reconstitute them, and then sell the new Treasury bond at a profit. (Or buy the bond, split it up, and sell the STRIPS at a profit.) In reality, such “arbitrage” opportunities were rare. These trades were easy to do and their costs were low. The Chicago economists who had predicted efficient markets could cite STRIPS as Exhibit A. There were no $20 bills lying around the STRIPS market.

  Jett was assigned to trade STRIPS of ten years or longer.111He had no training in STRIPS (he had been in mortgages at First Boston), and Mullin expected him to learn on the job. During the first few months, Jett’s results were discouraging, and Mullin was concerned. Then, on September 20, 1991, Jett had his eureka moment.

  Jett knew that STRIPS traded at a discount to their face value. In the above example, the 30-year STRIPS were worth $200,000; the 3-year STRIPS were worth $800,000. Over time, the value of STRIPS approached their face value. In other words, the day before maturity, the $1 million face value of STRIPS might be worth $999,900. This made sense: a person would pay a lot more to receive $1 million tomorrow than to receive it 30 years from now. STRIPS increased in value over time, just as money in the bank did.

  On September 20, Jett noticed that Kidder’s accounting system allowed him to record a STRIPS transaction in which the purchase and sale of bonds would occur, not right away, but instead at some future date. In other words, rather than buying STRIPS today, reconstituting them, and selling the resulting bond today, Jett could agree today to do the same transaction six months in the future.

  Why would he agree to do the reconstitution in the future, instead of today? This was the eureka. When Jett entered a future trade—a forward reconstitution—into Kidder’s accounting system, it produced an automatic profit. Why? Recall that a reconstitution involved a sale of STRIPS and a purchase of the corresponding bond. To make a profit, Jett needed to be able to buy cheap STRIPS. Kidder’s accounting system guaranteed that Jett could always buy cheap STRIPS, because it was misprogrammed to record the fact that STRIPS became more valuable over time, and to give the trader credit for that increase—not in the future—but today.

  In other words, Jett could buy STRIPS today for, say, $200,000, and agree to sell them in six months as part of a forward reconstitution. Kidder’s accounting system would record the sale price of the STRIPS as if time had moved forward by six months, when the STRIPS would be worth, say, $250,000. In this example, the system would record $50,000 as profit—the difference between the “forward price” and the current price.

  In reality, there was no profit at all. If Jett bought STRIPS today and committed today to sell them in six months, he would make nothing at all today. The value—in today’s terms—of the STRIPS he had bought was exactly equal to the value—again, in today’s terms—of the STRIPS he had agreed to sell in the future. Jett might make money over time, if he held on to those positions, depending on changes in interest rates. He might lose money, too. But he clearly would not make money up front. The accounting system was comparing forward values to today’s values, something that was as mathematically false as subtracting four from four and getting one.

  Jett did a few forward reconstitutions, and Kidder’s accounting system magically showed that he had made a profit. Suddenly, after four months of poor performance, Jett was making money. For November and December 1991, his forward reconstitutions showed a $265,000 gain. Not bad.

  Melvin Mullin was pleased with Jett’s turnaround. Another STRIPS trader complained to Mullin about Jett’s trading practices, and accused Jett of mismarking his positions. But Mullin brushed off these complaints. After all, the accounting system showed a profit. (Notwithstanding his mathematics Ph.D., Mullin later would claim he hadn’t understood Jett’s transactions.) A few months later, Jett ensured that the complaining trader would be fired, by delivering to Edward Cerullo a tape of a conversation in which the trader sought employment at another bank. Ironically, this trader ended up at First Boston, where he became a successful STRIPS trader in a group that made millions of dollars trading with Jett.

  Nevertheless, Mullin was cautious about Jett. In a year-end performance evaluation, Mullin gave Jett the next-to-lowest mark for “Overall Rating,” but the highest mark for “Performance Trend.” Mullin wrote that Jett had had a “lower start than anticipated” but “seemed to be improving.” The improvement wasn’t enough to justify any big money, though. Jett received a token bonus of $5,000.

