One commentator described Mozer’s tax-avoidance trades as “dirty jobs.”29 In one example, Mozer created sham transactions to minimize Salomon’s taxable earnings. In these transactions, Salomon bought and sold bonds at artificial prices to generate what appeared to be losses—thereby concealing more than a hundred million dollars of taxable earnings—when in reality the trades had no economic effect. (By the late 1990s, there would be billions of dollars of such tax-avoidance trades—especially among energy and telecommunications companies such as Enron and Global Crossing—with no economic purpose other than avoiding legal rules, including taxes.)
Not surprisingly, traders who were consistently designing transactions to avoid legal rules—and who were paid millions of dollars for doing so—developed a culture of supremacy and disdain. Mozer and other traders began to disparage other firms in the same way salesmen from Bankers Trust had derided their clients. Salomon’s customers were important to the Arbitrage Group, because they provided precious information about supply and demand that traders needed to spot inefficiencies. But many of Meriwether’s traders began to take that information for granted, and to view their customers as fools. Paul Mozer was said to “turn on the charm on sales calls. But then he’d turn around and say something about what a moron the person was.”30
Although Mozer rankled some of his colleagues and their clients, he was a consistent source of trading profits. When Mozer’s predecessor on the government-bond desk was involved in a Treasury-bond auction, dozens of people would stand behind and watch. But when Mozer took over, anyone who tried to watch him—even Chairman John Gutfreund—was castigated with “get away from here” or “get out of here.”31 Salomon’s chief legal counsel, Donald Feuerstein, thought Mozer had an “attitudinal problem.”32 Nevertheless, Meriwether and Gutfreund strongly supported Mozer, and the fact that he was abrasive didn’t matter much. Mozer made $3 million in 1988, $4 million in 1989, and $4.75 million in 1990.33
When Mozer learned that Salomon had paid Hilibrand $23 million in 1990—five times Mozer’s compensation—something snapped. As one executive described it, “Mozer looked at Larry Hilibrand and thought, he got 23 million; I want to get 23.”34 In a bank full of bonus-obsessed traders, Mozer was unusually neurotic about his pay.
For his plan to “get 23,” Mozer returned to the on-the-run/off-the-run trade idea. Again, the original idea was that many clients—particularly foreign insurance companies—always wanted to hold the most liquid Treasury bonds. As a result, the newest bonds often were worth more than slightly older bonds. That was the arbitrage opportunity: a trader could buy the cheap bonds, sell the expensive ones, and wait for the prices to converge. In Mozer’s case, the bonds that fit into his new plan were 2-year notes, not 30-year bonds, but the idea was the same.
A group of investment banks had created a when-issued market, in which traders could buy and sell future rights to the newest 2-year notes. In this market, Mozer could agree today to sell notes that did not yet exist but would be created in a few weeks “when issued” by the Treasury in its next auction. The Treasury approved of this market, for the same reason regulators allowed price talk for new issues of stock: the early market minimized surprises and gave traders more confidence in bidding for the actual notes at auction.
Because companies were eager to buy new notes, prices tended to rise during the when-issued period, and then fall when the Treasury auctioned the notes (companies then would begin looking forward to the next auction and a fresh batch of notes). That predictable price decline presented an arbitrage opportunity: traders could sell in the when-issued market, and then buy a few days later at the Treasury auction. The major government-bond dealers all did this trade, selling billions of dollars of when-issued notes and buying at auction to cover their positions.
Mozer had a diabolical idea: if he could control the Treasury auction, he could squeeze the other dealers doing the arbitrage, who were counting on the new notes to satisfy their obligations to sell in the when-issued market, forcing them to pay a premium to buy the precious, recently auctioned bonds. Mozer wasn’t the first person to think of this manipulative scheme: some of the clients Mozer so disdained—including Japanese insurance companies—had attempted to squeeze the Treasury market several years earlier. But Salomon was the largest of the handful of dealers approved to participate in Treasury auctions, so Mozer would be in a much better position to execute a big squeeze.
