Infectious Greed

by Frank Partnoy

  The two most prominent credit-rating agencies were Moody’s Investors Services and Standard & Poor’s Corp., known as S&P. Moody’s had been in the business since 1909, when John Moody began publishing ratings of railroad bonds. S&P followed soon thereafter, and Fitch Investors Service—a distant third to Moody’s and S&P—began rating bonds in 1924.

  The agencies labored in obscurity until 1973, when regulators began tying legal rules to ratings. Because only Moody’s, S&P, and Fitch were approved for regulatory purposes, those three agencies had a monopoly lock on the market.18 Not surprisingly, it was a profitable business. Nearly every company with publicly traded bonds paid the rating agencies directly for the ratings. The typical cost of a rating ranged from $30,000 to $100,000. With thousands of companies needing ratings, business boomed. The legendary investor Warren Buffett invested in Moody’s and touted its franchise. By 2002, Moody’s would be worth more than Bear Stearns, a prominent investment bank.

  As commentator Thomas Friedman put it in 1996, “There are two superpowers in the world today in my opinion. There’s the United States and there’s Moody’s bond rating service. The United States can destroy you by dropping bombs, and Moody’s can destroy you by downgrading your bonds. And believe me, it’s not clear sometimes who’s more powerful.”19

  Why were credit ratings so important? Again, the explanation related to legal rules. Since 1973, regulators had passed dozens of laws that depended on credit ratings, ranging from AAA to D, with BBB being the all-important investment-grade dividing line. By the time Allen Wheat arrived at First Boston, just about every major financial business was limited in some way based on these ratings.

  In simple terms, a better rating meant better regulatory treatment. For example, certain mutual funds couldn’t buy bonds rated below investment grade. Insurance companies had to set aside more capital for lower-rated bonds. Federal loan guarantees required a high rating. And so on.

  During the 1980s, no one understood and profited from credit ratings more than Michael Milken and his investment bank, Drexel Burnham Lambert. When Milken was in business school (also at Wharton), he had noticed that bonds rated below the investment-grade level of BBB—now known as junk bonds—were undervalued relative to higher-rated bonds. The prices of bonds dropped sharply when the credit-rating agencies downgraded them to a level below BBB.

  Building on work by previous scholars,20 Milken concluded that by buying a portfolio of bonds rated below investment grade, an investor could increase returns without increasing risk. Even though a portfolio of sub-investment-grade bonds might include a few companies that defaulted on their debts, overall the cheap prices of those bonds more than made up for the risks. In other words, the sharp divide between investment-grade bonds and non-investment-grade bonds made no sense. To Milken, junk bonds were like $20 bills lying on the ground, and Drexel began trading them, earning huge profits both from holding sub-investment-grade bonds for its own account and from selling them to other investors who had become believers in “junk.”

  Unfortunately, Drexel became involved in various insider-trading schemes and other alleged violations of securities law, and its bond-trading business deteriorated. When Drexel filed for bankruptcy in February 1990—at exactly the same time Allen Wheat joined First Boston—a baton was passed. Drexel’s collapse made credit ratings even more important, and it left a gaping hole in the market for sub-investment-grade bonds.

  Like Bankers Trust, Drexel had been active in swaps. Now, banks—and their regulators—were worried about whether Drexel would default on payments owed on its swaps. Regulators and market participants began focusing more on the credit ratings of counterparties in the swap market. And the more they focused on credit ratings, the more important the ratings became.

  By combining CSFP’s AAA credit rating with the complex deals developed by Bankers Trust, Allen Wheat could create new products and sell them in previously untapped markets. CSFP designed the instruments, but they often were sold through First Boston, which had a more established and better-known sales force than Bankers Trust.

  Allen Wheat’s marketing pitches were transparently directed at Bankers Trust’s derivatives business. One glossy brochure stated, “We solve problems for clients rather than push products on them. In fact, Credit Suisse FP’s success can best be understood by focusing on this single underlying theme.”

