220 Week 4 F /For WIZARD KIM

Part I

a The Subprime Market Takes Off

The astonishing thing about the subprime crisis is that something so small wreaked so much havoc. Subprime loans started out as just a pocket of the U.S. home loan market, then mutated like a virus into a crisis of global proportions. Along the way, brokers, lenders, investment banks, rating agencies, and—for a time—investors made a lot of money while borrowers struggled to keep their homes. The lure of money made the various actors in the subprime food chain ever more brazen and, with each passing year, subprime crowded out safe, prime loans, putting more homeowners at risk of losing their homes and ultimately pushing the entire world economy to the edge of a cliff.

C o p y r i g h t 2 0 1 1 . O x f o r d U n i v e r s i t y P r e s s .

A l l r i g h t s r e s e r v e d . M a y n o t b e r e p r o d u c e d i n a n y f o r m w i t h o u t p e r m i s s i o n f r o m t h e p u b l i s h e r , e x c e p t f a i r u s e s p e r m i t t e d u n d e r U . S . o r a p p l i c a b l e c o p y r i g h t l a w .

EBSCO Publishing : eBook Collection (EBSCOhost) - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM AN: 349547 ; Engel, Kathleen C., McCoy, Patricia A..; The Subprime Virus : Reckless Credit, Regulatory Failure, and Next Steps Account: s7348467.main.ehost

This page intentionally left blank

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

2

a The Emergence of the Subprime Market

Abusive subprime lending burst into public consciousness in 2007, but its legacy dated back years. As early as the 1990s, consumer advocates were reporting pred- atory lending in lower-income neighborhoods. This early period was the fi rst iteration of subprime lending. Only later did subprime loans morph into products that ulti- mately brought down the fi nancial system.

FROM CREDIT RATIONING TO CREDIT GLUT

To trace the emergence of subprime lending, we have to begin with the home mort- gage market in the 1970s. Back then, mortgage lending was the sleepy province of community thrifts and banks. Banks took deposits and plowed them into fi xed-rate loans requiring down payments of 20 percent. Consumers wanting mortgage loans went to their local bank, where loan offi cers helped them fi ll out paper applications. The applications then went to the bank’s back offi ce for underwriting. Using pencils and adding machines, underwriters calculated loan-to-value and debt-to-income ratios to determine whether the applicants could afford the loans. In addition, under- writers drew on their knowledge of the community to assess whether the customers were “good folk” who would repay their loans.

Banks kept their loans in their portfolios and absorbed the loss if borrowers defaulted. Knowing that they bore the risk if loans went bad, lenders made conserva- tive lending decisions. They shied away from applicants with gaps in employment, late payments on bills, and anything less than solid reputations in the community. People of modest means could rarely obtain loans because their incomes were too low and they couldn’t afford the high down payments. For people of color, obtaining credit was even harder. Many lenders refused to serve African-American and Hispanic borrowers at all, even when they had high incomes and fl awless credit histories.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

16 • PART I THE SUBPRIME MARKET TAKES OFF

Deregulation Just as mortgage lending was conservative, so was regulation. Throughout most of the 1970s, federal and state governments imposed interest rate caps on home mortgages. Some states banned adjustable-rate mortgages (ARMs), loans with balloon pay- ments, and prepayment penalties, which are charges for refi nancing loans or paying them off early. These regulations had the effect of limiting or delaying opportunities for homeownership.1

The interest rate restrictions and bans on certain types of mortgages did not last forever. From 1972 to 1980, the average interest rate on thirty-year fi xed-rate mort- gages rose from 7.38 percent to 13.74 percent a year.2 These high rates hurt lend- ers and borrowers alike. Mortgage lending and real estate sales declined. In states where market interest rates exceeded the state’s interest rate cap, some lenders stopped fi nancing home mortgages altogether. To add insult to injury, depositors were fl ocking to withdraw their money from banks to invest in money market funds, which offered higher returns because they were not subject to interest rate caps on bank accounts. The outfl ow of deposits meant banks had less money to lend, further curtailing the availability of mortgage loans.

Eventually, as the banking industry faltered and real estate sales dried up, Congress took action to dismantle the regulatory apparatus. First, it passed a law in 1980 elim- inating interest rate caps on fi rst-lien home mortgages. Then, in 1982, it permitted loan products other than fi xed-rate, fully amortizing loans. Overnight new products sprung up, including ARMs, balloon payment loans, and reverse mortgages.3 Con- gress, in a sweeping move, also overrode state and local provisions that were inconsis- tent with the 1980 and 1982 laws.4

Deregulation addressed the immediate pressures facing banks. The abolition of interest rate caps allowed banks and thrifts to charge market rates of interest. At the same time, the proliferation of new loan products broadened the array of loans avail- able to borrowers. Borrowers who knew they would only be in their homes for a few years could opt for low-interest loans with a fi ve-year balloon to be paid when they sold their homes. Other borrowers were attracted to ARMs offering initial interest rates below the rates on fi xed-rate mortgages. Many of these borrowers planned to refi nance later if fi xed-rate loans dropped in price.

Deregulation was not all good news. Without the constraint of interest rate caps, lenders were free to charge exorbitant interest rates. They also had carte blanche to dream up an endless menu of exotic loan products that borrowers had no hope of understanding.

Technological Advances Starting in the 1980s, technological innovation also transformed the home mortgage market and paved the way for subprime lending. Lenders, in the past, had been extremely careful about borrowing decisions. They had erred on the side of caution because they did not know how to calculate the risk that borrowers would default. When underwriting loans, they had used rules of thumb to help ensure repayment, such as a total debt-to-income ratio of 36 percent, a 20 percent down payment, and three months of savings in the bank.

When the mainframe computer arrived on the scene, lenders could suddenly ana- lyze vast stores of data on borrowers and their credit histories. Statisticians began

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 17

using the power of computing to identify the factors that best predicted whether borrowers would make their mortgage payments. They used these factors to develop models for determining the risk that individual borrowers would default. The models were called automated underwriting and were dubbed AU. With AU, loan offi cers and brokers could take information from the loan applications of potential borrowers and run it through a computer program to determine the applicants’ default risk and their eligibility for loans.