Nevertheless, Mozer’s first attempts failed. In May 1990, he tried to corner a Treasury auction by boldly bidding for more than 100 percent of the notes. By submitting more bids than there were bonds for sale, Mozer improved the chances that he would receive most, if not all, of the bonds being auctioned, just as some Wall Street traders placed reservations to buy, say, twelve cars when they learned that a dealership was receiving a shipment of ten of the latest model. (When the Mazda Miata was introduced, traders held the rights to buy many of the new cars—rights they were able to sell at a profit, even if they had no interest in buying a Miata.)
The U.S. government frowned on traders submitting such large bids, because they didn’t want one investment bank to be able to control the price of recently auctioned bonds. Michael E. Basham, a deputy assistant secretary of the Treasury, called Mozer immediately after receiving Mozer’s bid and reminded him of a “gentleman’s agreement” between the Treasury and Wall Street that no firm would bid for more than 35 percent of a Treasury auction. The Treasury wanted the dealers to compete, so that the government could borrow at lower rates. Although the auctions were only loosely regulated, the Treasury did not want to encourage anti-competitive behavior by dealers. Basham said the Treasury wouldn’t discipline Salomon this time, but cautioned that Mozer should take care to comply with the informal 35 percent agreement in future auctions.
After this incident, Mozer’s bad attitude got worse. He lashed out at Basham, admonishing him not to get in the way. He warned that John Gutfreund, Salomon’s chairman, would call the secretary of the Treasury (Nicholas Brady, a close friend of Gutfreund’s) if Basham interfered again.35 The nerve of a “lowly” government official getting in the way of a top trader from Salomon!
Mozer ignored Basham’s warning, and submitted a bid for 240 percent of the next month’s auction. Basham called again, saying, “Paul, quite frankly the nature of the process is designed to encourage you to bid a higher price than the next guy, not to try to game the system. We would prefer that you not do it again.”36 Basham asked the Federal Reserve, which managed the auctions, to adjust Mozer’s bid down to 35 percent.
Frustrated, Mozer again ignored Basham and bid for 300 percent of the notes in the next auction. Again, Basham ensured that Salomon’s bid was adjusted down to 35 percent. A few weeks later, on July 10, 1990, the Treasury converted the 35 percent “gentleman’s agreement” into a formal rule.37 Basham had not consulted Wall Street regarding this new rule, and government-bond traders were surprised by it. At the next day’s auction, Mozer angrily entered eleven separate bids for 35 percent each, a total of 385 percent of the auction.38 He assailed the Treasury, over the phone and in the media, telling the New York Times that “the Treasury had been rash in making their decision without any prior consultation with the primary dealers. The Treasury has tied the hands of large dealers who time in and time out are the ones who underwrite the Treasury’s debt.”39
At this point, Mozer had to be reined in. Thomas Strauss, the number two person at Salomon (junior to Gutfreund, senior to Meriwether), insisted that Mozer apologize to the regulators, and then go to London for a few weeks to calm down.40
Mozer apparently spent the next months contemplating ways to circumvent the new 35 percent rule, just as he previously had helped Salomon dodge paying taxes. On February 21, 1991, Mozer submitted a bid in Salomon’s name for precisely 35 percent of the auction, dutifully in compliance with the new Treasury rules.
But he also submitted a bid for 35 percent of the auction on behalf of Mercury Asset Ma
nagement, a subsidiary of S. G. Warburg & Co., then the top securities firm in England, and a Salomon client. And he submitted yet another bid for 35 percent in the name of Quantum Fund, the hedge fund run by billionaire George Soros (where Andy Krieger had worked briefly). Warburg and Quantum hadn’t authorized the bids, but Mozer—who derided clients anyway—didn’t seem to care. They would never know about the bids, and Mozer would simply keep the extra notes for Salomon, effectively avoiding the Treasury’s 35 percent rule.
The strategy worked as planned, and Salomon controlled well over half of the notes sold at auction. After the auction, Mozer executed fictitious trades in which Warburg and Quantum “sold” their notes—more than a billion dollars’ worth—to Salomon at less than their market price.41 Salomon made a huge profit, and Mozer was on his way to catching Hilibrand and “getting 23.”