  The pitches worked. CSFP earned numerous awards during this time, including “Swap House of the Year,” “Options House of the Year,” and “Best Derivatives House.” Wheat hired more people in 1991, and took over a floor of First Boston’s Park Avenue Plaza glass tower. CSFP developed new trades, which “solved problems” for clients, enabling them to make financial bets or hedge risks in increasingly sophisticated ways.

  There were two major innovations at First Boston during the early 1990s. One involved structured notes—essentially, highly rated bonds whose payments were linked to the same types of formulas Bankers Trust had used in its complex swaps with Gibson Greetings and Procter & Gamble (plus a few new formulas CSFP invented). The other innovation related to structured finance—a class of deals in which financial assets were repackaged to obtain higher credit ratings. Within a decade, structured finance would become a trillion-dollar industry, ranging from legitimate deals that enabled institutions to transfer risk more efficiently, to more dubious transactions (including those later involved in the collapse of Enron and the 2008 crisis).

  First, CSFP began issuing and selling structured notes. Unlike a typical bond, which has a series of standard coupon payments and then a fixed principal repayment, a structured note’s returns might vary wildly, based on different variables. In other words, the bond—or note—was structured so that its payouts were based on any conceivable financial instrument or index. One structured note was even linked to the number of victories by the Utah Jazz, a professional basketball team,21 although more typical variables were interest rates or currencies. As Richard Deitz, head of structured fixed income at CSFP, put it, “The types of structures we can come up with are infinite.”22

  For example, the first page of CSFP’s presentation to a new-associate training program in the early 1990s featured a structured note with a colossal 21 percent coupon and principal redemption leveraged twenty times. These numbers dwarfed those in the Gibson and P&G swaps, and the trades were even more complicated. Lest any new employees feel they could create pricing models for these trades on their own, the last page of CSFP’s presentation stressed that it was “necessary to exercise a great deal of care in modeling random yield curve movements” in valuing the trades. The message: don’t try this one on your own.

  Given Allen Wheat’s background in Asia, it was natural for CSFP to focus on structured notes linked to variables in Asian markets. For example, the bank sold huge numbers of what it called Thai Baht Basket-Linked Notes. The idea behind these notes was that the currency of Thailand, called the baht, had been fluctuating in a very narrow band; investors could profit by betting the baht would remain stable.

  Although the central bank of Thailand would not admit to buying and selling baht to keep its value within a particular range, the traders at CSFP had determined that a proxy basket of three other currencies closely tracked how the central bank of Thailand managed its own currency. Specifically, they said a basket of 84 percent U.S. dollars, 10.15 percent Japanese yen, and 5.85 percent Swiss francs had moved within a 3 percent range of the Thai baht’s actual movements since 1986.

  They then created a trade based on the proxy basket. Credit Suisse’s New York branch issued the notes, which paid an attractive fixed coupon of 11.25 percent, more than triple the rate the bank typically would pay.

  In return for the big coupon, investors took on the risk that Credit Suisse would not repay the full principal amount. Instead of a fixed principal repayment—as was typical for standard notes—the principal amount was linked to a formula based on the Thai baht; specifically, on the difference between the value
of CSFP’s proxy basket at the time the investor bought the bond and the value at maturity (typically, one year later).

  In other words, if the Thai baht went down relative to the proxy basket, the investor would lose money. If the baht stayed in its historical range, the investor would earn huge returns. According to CSFP’s research, this structured note had earned an average return of 15 percent since 1986, with the worst return being more than 9 percent. These returns were fantastic, given that most bond investors were earning less than half of that.

  Still, the notes were peculiar. Why would investors buy a structured note—and pay a substantial fee to CSFP—instead of just borrowing to invest in Thailand directly? After all, an investor buying a Thai-baht-linked note was making precisely the same bet as an investor simply borrowing money to invest in Thailand. (Either way, the investment did well if the Thai baht remained steady and lost money if the baht devalued.) Why the fancy, AAA-rated, proxy-basket package?