AU dashed a number of hoary maxims about traditional loan underwriting. Out went requirements that borrowers make down payments of 20 percent and have savings equal to three months of expenses. Out, too, went an insistence on pristine credit records, low debt-to-income ratios, and full documentation of income. The old-fashioned under- writing rules and underwriters’ seat-of-the-pants judgment gave way to fancy statisti- cal models, giving lenders the confi dence to lend to borrowers with damaged credit or no credit history at all.

Equally important, AU made underwriting quick and cheap. In the “old days,” it took weeks to get a loan approved. With AU, lenders could shorten the underwriting period to seconds. New Century Financial, now a bankrupt lender that approved loans through a call center, advertised: “We’ll give you loan answers in just 12 seconds.” AU not only saved time. It also saved money. AU software reduced underwriting costs by an average of $916 per loan.5

The mortgage fi nance giants Fannie Mae and Freddie Mac set the gold standard for AU systems with their Desktop Underwriter and Loan Prospector programs for prime loans. Later, Fannie Mae designed a program called Custom DU, which was supposed to automate the underwriting of subprime loans. Other companies designed their own AU models for subprime mortgages.6

Although automated underwriting was a valuable innovation, it had downsides, especially when it came to subprime loans. One problem was that many models assumed that housing prices in the United States would go up indefi nitely, which was an unfounded and foolish assumption. AU systems also had a garbage in, garbage out problem. AU is only as good as the data that are entered. For example, if a broker entered false information, by infl ating borrowers’ income or the value of their property, the computerized assessment of the borrowers’ risk would come out wrong.

When it came to subprime loans, there was even greater reason to question the reliability of automated underwriting. AU was originally developed for the prime market, using decades of data on the performance of prime loans. There was scant evidence, however, that these models yielded accurate results for subprime loans because there was little historical data on subprime loans. Despite these problems, AU gave the appearance of reliable underwriting, which was enough to embolden the market.

Securitization Perhaps the biggest factor contributing to the subprime boom was the securitization of home mortgages. Securitization quietly entered the scene in the 1970s. The idea behind securitization is ingenious: bundle a lender’s loans, transfer them to a legally remote trust, repackage the monthly loan payments into bonds rated by rating agen- cies, back the bonds using the underlying mortgages as collateral, and sell the bonds to

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

18 • PART I THE SUBPRIME MARKET TAKES OFF

investors. It is a bit more complicated than this description suggests; we save the nitty- gritty of securitization for the next chapter.

The roots of securitization date back to the 1930s, when Congress established the Federal National Mortgage Association (Fannie Mae) as a federal agency to increase the money available for home mortgages. Initially, Fannie Mae purchased FHA- insured mortgages and in the process replenished the funds that lenders had on hand to make home mortgages.7 Thirty years later, Congress spun Fannie Mae off into a government-sponsored entity (GSE) and created a new GSE, the Federal Home Mortgage Corporation (Freddie Mac). Both securitized mortgages and eventually became private sector companies owned by shareholders. The government exempted the GSEs from state and local taxes. In exchange, Fannie and Freddie agreed to meet affordable housing goals set by the U.S. Department of Housing and Urban Develop- ment (HUD). This public mission meant that Fannie and Freddie had two masters to serve: their shareholders and the government.

The way that GSE securitizations work is that lenders originate mortgage loans that they sell to the GSEs. Only loans that meet Fannie’s and Freddie’s underwriting standards and that fall below a certain dollar threshold are accepted for securitization by the GSEs. In the industry, these loans are called “conforming loans.” Once they acquire the loans, the GSEs package them into pools. Those pools then issue bonds backed by the loans. As part of the bond covenants, Fannie and Freddie guarantee investors that they will receive their bond payments on time even if the borrowers default on their loans.8

Seeing the success of GSE securitization, investment banks and other fi nancial institutions wanted in on the game. Fannie and Freddie had captured most of the prime mortgage market, but had not yet tapped subprime mortgages for securitiza- tion. This set the wheels in motion for “private label” securitization of subprime loans. Private label is the term used for any securitization other than those orchestrated by one of the GSEs. Some private-label securitizations were done by lenders. For exam- ple, Countrywide Financial Home Loans, one of the largest subprime lenders, pack- aged and securitized the loans it originated. More often, however, subprime loans were securitized by Wall Street investment banks. By 2006, up to 80 percent of subprime mortgages were being securitized.9

Securitization revolutionized home mortgage fi nance by wedding Wall Street with Main Street. It tapped huge new pools of capital across the nation and abroad to fi nance home mortgages in the United States. Lenders, in a continuous cycle, could make loans, sell those loans for securitization, and then plow the sales proceeds into a new batch of loans, which in turn could be securitized.

Securitization also solved an age-old problem for banks. In the past, banks had held home mortgages until they were paid off, which meant they were fi nancing long-term mortgage loans with short-term demand deposits. This “lending long and borrowing short” destabilized banks. If interest rates rose, banks had to pay depositors rates that exceeded the interest rate borrowers were paying on older mortgages. And if interest rates dropped, borrowers would refi nance to less expensive loans. This “term mis- match” problem was a direct cause of the 1980s savings and loan crisis. Securitization solved that problem by allowing banks to move mortgages off their books in exchange for upfront cash.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 19

It was not only banks that benefi ted from the advent of securitization. All of a sud- den, thinly capitalized entrepreneurs could become nonbank mortgage lenders. They fi nanced their operations not with deposits, but by borrowing money to fund loans, which they paid back as they sold the loans for securitization. The new lenders oper- ated free from costly and time-consuming banking regulation and fl ew under the radar by making loans through brokers. Many had no physical presence in the communities where they operated and were anonymous unless borrowers read the fi ne print.

By the time everyone was toasting the millennium, subprime lending was poised to take off. Soon what had been a credit drought would become a glut of credit.