However, even the brainiest criminals frequently make one critical mistake. For Mozer, it was his failure to consider the possibility that a Salomon client might independently submit a bid in the auction. Mozer had entered bids for Mercury Asset Management and Quantum Fund at exactly the 35 percent limit, so even a small independent bid would put one of these clients over the limit, and in violation of the Treasury rule. There was no room for error.
Unfortunately for Mozer, Warburg—the parent company of Mercury Asset Management—entered a relatively small, $100 million bid of its own in the February 21, 1991, auction. This bid pushed Warburg over the 35 percent limit, and red lights began flashing at the Treasury. After several weeks of investigation, Treasury officials sent a letter questioning the bids to Warburg, with a copy to Mozer, who regulators knew had submitted the bid on behalf of Mercury Asset Management.
In an attempt to cover his tracks, Mozer immediately called a senior manager at Mercury Asset Management and tried to explain that the $3.15 billion bid he had submitted in Mercury’s name was merely a clerical error. Mozer asked, please, would he keep the matter confidential?
That same day, April 24, 1991, Mozer showed Meriwether the letter. Meriwether told Mozer his “behavior was career-threatening” and called to arrange a meeting with his own boss, Thomas Strauss, and Salomon’s general counsel, Donald Feuerstein.42 The three men met the next morning, discussed Mozer’s actions, read the Treasury letter, and decided they needed to speak to Salomon’s chairman, John Gutfreund, who was out of town. But they did not reprimand or discipline Mozer.
Meanwhile, that very day, Mozer was busy violating the 35 percent rule yet again, in the April auction. Mozer submitted a 35 percent bid for Salomon, along with a smaller authorized bid on behalf of Tudor Investment Corporation, a hedge fund run by Paul Tudor Jones. But Mozer couldn’t resist adding an extra $1 billion—unauthorized—to Tudor’s bid. This time, he was careful to leave some cushion, so that Tudor would not appear to have violated the 35 percent rule; in reality, Salomon was bidding the extra $1 billion, furtively putting itself over 35 percent. Mozer succeeded, and Salomon kept the additional $1 billion in notes, falsely purchased on behalf of Tudor, at a substantial profit.
A few days later, Gutfreund finally met with Meriwether, Strauss, and Feuerstein. Meriwether described the Warburg incident as an “aberration” and said he hoped it would not end Mozer’s career at Salomon. Even Feuerstein—who reportedly loved converting “chinks in the regulators’ armor”43 into advantages for the firm—thought Mozer had gone too far. He thought Mozer had committed a crime, and said Salomon should tell the government about it, although he noted that there probably was no legal duty to do so. They debated whether to notify the Treasury, or perhaps the New York Federal Reserve (Chairman E. Gerald Corrigan and Strauss were friends). They did not mention disciplining Mozer or limiting his activities.44
Perhaps nothing could have stopped Mozer at this point. He had submitted billions of dollars of false bids, including a $1 billion bid as a practical joke on a retiring Salomon salesperson from San Francisco.45 His reckless trading had not remained confidential, and when several hedge-fund managers learned about it, they hopped on Mozer’s wagon. Two large hedge funds—Caxton Group and Steinhardt Partners—had been involved in squeezing the April auction, too, and they were interested in participating, through Salomon, in the upcoming May auction. 46 Julian Robertson, the famed head of Tiger Management, another hedge fund, also wanted in on the action.
In all, Mozer bought 86 percent of the May auction. Mozer included Tiger in his bid for the May 22, 1991, auction, and added $500 million for Salomon—unauthorized, of course—to Tiger’s bid. Caxton and Steinhardt bought even more notes. This time, Mozer also bought 2-year notes in the when-issued market, and did not disclose those purchases, as Treasury rules required.47 With these secret purchases, Mozer was swinging for the fences, hoping he could sell all of these positions at a huge profit.
The big squeeze was on. The price of the 2-year notes skyrocketed, and the holders of the notes made $30 million.48 Mozer had hit a home run. Yes, he was out of control, but if he could just hold on until the end of the year, he would earn a huge bonus, perhaps even set a new record at Salomon. It seemed unlikely that regulators would catch him, and inconceivable that he would go to jail.