  The answer—again—related to legal rules and credit ratings. Just as Japanese insurers couldn’t buy stocks, many investors were prohibited by law from investing in Thailand, where even government bonds had a low credit rating. Thailand was a risky place, and U.S. regulators had adopted rules that prevented many money managers from investing there. According to these rules, pension funds (to give but one example) weren’t supposed to be speculating on the baht.

  But even if a fund manager couldn’t buy Thai-denominated investments, he could buy a note issued by a AAA-rated bank. What was wrong with a one-year note issued by Credit Suisse? It was a perfectly legal investment, and it would look safe to a regulator, a shareholder, or even a boss, who hopefully would not see the fine print describing how principal redemption was linked to the Thai-baht proxy basket. Thus, CSFP’s Thai Baht Basket-Linked Notes allowed regulated U.S. investors to play in Thailand, just as Bankers Trust’s equity derivatives had allowed Japanese insurers to play in their own markets.

  With trades like this one, Allen Wheat’s optimism about CSFP was justified. The fees were comparable to those Bankers Trust charged for complex swaps. Credit Suisse’s New York branch set up a borrowing program—called a medium-term note facility—to use in issuing these notes. Setting up the facility was like putting securities on the shelf. CSFP could then take down structured-note issues when it wanted, just as Bankers Trust had taken down its private placements.

  How optimistic was Wheat? The size of the medium-term note facility for the Thai-basket notes—the maximum dollar value of notes that could be issued—was $1.5 billion. As long as the Thai baht remained steady, both CSFP and its investors would get rich, and no one else would need to know exactly how they had done it.

  The idea of structured notes quietly spread throughout Wall Street during 1990. Twenty billion dollars’ worth were sold in 1991, $30 billion the next year, and $50 billion the next.23 By the end of 1993, every major bank sold structured notes. They were the rage among large institutional investors, and one of the hottest products among corporate and government treasurers.

  Who bought these notes? The better question was, who didn’t buy them? The list included major mutual funds, insurance companies, pension funds, and corporations. Many were less sophisticated than Gibson and P&G. State and local governments—even some school districts—bought the notes. One of the biggest buyers was the elderly treasurer of otherwise-conservative Orange County, California. The most common structured notes resembled the Bankers Trust swaps; essentially, they were hidden bets that interest rates would stay low.

  Investment banks soon found that—with their meager, single-A credit rating—they could not compete with AAA-rated CSFP for structured note issues, or for related swaps. To get a higher rating, they set up derivatives subsidiaries with their own capital. In a typical deal, a bank created a subsidiary and gave it several hundred million dollars of seed capital. Then the subsidiary agreed to strict limits on what it was permitted to do. A bank promised not to do business on behalf of the subsidiary, or to intervene in its transactions. With these protections, the credit-rating agencies agreed to give subsidiaries much higher ratings than their parent banks, typically the highest rating of AAA. As business shifted to these special-purpose subsidiaries, credit-rating agencies became even more important. One Merrill Lynch employee remarked, “We can’t blow our noses without getting approval from them.”24

  Other structured-note issuers included U.S. government-supported agencies and well-known corporations. (The issuers were the institutions that were using the notes to borrow money; the sellers were the banks that marketed the notes to their investors.) The Federal Home Loan Bank was the 700-pound gorilla, issuing tens of billions of dollars of structured notes.25 Two government-sponsored mortgage agencies—known as Fannie Mae and Freddie Mac—were big players, as were quasi-governmental entities, including the World Bank. For example, one of CSFP’s very first deals in New York was a $252 million structured note for Sallie Mae, the government-sponsored entity that makes student loans.26 By issuing structured notes, some of these agencies sometimes borrowed at lower rates than even the U.S. Treasury, which wouldn’t soil its name by participating directly in these markets.