Macroeconomic and Public Policy Factors Macroeconomic forces also helped spawn the subprime boom. Ironically, two fi nancial busts helped clear the way for subprime lending’s phenomenal growth in the 2000s. One of those grew out of the Asian fl u. In July 1997, the Asian fi nancial crisis ignited in Thailand, driving down the value of assets and currencies throughout Southeast Asia. In a domino effect, the crisis reduced the demand for oil, which contributed to a fi nancial crisis in Russia the following year. After Russia defaulted on its debt, fearful investors began dumping both Asian and European bonds. The crisis spread to the United States when Long-Term Capital Management (LTCM), a highly leveraged hedge fund that made its money through arbitrage on bonds, lost money and experi- enced crippling redemptions. The Federal Reserve Board (the Fed) orchestrated a private bailout of LTCM of over $3.5 billion. With LTCM’s collapse, the bond mar- kets erupted in chaos, briefl y paralyzing private-label securitization and resulting in a liquidity crunch. Several subprime lenders found themselves unable to raise working capital, and ultimately their businesses failed.10

During the same period, the dot-com bubble was swelling. In 2000, the bubble burst and stock values plunged. By August 2001, the S&P 500 Index was off 26 per- cent from its former high. Then on September 11, 2001, terrorists attacked the United States. As the country grieved, the faltering economy attempted to revive, only to sus- tain another body blow in December 2001 when Enron fi led for bankruptcy. As one corporate scandal after another came to light, confi dence in the stock market crum- bled. The S&P 500 dropped another 15 percent and the country slid into a recession.

Throughout it all, the housing and credit markets were a beacon of hope for the economy. Alan Greenspan, the chairman of the Federal Reserve Board, seized on mortgage loans and other consumer credit as the way out of the slump. In mid-2000, the Fed exercised its “Greenspan put” and slashed interest rates, causing housing prices to grow at a steady clip of 10 percent a year nationally. After the 9/11 attacks, with the recession in full swing, the Fed ordered further rate cuts in order to jump-start the economy. Between August 2001 and January 2003, the Fed chopped the discount rate from 3 percent to 0.75 percent. This series of cuts drove down interest rates on prime loans. The cuts also made it possible for subprime lenders to borrow money at low rates, charge high rates to borrowers who couldn’t qualify for prime loans, and make money on the spread when they sold the loans.11

Low interest rates answered President Bush’s post-9/11 call for Americans to go shopping. Suddenly spending money became patriotic, and many consumers fi nanced their purchases with credit cards that charged exorbitant interest and late fees. Too

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

20 • PART I THE SUBPRIME MARKET TAKES OFF

often, families converted their credit card debt into mortgage debt or refi nanced their homes to pull out equity. As Greenspan noted, “Consumer spending carried the econ- omy through the post-9/11 malaise, and what carried consumer spending was hous- ing.”12 Programs advertising “Your Home Pays You Cash” urged people to borrow against their homes. Companies also promoted the idea that credit was the way to live the good life. Citibank spent $1 billion on a “live richly” campaign designed to lure people into home equity loans. PNC ads for second mortgages showed a wheelbarrow with the slogan, “The easiest way to haul money out of your house.”13

The constant message was that people should feel good about using credit. Debt, which used to be considered embarrassing and a sign of poor discipline, had stopped being shameful. As a sign of this cultural shift, between 2001 and 2007, overall house- hold debt grew from $7.2 trillion to $13.6 trillion, a 10 percent increase each year.14

The Fed under Greenspan not only kept interest rates low, but also refused to intervene to protect consumers despite growing evidence of abusive mortgages. Like- wise, Congress and federal regulatory agencies were unmoved by stories of defrauded consumers. The dominant ideology was that if there were problems with mortgage lending, the market would solve them. In addition, if consumers were taking on credit they couldn’t afford, that was their choice and their problem. The market’s job was to offer consumers choices, and consumers’ job was to take personal responsibility for the choices they made. On the corporate side, responsibility meant maximizing the bottom line for the benefi t of shareholders, without regard for the consequences of abusive lending to consumers or society.

These dynamics coincided with a huge federal push for homeownership. This push began in the mid-1990s under President Bill Clinton, when HUD coordi-

FIGURE 2.1. U.S. President George W. Bush makes remarks on home ownership at the Department of Housing and Urban Development. (Luke Frazza/ AFP/ Getty Images).

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 21

nated a public-private partnership designed to increase homeownership.15 When President George W. Bush came into offi ce in 2001, he went further, advocating that everyone should own a home as part of his vaunted “Ownership Society” ini- tiative. In response, HUD increased its pressure on Fannie Mae and Freddie Mac to fi nance an ever greater number of mortgages to people with modest incomes and to borrowers of color. The Bush administration embraced subprime loans as the key to growth in homeownership. By 2004, even the chief counsel of the Offi ce of the Comptroller of the Currency, Julie Williams, was lauding “the rise of the subprime segment . . . in advancing homeownership, especially for minority Americans.”16

Ultimately, the forces of technology, fi nancial engineering, and public policy converged to fuel the growth of the subprime market. Starting in 2000 the subprime market grew exponentially, capturing 36 percent of the mortgage market at its height in 2006, up from 12 percent in 2000, before crashing and infecting the world economy.17

PREDATORY LENDING

The fi rst iteration of subprime lending—coined predatory lending—began in the 1990s and was targeted at people who historically had been unable to get loans. Some had blemishes on their credit or limited credit histories that made them ineligible for prime credit with its stiff underwriting standards. Others were eligible for prime loans, but did not know how to go about applying for credit or, because of past discrimina- tion, mistrusted banks. These people were ready prey for a new class of brokers and lenders, who targeted unsophisticated borrowers.

In these early days, mortgage brokers were small-time operators, soliciting bor- rowers over the phone or door-to-door like Fuller Brush salesmen of yore, armed with a menu of loan products from various lenders. Lenders back then were often small fi nance companies that generated money for loans through warehouse lines of credit. Some lenders worked solely with brokers, but many had storefronts where they took applications directly. One of the early entrants was Citigroup, which bought the Baltimore subprime lender Commercial Credit and later renamed it CitiFinancial, CitiFi for short.