Arbitrage wasn’t the only profitable business at Salomon. The firm also was a pioneer in the market for complex mortgage bonds, financial instruments based on interest and principal payments made by homeowners. Lewis Ranieri of Salomon was a crusader for mortgage bonds in the same way Michael Milken was a crusader for junk bonds. They visited the same accounts and gave the same forceful sales pitches. Ranieri illustrated how homeowners’ payments of principal and interest flowed through a U.S. government agency into a specially created trust, and then into some lucky investor’s portfolio. Ranieri was just as colorful as vice chairman of Salomon as he had been as a mail clerk at the firm who was promoted to supervisor after he “got the brilliant idea one day to put a map of the U.S. on the wall and outline the postage zones in Magic Marker.”49 One of his favorite comments to investors was that “mortgages were so cheap your teeth hurt.”50
Mortgage bonds and junk bonds had much in common: they were undervalued relative to plain-vanilla bonds, they were more difficult to evaluate than most other investments, and there were hundreds of billions of dollars of them. Just as First Boston had securitized junk bonds in CBOs, Salomon would securitize mortgage bonds in similarly complex deals.
Mortgages had one key advantage over junk bonds: they were rated AAA by the major credit-rating agencies. The U.S. government felt that home mortgages were important and it subsidized them, not only by allowing taxpayers to deduct interest payments, but by implicitly backing the payments on mortgage bonds.
This AAA rating didn’t mean mortgage bonds were risk-free, of course. What the bonds lacked in credit risk, they more than made up for in interest-rate risk. Because homeowners could refinance their mortgages anytime they wanted, the owner of a mortgage bond never really knew if the bond would last for 30 years or 30 months. If interest rates declined and everyone refinanced, the purchaser of a long-term, 10 percent mortgage bond might have her principal back in a year, when available investments paid a much lower rate. This risk—known as prepayment risk—was very difficult to quantify, and was one reason mortgages were cheap.
Salomon stripped these mortgages into pieces in the same way First Boston had stripped junk bonds. Salomon created a trust (typically backed by a U.S. government agency), transferred a pool of mortgages into the trust, and then created a structure to separate the mortgages into different tranches. Mortgages could be split in various ways: interest and principal, short-term and long-term, West Coast and East Coast. They also could be separated based on prepayment risk, with a risky set of bonds immunizing others. For example, one tranche of bonds might suffer first from prepayments, and the second tranche would not be prepaid until all of the first tranche had been.
These strips of mortgages generally were known as Collateralized Mortgage Obligations, or CMOs, and the different varieties had fa
ntastically colorful acronyms, from interest-only bonds (IOs) and principal-only bonds (POs) to PACs, TACs, IOettes, VADMs, and Z-bonds. In most cases, the wilder the name, the riskier the bond. The riskiest versions were sometimes just called “nuclear waste.” Salomon became the biggest mortgage dealer in the world (although First Boston—the long-standing expert in securitization—had been the first firm to create a CMO). Salomon made money from mortgage bonds in three ways.
First, it earned a fee from setting up mortgage deals. The more complex the tranches, the higher the fee.
Second, Salomon made a commission selling mortgages to clients, especially those who wanted to take risks but were restricted by law to highly rated instruments. For example, regulated financial institutions, such as savings and loans, which were permitted by law to buy only highly rated instruments, could nevertheless buy CMOs, even though they were riskier than plain-vanilla bonds. Ranieri understood the importance of legal restrictions on investors, and he refused to accept state-imposed limits on mortgage investing. As one of his traders put it, “If Lewie didn’t like a law, he’d just have it changed.”51
(From a fund manager’s perspective, Salomon’s mortgage bonds resembled CSFP’s structured notes. They had very high credit ratings and their returns were potentially higher than any comparably rated investment. But even the more sophisticated fund managers didn’t understand mortgages much better than they understood structured notes, and they paid substantial fees to Salomon and other sellers as a consequence. Typically, fund managers did understand that mortgages had a much greater upside than other highly rated government bonds, and they stuffed their government-bond funds with risky mortgages—whatever the fee—hoping to outperform their peers; more on the effects of these risks in the next chapter.)
Infectious Greed Page 13
