13 Bankers: The Wall Street Takeover and the Next Financial Meltdown (22 page)

In theory, then, structured finance could increase the pie for everyone. More important, in practice it was sure to increase revenues for the banks arranging these complex transactions. The ordinary trajectory for most products and services in the business world is for profit margins to decline as competition increases. For example, the amount that a bank could make from a plain-vanilla loan to a highly rated company was minimal, because many other banks would be willing to make that loan. Like all businesses, banks needed to invent new products that were not yet commoditized and that could command high margins. Structured products were the perfect answer. They were complex products that bank customers could not arrange on their own. Moreover, because selling these products required the ability to hedge risks in multiple markets, it was difficult for new banks to break into the business, which became dominated by a small number of players who could charge hefty fees for their services. In his memoir, former trader Frank Partnoy described how Morgan Stanley earned $75 million on a single trade.
³

In addition to pure derivatives such as interest rate swaps, currency swaps, and credit default swaps, asset-backed structured products were a mainstay of Wall Street derivatives desks in the early 1990s. The Repackaged Asset Vehicles that played a starring role for Morgan Stanley were structured products, in which a special-purpose vehicle (SPV, a new company that exists only on paper) bought a set of existing securities (say, bonds issued by the state electric utility of the Philippines) and paid for them by selling investors a new set of custom-designed securities.
⁴

Asset-backed structured products became Wall Street’s new cash cows, in the form of mortgage-backed securities (MBS) and their cousins, collateralized debt obligations (CDOs). The original mortgagebacked securities created by Ginnie Mae in the late 1960s were “pass-through” securities: mortgages were combined in a pool, and each security had an equal claim on the mortgage payments from that pool, spreading the risk evenly. Private MBS, however, are typically divided into different tranches, or classes, that have different levels of risk and pay different interest rates. Because the “senior” tranches have the first claim on all the mortgage payments, they have the least risk, and the credit rating agencies routinely stamped them with their AAA rating—the same rating given to U.S. government bonds. The “junior” tranches are riskier, but therefore pay higher interest rates to investors.
*

A CDO is similar, except that instead of being built out of whole mortgages it is built out of mortgage-backed securities or securities backed by other assets (such as credit card loans, auto loans, or student loans).
^†
By building CDOs out of junior, high-yielding MBS tranches, banks were able to engineer new securities that offered high returns with relatively little risk—at least according to their models. It was possible to combine low-rated MBS tranches, mix them together, and create a new CDO, 60 percent or even 80 percent of which was rated AAA; even though the MBS (the inputs) had low ratings, it was unlikely that many of them would default at the same time—at least according to the models. Financial engineers even created CDO-squareds—CDOs whose raw material was other CDOs—and higher-order variants, in order to squeeze out higher yield at lower supposed risk.
⁵

In the late 1990s, Wall Street became addicted to mortgage-backed securities and CDOs. As housing prices took off, it became easy to build models showing that MBS and CDOs had virtually no risk, because borrowers could always refinance their mortgages as long as prices were rising; even if they defaulted, rising prices meant that the investors would own valuable collateral. But because borrowers—especially subprime borrowers—were individually risky and paid interest rates to match, it was possible to manufacture AAA-rated securities that paid higher interest rates than other low-risk assets, such as U.S. Treasury bonds. Comforted by their AAA ratings, investors bought those CDOs without worrying about what was inside them. Most important, U.S. homeowners and homebuyers represented an enormous pool of potential borrowers that could be tapped over and over again as home prices rose and as they bought bigger houses or refinanced to turn their home equity into cash for home improvements or flat-screen televisions. Those mortgages and home equity loans were the raw material that Wall Street transformed into gleaming new CDOs for investors, taking a flat fee with each turn of the assembly line.

In comparison with MBS and CDOs, credit default swaps (insurance against default), introduced in
chapter 3
, are a relatively simple product, but they played a special role in the finance boom. Because the boom was based on creating, packaging, and selling debt, it depended on the assumption that borrowers would pay off their debts—or that someone else would pay in their place. Credit default swaps made it possible to insure
any
pool of mortgage loans or mortgage-backed securities, seemingly eliminating the risk of default.

In 1997, J.P. Morgan (part of today’s JPMorgan Chase) pioneered the use of credit default swaps to shift the default risk of loans off of its balance sheet. In the “BISTRO” transaction, J.P. Morgan’s derivatives team created a new special-purpose vehicle to insure loans the bank had made. J.P. Morgan paid insurance premiums to the SPV, and the premiums backed new bonds issued by the SPV to investors.
⁶
This complex piece of engineering had two benefits for the bank. First, because the risk of default on the underlying loans had been transferred from the bank to the SPV, the bank did not have to maintain capital reserves for those loans, so it could make more loans and hence more lending profits.

The other implication was more far-reaching. In effect, J.P. Morgan had created a new CDO out of thin air, without any of the raw material—loans or asset-backed securities—usually required. BISTRO was the first of what came to be known as “synthetic CDOs.” If the borrowers paid off their loans, the SPV would receive a steady stream of insurance premiums from J.P. Morgan to pay off its investors; but if the borrowers defaulted, the SPV would have to make a large cash insurance payout to J.P. Morgan, and its investors would lose their money. From an economic standpoint, it was as if the SPV actually held the underlying loans. This meant that a bank could create a CDO based on the housing market without having to buy a pool of mortgages or mortgage-backed securities; instead, it only needed to find someone who would buy insurance (using credit default swaps) on securities that already existed in the market. No one, in other words, had to go to the trouble of lending new money.