Finding potential borrowers and getting them to commit to loans was the key to success. Existing homeowners were the most frequent targets because they had equity and were easy to identify through property records, unlike prospective homeowners. Brokers and lenders perfected marketing strategies to fi nd naïve homeowners and dupe them into subprime loans. Some hired “cold callers” who would contact home- owners to see if they were interested in a new mortgage. The cold callers got paid a few hundred dollars for each successful call. Brokers and lenders also used municipal records to identify prospects. They scoured fi les at city offi ces to fi nd homes with outstanding housing code violations, betting that the homeowners needed cash to make repairs. They read local obituaries to identify older women who had recently lost their husbands, surmising that widows were fi nancially gullible. They also identifi ed potential borrowers through consumer sales transactions. For example, in Virginia, Bennie Roberts, who could neither read nor write, bought a side of beef and over 100 pounds of other meat from a roadside stand on credit from the notorious subprime

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

22 • PART I THE SUBPRIME MARKET TAKES OFF

lender Associates First Capital. In talking with Mr. Roberts to arrange the consumer loan, the loan offi cer from Associates learned that Mr. Roberts had no mortgage on his home. He soon convinced his new client to take out a loan using the client’s home equity. Associates refi nanced that mortgage ten times in four years. The principal after the refi nancings was $45,000 of which $19,000 was paid to Associates in fees.18

High fees were not the only thing that typifi ed predatory loans. Interest rates, too, could be astronomical. In 2000, the Baltimore City Paper told the story of the Pulleys, who had overextended themselves with credit card debt. In 1997, the Pulleys “were bar- raged with letters and calls from mortgage lenders offering to consolidate [their] exist- ing mortgage . . . and all their other debts into a new loan,” which would supposedly save them $500 per month. “Needing the cash and not well-versed in such dealings,” the Pulleys made a deal with Monument Mortgage for an adjustable rate mortgage loan with an annual interest rate that increased every six months up to 19.99 percent.19

Some brokers and lenders had understandings with real estate agents and home improvement contractors to refer homeowners to them for loans. This network also worked in reverse, when mortgage brokers received kickbacks for suggesting contrac- tors to borrowers who were seeking loans for home repairs. These referrals generated good money for everyone except the borrowers, who ultimately paid for the referrals out of the loan proceeds or through up-front fees.

Shady contractors who helped homeowners fi nance repairs were rife. In Cleveland, Ruby Rogers had a mortgage-free home she had inherited from her uncle. Citywide Builders, a contractor, helped her obtain a loan through Ameriquest Mortgage to update the home. Over six months, the contractor arranged repeated refi nancings of Ms. Rogers’ loan until the principal hit $23,000. Of that amount, Ms. Rogers only saw $4,500. Meanwhile, Citywide Builders walked off the job after doing $3,200 of work on the house. Ms. Rogers was left with a leaking roof, peeling tiles, warped wall paneling, and a hole in the wall. After Citywide Builders went bankrupt, Ameriquest sued Ms. Rogers for foreclosure.20

Brokers and lenders also targeted black and Latino neighborhoods, where they knew credit had been scarce and demand for loans was high. As electronic databases of consumers became more sophisticated, lenders could “prescreen for vulnerability,” picking out people they could most easily dupe.21 Loan offi cers at one lender report- edly referred to neighborhoods with a high percentage of borrowers of color as “never- never land.’”22

For homeowners, the arrival of brokers and lenders offering them credit seemed like manna from heaven. Some lenders even invoked heaven in luring borrowers. Gospel radio station Heaven 600 AM aired advertisements for refi nance loans through Promised Land Financial. To help brokers win customers’ confi dence, First Alliance Mortgage Company, nicknamed FAMCO, had brokers watch movies to help them understand borrowers’ points of view. They were instructed to watch Boyz N the Hood to experience inner-city life and Stand and Deliver to get a feel for His- panic borrowers.23

Loan offi cers and brokers were trained to make customers feel that they were act- ing in their best interests, even going so far as to provide attorneys to “represent” par- ticularly leery borrowers.24 They were told to “establish a common bond . . . to make the customer lower his guard.” Suggested common bonds included family, jobs, and

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 23

pets.25 Clueless that they were being targeted, residents welcomed the salespeople who befriended them into their homes. There the pitchmen would ply them with offers of loans to fi x a sagging porch, pay for a child’s education, or buy a car. One borrower, whose loan was fl ipped multiple times, said, “Everyone was just so buttery and nice.”26

Some brokers were people that borrowers knew through work or church. For many people, especially those who had been victims of redlining in the past, working with someone familiar or recommended felt safer than going to a bank. Often that was a mistake. The head deacon of the Message of Peace Church in South San Francisco allegedly used his position to exploit Brazilian immigrants who were parishioners at his church, by encouraging them to fi nance their home purchases through him. After they placed their trust in the deacon, he completed their loan applications, reportedly falsifi ed documents, and agreed to terms on their behalf. One borrower said that when she uncovered what the deacon had done, he threatened to report her to the Immigra- tion and Naturalization Service for overstaying her visa and then tried to bribe her with $5,000 to keep quiet.27

Much early predatory lending involved extracting equity from people’s homes. Lenders or brokers would convince homeowners to take out high-cost loans that the salespeople knew would eventually become unaffordable. The loans might contain balloon payments coming due in a few years or adjustable rates that would only go up. Just when borrowers were on the brink of defaulting, the brokers or loan offi cers reappeared on their doorsteps, ready to refi nance the borrowers into new loans. Some went so far as to adopt systems for tracking the amount of equity borrowers had in their homes. Each “loan fl ip” resulted in more fees for the brokers and lenders, which they tacked onto the principal. With each fl ip, the borrowers’ equity shrank and their monthly payments went up, until their equity disappeared and they could no longer qualify for loans.28

By design, these subprime loans were unaffordable. The easiest loans to fl ip were those that borrowers couldn’t afford in the fi rst place. The higher the interest rate, the bigger the monthly payment and the more likely the borrower would default. Reports abounded of subprime mortgages with fi xed rates of 18 percent and adjustable rates of close to 30 percent.29

One of the sadder instances of loan fl ipping involved Mary Podelco, a former waitress with a sixth-grade education who had lost her husband in 1994. She used his life insurance to pay off the mortgage on her family home. A year later, in need of new windows and a heating system, she took out a loan with Benefi cial Finance for $11,921. Just one month later, Benefi cial convinced her to refi nance the loan for $16,256. Soon other lenders got into the game, each promising Ms. Podelco a loan that was superior to the one she had. Over the course of a year, lenders fl ipped her loan at least fi ve times, increasing her outstanding debt to over $64,000. Unbeknownst to Ms. Podelco, she was paying exorbitant charges with every fl ip. On July 26, 2001, long before the subprime heyday, she told her story to the Senate Committee on Banking, Housing and Urban Affairs.30

Sometimes lenders urged borrowers to take out mortgages and use the funds to pay off outstanding medical debts, credit cards, or other bills. By consolidating their debts, lenders argued, borrowers would get lower interest rates and lower monthly payments. What the lenders didn’t say was that by converting unsecured debt into debt secured