  The most active corporate issuers of structured notes included new financial subsidiaries of many industrial companies, such as General Electric, IBM, and Toyota. DuPont Co. issued more than $1 billion of structured notes during the early 1990s.27 By issuing structured notes, these companies could save a half percent or more in interest costs. The companies didn’t really care if the note payments were linked to a complex formula involving interest rates or currencies—or some other wild bet—because the investment bank selling the structured notes always agreed to hedge the issuer’s risks with swaps, in the same way Bankers Trust had agreed to hedge the Canadian bankers’ risks in the Japanese markets. Companies often preferred raising money through structured notes instead of issuing stock or taking loans, both of which were typically more expensive.

  Structured notes quickly transformed these issuer companies. Even as early as 1991, 10 percent of GE Capital’s borrowing consisted of structured notes. By 1993, almost half of GE Capital’s medium-term-note program was in structured notes.28 GE Capital was especially savvy about credit ratings, adding a ratings trigger—stating that if a counterparty’s rating fell below a certain level, the swap was automatically unwound and settled—to many of its deals, beginning in April 1992.29 IBM and Toyota had similar approaches. The big three U.S. car makers also were involved, as were many insurance companies, especially American International Group.

  Shareholders of these companies were largely oblivious to these new instruments, because the securities laws did not require disclosure. Moreover, the notes were issued by financial subsidiaries, not by the parent company, and disclosure rules applied primarily to the parent. As a result, many of the details of the financial subsidiaries of major corporations—GE, IBM, and others—were kept secret.

  Likewise, the purchasers of structured notes—including many mutual funds, pension funds, and government-treasury departments—did not tell anyone about them. Public filings did not require descriptions, and fund managers certainly weren’t talking.

  The term structured note—ubiquitous in finance today—was not even mentioned in a major newspaper or business magazine during this period. The first listing in the general-news database of Lexis-Nexis, a computerized information service, was on April 27, 1992, in an article by Michael Liebowitz, of the specialty publication Investment Dealers’ Digest, noting in passing that First Boston had hired Tom Bruch to become the firm’s second structured-note trader. There were no substantive discussions of structured notes until well over a year later. The first reference to “structured note” in the New York Times was not until June 16, 1994; BusinessWeek on May 16, 1994; Fortune on November 29, 1993; and the Wall Street Journal on August 10, 1993. As a result, even a sophisticated, well-read investor following the markets during the early 1990s would not have heard of stru
ctured notes.

  Although structured notes were very profitable for the banks at first, just as complex swaps had been, inevitably the business became more competitive and margins declined. To maintain their profit margins, bankers needed to invent new versions of the notes, new structures that could not easily be copied, with margins that would last.

  CSFP liked to describe itself as solving problems for clients. However, a closer analysis reveals that CSFP was not so different from Bankers Trust in its approach to client relationships. Many of the “problems” CSFP was solving involved enabling clients to avoid regulation.

  Bankers Trust had demonstrated that regulation-avoiding trades were not only very profitable—they had staying power. And as long as clients couldn’t accurately value the trades and were using them to avoid legal rules, they would pay a premium. CSFP was about to take Bankers Trust’s approach to a new level of complexity.

  One of CSFP’s most innovative trades was called a Quanto, a turbocharged version of the structured notes the bank already had been marketing. A version of this trade was awarded “derivatives trade of the year” in a 1991 survey by Institutional Investor magazine. CSFP was widely credited with inventing the Quanto concept, which it first marketed to Japanese companies, although the concept grew from earlier deals Wheat did at Bankers Trust, and it resembled the swaps sold to Gibson Greetings and P&G.

  In a Quanto, the investor received payments based on foreign interest rates, with the unique twist that all payments were in the investor’s home currency. In other words, a U.S. investor could bet on European interest rates, but receive payments in U.S. dollars. In one typical deal, the Federal Home Loan Bank of Topeka, Kansas, issued $100 million of notes paying a coupon of twice U.S. LIBOR, minus British LIBOR plus 1.5 percent, all payable in U.S. dollars.30

  This twist—putting all payments in one currency—looked simple enough. But in reality, it was unimaginably complex.