  In 1992, Jett finally began to impress Mullin, and he began trading in forward reconstitutions at a frenetic pace. At the end of the year, Kidder’s accounting system recorded profits of $32 million for Jett, well above what anyone at Kidder had ever made trading STRIPS. Mullin was ecstatic, and recommended that Jett be promoted to senior vice president. In reality, Jett had lost $10 million. His profits were false, although apparently no one at Kidder knew it. Jett received a bonus of $2.1 million. He was a new man.

  A few months later, Edward Cerullo—the head of bond trading—told Mullin he wanted Mullin to run a new derivatives desk. Other banks—including Bankers Trust, First Boston, and Salomon Brothers—were making a fortune in derivatives, and Kidder wanted to be involved, too. The only question was who should replace Mullin? The two men decided on Jett, and in 1993—after less than two years at the firm—Jett became head of government-bond trading, supervising about twenty traders. Jett began reporting directly to Cerullo.

  Again, Jett didn’t disappoint, at least on paper. In 1993, he showed trading profits of nearly $151 million—more than a quarter of the profits in Kidder’s entire bond-trading operation. In reality, he had lost almost $100 million dollars, but Cerullo didn’t know it. Cerullo decided to pay Jett a record bonus of $9.3 million—more than triple what Andy Krieger had earned at Bankers Trust a few years earlier. If any managers at Kidder questioned how Jett, previously a total flop, had become the firm’s wunderkind, they didn’t voice their concerns. Instead, Jett was a hero and was named Kidder’s “Man of the Year.”112 At Kidder’s annual retreat, in Boca Raton, Florida, Jett gave an intense motivational speech, which Kidder’s general counsel described as “an emotional let’s-go-out-and-win kind of thing.”113

  When Jett began 1994, he was out of control. In the first two months of 1994—even as fund managers including Robert Citron, Worth Bruntjen, and David Askin were imploding—Jett set another record: $66 million of profits. Finally, his bosses began to ask some questions. How was Jett making so much money? When Cerullo looked at Jett’s trades and discovered more than $40 billion of forward reconstitutions—in some cases, more than the entire amount outsta
nding of a particular U.S. government bond—he was stunned.

  Cerullo’s first reaction was that he wanted to be sure those transactions wouldn’t appear on the firm’s financial statements. Even General Electric wasn’t a big-enough company to have one trader with so many assets and liabilities. An extra $40 billion of entries on GE’s balance sheet would jeopardize its AAA credit rating, and investors would bail out of the stock. As Kidder looked at ways of avoiding disclosure of these trades, no one focused on the issue of whether the profits were real. Everyone assumed they were real; they just wanted to hide the size of the bets. Jett apparently believed Kidder’s senior management supported his trading as a way of “window dressing” the firm’s balance sheet.114 The forward reconstitutions were classified as a type of over-the-counter derivatives transaction, and—remember—those were largely unregulated and did not need to be disclosed. For GE, they were safely “off balance sheet.”

  Finally, in late March 1994—just as Worth Bruntjen was struggling to evaluate his inverse IOs and David Askin was receiving margin calls—one of Cerullo’s deputies discovered the accounting glitch. Cerullo demanded that Jett explain his trading strategy in writing. The explanations made no sense, and Cerullo’s deputies calculated that Jett had lost about $350 million. Now, Cerullo had to tell Jack Welch.

  Welch was advised that Kidder had a “reconciliation problem,” but, ten days later, Kidder officials still could not figure out what had happened. The $350 million loss caused GE to report only $1.068 billion in profits for the first quarter of 1994, less than the $1.085 billion in 1993, the first time in fifty-two otherwise-perfect quarters that earnings were less than those of the previous year.115 At this point, it is possible that even Robert Citron, the Orange County treasurer, heard Jack Welch scream.