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

24 • PART I THE SUBPRIME MARKET TAKES OFF

by their homes, the borrowers put their homes at risk. Going into bankruptcy, more- over, would not wipe out their mortgages, unlike other debts.31

Another early predatory tactic was charging borrowers for credit insurance that would be used to pay off their loans if they became disabled or died. These policies charged a one-time premium in the many thousands of dollars that was paid at the closing and fi nanced as part of the loan. Because borrowers had to pay interest on the premiums, the effective cost was as much as three or four times the original amount. This practice made the front page in 2002 when Citigroup ponied up $240 million to settle litigation against Associates First Capital, which Citigroup had purchased in 2000 for $31 billion amid allegations of similar abuses. These were not the only allegations against Citigroup. Reporter Michael Hudson, who wrote an early exposé on CitiFi, told of a borrower whom CitiFi convinced to take out not only credit life insurance but also disability and unemployment insurance.32

Lenders packed other exorbitant fees into loans. FAMCO, one of the largest and earliest predatory lenders, reportedly charged borrowers as much as 25 percent of their loan amount in discount points. This meant a borrower with a $100,000 loan would pay $25,000 in points. Typically, FAMCO’s loans also included prepayment penalties and high interest rates. Eventually, FAMCO limited its points to 10 per- cent of the loan amount because of concern about the “sound-bite effect of high origination fees.”33

Bait-and-switch schemes were also rife. At the time of application or shortly after- ward, lenders would describe the loan terms to borrowers, but not actually lock in the terms. Lenders would then change the terms after the borrowers were psychologically and fi nancially invested in the loans. This was countenanced by federal disclosure laws, which only prohibited lenders from changing loan terms if they had made binding offers. In the subprime market, offers were almost never binding. Borrowers would show up at their loan closings expecting the promised loan terms, only to fi nd that the terms had become worse in major ways. Fixed-rate loans became adjustable and interest rates soared. Surprise fees popped up in the loans. Second mortgages suddenly appeared in the documents.

Often the borrowers did not even recognize these changes in the hubbub of the closing. The closing agents would sit the borrowers down with a big stack of papers and fl ip the pages, directing borrowers to sign next to the sticky arrows. If the borrow- ers protested that things were moving too fast, that they wanted to review the docu- ments, or that the terms appeared different, the response would be, “Don’t worry, I’ll take care of that, just sign here.”34

FAMCO reportedly pulled such a bait and switch on Bernae and Scott Gunder- son. After the loan closing, Ms. Gunderson looked through the loan documents and saw that the terms were worse than the ones she and her husband had agreed to. She talked with a manager at FAMCO, who assured her that the loan terms were as promised. Unbeknownst to the manager, Ms. Gunderson recorded the conversation. That recording proved invaluable after the Gundersons determined that FAMCO had added $13,000 in fees to the loan and put them in a loan with an interest rate that rose 1 percent every six months.35

Others were not so lucky. Roberta Green thought she was applying for a $6,000 home equity loan at a fi xed rate. The broker fi lled out the loan application for Ms. Green

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 25

and, as required by law, disclosed the interest rate and fees. Later, the loan closing was so rushed that Ms. Green did not realize she had agreed to refi nance her current mortgage for $76,500 with a higher adjustable interest rate and $6,500 in additional fees, none of which the broker had previously mentioned.36

Lenders and brokers even resorted to duress to close loans. Back in 2001, Crain’s Chicago Business published a report about a mentally disabled couple who had fallen behind on their real estate taxes. A broker approached them offering a loan to cover their taxes, using the equity in their home. On the day of the closing, a limousine brought the couple from Chicago’s South Side, where they lived, to an offi ce on the north side of Chicago near O’Hare International Airport. When the couple examined the loan documents, they discovered that the loan terms had been changed. They were far from home and did not know where they were or how to get home so they caved in and signed the papers.37

Intimidation was another tool of the subprime trade. A borrower with a CitiFi- nancial loan reported that when she missed some loan payments, a CitiFi manager threatened to have her arrested and to tell her boss that “she was a deadbeat.”38

The most brazen lenders and brokers lied about loan terms. It was common to tell borrowers that their loans were for fi xed rates when, in fact, they were not. Other misrepresentations took the form of false promises. Lenders would tell borrowers that they would quickly refi nance their loans to lower the interest rate. Later, when the bor- rowers pressed the lenders to honor their promises, the lenders would concoct excuses why better rates were not possible.39

The exploitative practices of early predatory lenders were summed up in the 1998 testimony before the Senate Special Committee on Aging by a former fi nance com- pany employee under the pseudonym of Jim Dough:

My perfect customer would be an uneducated widow who is on a fi xed income— hopefully from her deceased husband’s pension and social security—who has her house paid off, is living off of credit cards, but having a diffi cult time keep- ing up [with] her payments, and who must make a car payment in addition to her credit card payments. . . .

We were instructed and expected to fl ip as many loans as possible. . . . The practice is to charge the maximum number of points legally permissible for each loan and each fl ip, regardless of how recently the prior loan that was being refi - nanced had been made. The fi nance companies I worked for had no limits on how frequently a loan could be fl ipped, and we were not required to rebate any point income on loans that were fl ipped. . . .

Our entire sale is built on confusion. Blue-collar workers tend to be less educated. I know I am being very stereotypical, but they are the more unso- phisticated. They can be confused in the loan closings, and they look to us as professionals. . . . [T]hey are more trusting toward us.40

SUBPRIME GOES MAINSTREAM

Over time, the subprime industry began moving from fringe to mainstream as commer- cial banks, investment banks, hedge funds, insurance companies, and other fi nancial

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

26 • PART I THE SUBPRIME MARKET TAKES OFF

giants saw the profi ts that could be made. They soon started buying subprime lenders. Small mortgage banks mushroomed into large national behemoths, absorbing smaller entities along the way. All this buying of subprime lenders led to widespread consoli- dation in the industry. For example, one of Cleveland’s leading bank holding com- panies, National City Corporation, bought the subprime lender First Franklin in 1999 for $266 million. In a few short years, First Franklin’s subprime lending volume sky- rocketed from $4 billion to $30 billion. With a subprime lender in its pocket, National City Bank could originate prime loans in its own name and, in the words of its chief executive, David Daberko, refer otherwise lost prospects “immediately to the nonprime company.”41 Later First Franklin became infamous for bringing down both National City Bank and Merrill Lynch.42

The global banking giant HSBC bought its own subprime lender, Household International, in 2003 for $14 billion. Investment banks Credit Suisse, Bear Stearns, Lehman Brothers, Merrill Lynch, Morgan Stanley, and Goldman Sachs all bought or founded nonbank subprime lenders to feed their securitization machines. As a Wall Street Journal reporter noted, “Without a production-line of mortgages, the inventory for all those fee-paying securities would dry up.”43 Private equity fi rms like New York’s Capital Z Partners snapped up subprime lenders. Even blue-chip companies got swept up in the buying frenzy. General Electric bought WMC Mortgage Corporation in 2004. H & R Block bought Option One Mortgage Corporation in 1997.