In the 2000s, as demand from investors and Wall Street banks for subprime loans outstripped supply, credit default swaps were used to fill the gap. As hedge fund manager Steve Eisman said, “They weren’t satisfied getting lots of unqualified borrowers to borrow money to buy a house they couldn’t afford. They were creating them out of whole cloth.”
⁷
This practice ultimately magnified the impact of mortgage defaults; as borrowers stopped paying, their defaults hurt not only the CDOs that held bits and pieces of their mortgages, but also the synthetic CDOs that mirrored them.

Credit default swaps also made possible another Wall Street business model. With a stable economy and rising housing prices, the default risk of the senior tranches of mortgage-backed securities and CDOs (the ones that got paid off first) seemed vanishingly small. Selling credit default swaps on these securities looked an awful lot like free money, and hedge funds stepped forward to take it. Notably, American International Group (AIG), the world’s largest insurance company, had a Financial Products group that was willing to insure AAA-rated structured securities for almost nothing. In the late 1990s, AIG agreed to insure the “super-senior” portion of J.P. Morgan’s CDOs—the part that was even safer than the AAA-rated bonds issued by the SPV
⁸
—for only two basis points (hundredths of a percentage point) per year. In other words, in return for insuring $100 million of loans against default, AIG would get $20,000 per year.
⁹
According to AIG’s models, it was free money. From J.P. Morgan’s perspective, because AIG was considered one of the world’s safest companies—it had a AAA rating of its own at the time—it was fully insured, for a cheap price. Everyone was happy. Of course, risk never disappears, and in this case it would reappear with a vengeance in September 2008.

The third ingredient of this money machine was a wave of innovation in mortgages, often described as (but not confined to) subprime lending. Traditionally, since the 1930s, home mortgages had been relatively conventional products and mortgage lending a relatively staid business. Most mortgages were long-term, fixed-rate, “prime” mortgages, where the borrower met the lender’s standards for creditworthiness, capacity (income) to repay debt, and collateral (real property sufficient to protect the lender in case of default).
¹⁰
Subprime loans, where the borrower did not meet one of these criteria,
*
were relatively rare. In 1993, there were only 24,000 subprime mortgages used to purchase homes and 80,000 subprime refinance mortgages. By contrast, there were 2.2 million prime home purchase mortgages and 5.2 million prime refinance mortgages; in aggregate, there were seventy prime mortgages for every subprime mortgage.
¹²

The 1990s and 2000s, however, saw an explosion in all types of mortgage lending, although most attention has focused on subprime lending (and “Alt-A,” a new designation for higher-end subprime loans), in which lenders lowered their standards for creditworthiness, capacity, or collateral, or all three at the same time. Before the market for private mortgage-backed securities took off in the 1990s, subprime lending was constrained by the fact that subprime lenders wanted to be paid back. Subprime loans had to conform to underwriting standards, like all loans, and were used primarily to refinance prime mortgages for borrowers who had poor credit histories but otherwise had the capacity to repay the debt. In addition, subprime loans were generally made by nonbank mortgage lenders who could not raise funds by taking deposits from customers. With the advent of securitization, however, investors and the investment banks that served them became particularly hungry for subprime loans because of the higher interest rates they paid, which were crucial to manufacturing high-yielding CDOs.

Now that loans could be resold to Wall Street, mortgage lending became a fee-driven business, where volume was the key to profits. Lenders responded by inventing new mortgage products that made it easier for borrowers to afford their monthly payments, at least for the first few years. These products went beyond the standard adjustable rate mortgage to extreme forms such as “pay option” mortgages where borrowers could choose to pay
less
than the monthly interest on the loan, causing the principal balance to go up instead of down. Lenders relaxed traditional underwriting practices, such as verifying the income and assets of the borrower; in stated-income mortgages, the lender explicitly
did not
confirm that the borrower had the income he or she claimed, and told the borrower as much. They accepted smaller down payments, resulting in higher loan-to-value (LTV) ratios, or used second mortgages to eliminate the down payment entirely; this meant that the collateral would not be sufficient to protect the lender from default unless housing prices rose.
¹³

Many of these innovations applied equally to prime and subprime loans, and business in both categories boomed. In 2005, 1.0 million subprime loans were used to buy houses and 1.2 million were used for refinancing—in aggregate, a twenty-fold increase over 1993.
¹⁴
The result—whether due to mortgage brokers who pushed borrowers into inappropriate loans, or due to house “flippers” who took on as much leverage as possible to buy as many houses as possible while the market was hot—was mortgages that many borrowers would have little chance of actually paying off out of their income.

But that no longer mattered—at least not to the lenders or the investment banks—because the lending business model detached itself from the requirement that borrowers pay back their loans. Lenders made fees for originating loans; the higher the interest rate, the higher the fees. Then, when interest rates reset and borrowers became unable to make their monthly payments, lenders could earn more fees by refinancing them into new, even-higher-rate mortgages. As long as housing prices continued to rise, a single borrower could be good for multiple loans, each time increasing his debt. (This business model had been pioneered by credit card issuers, who discovered that they could make money off of borrowers even if they never fully paid off their card balances.)
¹⁵
If he finally became unable to refinance, any loss would typically be taken by a CDO investor, not by the mortgage lender, and would be confined to the junior CDO tranches. As late as 2007, according to the International Monetary Fund, “Stress tests conducted by investment banks show that, even under scenarios of nationwide house price declines that are historically unprecedented, most investors with exposure to subprime mortgages through securitized structures will not face losses.” (The stress test cited by the IMF was conducted by Lehman Brothers.)
¹⁶