The new owners of the subprime lenders piously avowed that they had cleaned up “shop” and would never sanction abusive lending. In testimony before the House Com- mittee on Government Oversight and Reform, the former chief executive offi cer of Lehman Brothers, Richard Fuld, said: “When we bought [subprime lenders], we changed management, we changed underwriting standards to make them much more restrictive, to improve the quality of the loans that we did in fact originate so that those loans that we did then put into securitized form would be solid investments for investors.”44

At the same time, independent mortgage banks that once had bit roles grew into mammoth institutions fed by securitization. Ameriquest and Countrywide were two of the most egregious mega-subprime lenders. Initially a small California thrift called Long Beach Savings and Loan, Ameriquest became a privately held mortgage lender in 1994 and quickly grew to secure a position as one of the largest mortgage companies in the United States. Ameriquest made loans through retail operations and indepen- dent mortgage brokers. The latter branch of the company went by the name Argent. Ronald Arnall, who founded Long Beach and stood at Ameriquest’s helm, was the country’s 106th wealthiest billionaire by 2004. By 2005, Ameriquest was reaping suf- fi cient fees to sponsor the Super Bowl XXXIX half-time show in Jacksonville, Florida. The Ameriquest blimp hovered over the stadium touting the company’s success.45

Arnall and his wife, Dawn, were big political contributors, raising over $12 million for President George W. Bush and various conservative advocacy organizations.46 On August 1, 2005, President Bush nominated Arnall to be the next ambassador to the Netherlands, a position he held until 2008. On the same day that Bush nominated Arnall for the ambassadorship, Ameriquest agreed to fork over $325 million to settle pending lawsuits and investigations centered on Ameriquest’s and Argent’s question- able lending in dozens of states.47 By 2007, Ameriquest had tanked. It shut down its retail mortgage shop and sold what was left of the company to Citigroup.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 27

Countrywide, like Ameriquest, grew at an astounding rate and generated huge returns for its investors. Between 2000 and 2006, its securities trading volume went from $647 billion to $3.8 trillion. Fortune magazine reported that a $1,000 investment in the company in 1982 was worth $23,000 twenty years later. This 2,200 percent return more than outpaced returns at Wal-Mart and even Warren Buffett’s Berkshire Hathaway. Members of Countrywide’s board of directors were handsomely compensated, with some receiving over half a million dollars a year. Countrywide’s chief executive offi cer, Angelo Mozilo, was paid up to $43 million a year.48

By 2005, Countrywide had become the nation’s largest subprime lender. Two years later, the company went into a subprime skid and, in 2008, was acquired by Bank of America amid rumors that the lender was on the verge of bankruptcy. Shortly after Bank of America completed the sale, Countrywide committed over $8 billion to settle abusive lending claims with dozens of states.49

Lending Channels Borrowers could get subprime loans through three main channels: the retail channel, the wholesale channel, and the correspondent channel. The retail channel is the sim- plest to explain. Retail lenders took applications in person, over the Internet, and through call centers, using in-house loan offi cers instead of outside mortgage brokers. These lenders processed the applications, underwrote the loans, and funded them once approved. Many retail lenders were depository institutions. For example, Wash- ington Mutual Bank (WaMu), the savings and loan giant, made subprime loans directly to borrowers through its retail branches.

More subprime loans, however, came through independent mortgage brokers, not loan offi cers at retail lenders. This was known as the wholesale channel because bro- kers generated loan applications for wholesale lenders who underwrote and funded the loans. Wholesale lenders could either be depository institutions or nonbank fi nance companies that raised money on the capital markets and used the money to fund their loans. Eventually, the loans were sold, at which time the wholesale lenders would profi t from the difference between the sales price and the cost of funding the loans. At the peak of subprime lending, almost 80 percent of subprime loans were originated through some form of wholesale lender.50

The third channel was called correspondent lending. Correspondent lenders, which could be depository institutions or fi nance companies, had retail operations where they took applications and made loans pursuant to underwriting standards set by a wholesale lender, who committed in advance to buy the loans at a set price. In the correspondent setting, the wholesale lenders served as loan aggregators.

Mortgage channels had two more twists. The fi rst was an arrangement known as table-funding, where on paper brokers appeared to make the loans, but in reality, the brokers only held the loans for a matter of seconds. The brokers would immediately endorse the loan notes over to wholesale lenders who were the true funders of the loans. Table-funding enabled brokers to make more fees.

The other twist was net branch banking, through which brokers became temporary employees of wholesale lenders, typically to avoid disclosing to borrowers the commis- sions that they received and to circumvent state licensing requirements. Many lenders

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

28 • PART I THE SUBPRIME MARKET TAKES OFF

used multiple channels. One lender could have retail and wholesale operations and also purchase loans from correspondent lenders.51

Housing Bubble The supply of subprime loans increased as the subprime industry went mainstream and investors fl ooded the market with capital. At the same time, there was increased demand for subprime loans. Demand was driven by a combination of stagnant real wages for most workers, rising interest rates, and rising home prices. The average U.S. home rose over 50 percent in value between 2001 and 2005. On the West Coast, the increases were even higher, with homes increasing 20 percent in just one year between 2004 and 2005. By 2006, only 17 percent of California households could afford a median-priced single-family home.52

The subprime industry contributed to rising home prices. Easy credit generated greater demand for housing, which, in turn, drove up housing prices. Higher-priced homes meant that homebuyers had to borrow ever larger amounts of money. Even borrowers with good credit histories found that they could not qualify for safe loans because they could not afford the monthly payments for fi xed-rate, fully amortizing mortgages. In high-priced markets, some borrowers had to resort to jumbo nonprime loans because the loans they needed exceeded Fannie and Freddie limits on the size of the loans they would buy. Real estate investors also added to the demand. These purchasers bought homes with the intention of selling them in a few years and cashing in on the expected rise in home values.

Housing developments sprung up like weeds in western deserts, trash-strewn lots in cities, and farmland in the Midwest. Some developments were geared to high- end buyers who wanted McMansions with in-home movie theaters. Others marketed cheap, remote tracts to fi rst-time homeowners. Developers partnered with lenders and brokers, providing buyers with one-stop shopping for a house and a loan. Country- wide established nearly 800 local offi ces that solicited loans through home builders and real estate agents. Developers and lenders sent a message that anyone could own a home and that people should dream big.53

Feeding the Mortgage Machine From lenders’ perspective, the mortgage machine needed constant feeding in order to generate constant fees. Volume was what mattered. Lenders bought “leads” from fi rms that gathered consumer information from banks and credit bureaus and culled public records to create mega-databases on consumers. These databases, referred to as “farm- ing kits,” contained information on individual borrowers that lenders used to personal- ize their solicitations. Between 2004 and 2006, Countrywide mailed between six and eight million targeted solicitations each month and made many tens of thousands of phone calls. The solicitations made people believe that the new loan products were designed for them and were all pleasure and no pain.54

The president of one data fi rm explained that the people his company identifi ed as prospects included consumers who were in fi nancial trouble and likely needed to refi nance to avoid default. Lenders could buy a list of 2,500 distressed subprime bor- rowers for about $500. Some lenders mined their own data to pinpoint borrowers in need of refi nancing. Countrywide reportedly identifi ed borrowers who were behind in their payments and offered to refi nance their loans. In 2004, one out of every nine

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 29

loans by Ameriquest refi nanced a loan that Ameriquest had made within the previous two years.55

Credit bureaus offered other ways to identify people to target. Each time indi- viduals submitted applications for mortgage loans, lenders contacted credit bureaus to check their credit. The credit bureaus compiled lists of borrowers on whom inquiries were made and sold the lists to other lenders, who would then solicit the borrowers. In addition to supplying data, these fi rms developed models, using information on bor- rowers’ behavior, to create profi les of borrowers who showed a “statistical propensity to acquire new credit.”56

Dodgy Practices Although subprime lending moved from sketchy to mainstream institutions, the industry did not clean up its practices. Bait-and-switch tactics persisted, with lenders surprising borrowers at loan closings with thousands of dollars in unexpected fees and higher interest rates. Often the new interest rates were adjustable and could double or triple over time. When borrowers protested, lenders assured them that they would soon be able to refi nance. What lenders did not say was that when the borrowers refi - nanced, they would have to pay origination and closing fees all over again plus likely prepayment penalties for the early payoff of their loans.57

Lenders made out big from steering, which was the practice of conning borrowers who qualifi ed for cheap prime loans into agreeing to costlier subprime loans. Behind the scenes, lenders set minimum prices (a combination of the interest rate, points, and fees) they would accept for each type of loan, taking into account borrowers’ credit scores and factors like the amount of equity they had in their homes. This resulted in a sliding scale of prices, a practice known as risk-based pricing. Subprime lend- ers kept the real risk-based price a secret; borrowers only knew the price they were offered, which often exceeded the risk-based price. When borrowers paid more than the minimally acceptable price, lenders made more money. According to the Wall Street Journal, 55 percent of all subprime loans in 2005 went to people with suffi ciently high credit scores to qualify for prime loans.58

Another gambit was to generate hundreds of dollars in “junk” fees for check- ing borrowers’ credit, processing applications, and preparing loan documents. Firms charged borrowers up to $75 just to send emails. Lenders captured profi ts through other means, too. Investors paid more for certain loan attributes, with default interest rate clauses being a favorite. Under these clauses, the interest rate would skyrocket if a borrower missed a payment. Of course, if the interest rate went up, so did the likeli- hood of future defaults.59

Prepayment penalties were found in about 80 percent of subprime loans and typi- cally required borrowers to pay six months’ worth of interest if they refi nanced within a stated period, commonly one to fi ve years. On a $150,000 loan with a 12 percent interest rate, the prepayment penalty would be $7,500. A 2004 study estimated that prepayment penalties cost borrowers a total of $2.3 billion a year.60 Prepayment penal- ties had pernicious effects, including locking borrowers into high-cost loans. As their credit profi les improved, borrowers could not refi nance into cheaper loans unless they generated the cash to pay the penalties. Prepayment penalties also increased the odds of default by trapping borrowers in high-cost loans. One study found that loans with

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

30 • PART I THE SUBPRIME MARKET TAKES OFF

prepayment penalties longer than three years were 20 percent more likely to default than comparable loans without those penalties.61

Lending According to Stereotype The race-based targeting that was a signature of early predatory lending also contin- ued unabated. Subprime mortgages were more heavily concentrated in predominantly black and Latino neighborhoods relative to white communities, even when controlling for income and credit scores. Researchers estimated that half of home mortgage loans made to African-Americans before the housing market collapsed were expensive sub- prime loans.62

Damning evidence to this effect appeared in an affi davit by former Wells Fargo loan offi cer Tony Paschal. Paschal asserted that Wells Fargo employees referred to borrowers of color as “mud people” and their loans as “ghetto loans.” Loan offi cers, he reported, lowered interest rates for whites when interest rates fell before closing, but told black borrowers that their rates were “locked” and they could not take advantage of interest rate declines. Paschal also accused Wells Fargo of deliberate marketing practices directed at African-Americans, including a software program that would send fl yers to consumers based on their “language.” One such “language” was “African- American.”63 Another former Wells Fargo loan offi cer, Elizabeth Jacobson, attested that the company approached black churches in hopes that each minister would “con- vince the congregation to take out subprime loans with Wells Fargo.”64

Lenders and brokers saved the worst deals for borrowers they considered easy marks. In a study of loans insured by the Federal Housing Administration, researchers looked at differences in the amounts borrowers paid in fees based on race and education. The study controlled for factors such as credit scores and home values that could have infl u- enced default risks and therefore the cost of credit. On average, African-American borrowers paid $414 more in fees and Latinos $365 more than equivalent whites.65

These fi ndings were not unique. Studies by governmental agencies and consumer groups have consistently found that people of color pay more for credit even after con- trolling for creditworthiness. Setting the cost of mortgages based on borrowers’ race is blatantly unfair and discriminatory. It also increases the likelihood that borrowers will lose their homes because their mortgages are less affordable than those offered to equivalent white borrowers.66

Education mattered, too. In one study, borrowers with college degrees paid, on average, $1,100 less than those without a college education even though the latter borrowers purchased essentially the same products, presented the same level of default risk, and lived in similar communities and types of housing.67 Not surprisingly, given these results, subprime loans are concentrated in areas where people have lower educa- tional levels, even when those areas have the same default risk as neighborhoods with better educated residents.68

Outright Fraud The quest for ever higher revenues went hand in hand with fraud. Too often, brokers and lenders did whatever it took to close a loan. That might involve padding a bor- rower’s income or assets (with or without the borrower’s knowledge), commissioning infl ated appraisals, manufacturing fake pay stubs and W-2s, altering credit reports, and creating fi ctitious checks and investment statements. These practices, according to a

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 31

study by Fitch Ratings, were more pronounced among loans made through indepen- dent brokers than through loan offi cers. This was probably because brokers didn’t make any money if borrowers could not satisfy lenders’ underwriting criteria.69 In Cleveland, for example, eighty-three people were charged with mortgage fraud in a scheme to defraud lenders by fabricating employment and income records and then hiring people called “backstoppers” who would verify the information when lenders called as part of the underwriting process.70

Lenders and brokers did not work alone. There are stories of professionals who helped them commit fraud, like the CPA who allegedly was in the “back pocket” of a New Century account executive, providing letters verifying borrowers’ self-employment income that did not exist.71 A bank teller reportedly doctored bank statements for loan originators to help borrowers qualify.72 According to the New York Times, some bro- kers even offered bribes to underwriters to accept loan applications that other lenders had rejected.73 In 2002, following an intense FBI investigation, “two accountants, four title companies, fi ve appraisers, eight underwriters and forty mortgage brokers” were indicted for their involvement in a mortgage fraud scheme allegedly orchestrated by American Home Loans.74

Brokers and lenders were skillful at fi nding appraisers who would come in with property valuations that would satisfy underwriters. In a practice called “hitting the bid,” they would handpick appraisers and tell them what the property needed to be worth for the loan to go through.75 Appraisers reported feeling “bulldozed” into infl ating the value of homes and misrepresenting the condition of properties.76

When the State of New York got wind of allegations of such practices at Washington Mutual, it brought a lawsuit against an appraisal outfi t that reportedly succumbed to pressure from WaMu. According to the complaint, WaMu generated a list of preferred appraisers from eAppraiseIT who could be counted on to bring in high valuations.77

Some loan originators engaged in even more blatant forms of fraud. One mort- gage bank made a name for itself by having borrowers sign duplicate copies of their loan and mortgage agreements. The lender would then sell the notes to two different entities, each of which thought it held the “real” note. This practice, called “double- booking,” enabled the mortgage bank to retain all the funds from the sale of both notes. The borrowers would only pay on one of the notes, not realizing they owed on two. The investor that was not receiving any payments would then go after the bor- rowers, only to discover the fraud.78 In other situations, borrowers would attempt to refi nance their loans, only to learn that there were multiple fi ctitious mortgages on their property.79

In a damning affi davit, a former account executive for a subprime lender described his fi rm’s practices:

• Ameriquest taught . . . Account Executives to infl ate the stated value of the cus- tomer’s property for the purpose of qualifying them for a refi nance loan. I recall an Ameriquest area manager indicating that appraisal values should regularly be pushed by at least 10–15 percent.

• It was a common and open practice at Ameriquest for Account Executives to forge or alter borrower information or loan documents. For example, I saw Account Executives openly engage in conduct such as altering borrowers’ W-2

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

32 • PART I THE SUBPRIME MARKET TAKES OFF

forms or pay stubs, photocopying borrower signatures and copying them onto other, unsigned documents, and similar conduct . . .

• Account Executives regularly concealed or obfuscated that a loan was an adjustable rate mortgage, rather than a fi xed. In fact, it was common practice for Account Executives to refer to adjustable rate loans as “fi xed adjustable” loans.80

FOLLOW THE MONEY

People wonder why brokers and loan offi cers wanted to rip-off their customers with high-cost products. The answer is commissions. Think used car salespeople. In fact, the disreputable subprime lender FAMCO “recruited highly paid automobile sales representatives, who were at ease with the ‘hard sell’ but tired of the long evening hours and weekend work that auto sales involved.”81

Every day lenders sent brokers and loan offi cers—if they had retail operations— rate sheets refl ecting the minimum price they would accept for each type of subprime product they offered. Products varied by interest rate, points, and prepayment pen- alties. The prices took into account product features and borrower information like credit scores.82 Most importantly for brokers, the rate sheets spelled out the commis- sions, called yield spread premiums (or YSPs), that brokers could earn if they steered borrowers to higher interest loans with prepayment penalties. Banks often had parallel compensation systems for their loan offi cers, called “overages,” if they induced bor- rowers to take out loans on costlier terms. Both types of compensation were essentially legalized kickbacks.

Brokers and lenders virtually never gave borrowers copies of the rate sheets or told them the lenders’ bottom line prices. And, most lenders expressly prohibited disclo- sure of their rate sheets to borrowers. Every incentive was for brokers and loan offi cers to use the rate sheets to steer borrowers into high-priced loan products to maximize their commissions.83

Some fi rms were particularly generous with their YSPs. New Century, which is now among the ranks of bankrupt lenders, rewarded brokers with handsome YSPs. Amber Barbosa, who got into the loan origination business as an employee at New Century, eventually became an independent mortgage broker. At twenty-eight years old, with no college degree, Ms. Barbosa made $500,000 a year in YSPs. She drove a Mercedes CLS 500 and a Cadillac Escalade and owned three pieces of property, including one with an ocean view. She described YSPs of $15,000 to $20,000 “as kickbacks.”84

YSPs and the complex loan terms that usually accompanied them had signifi cant effects on the price of credit. One study found that borrowers paid between $800 and $3,000 extra per loan if the lender paid a YSP to the broker, with an average added cost to borrowers of $1,046.85 Another study revealed that subprime borrowers who did not use brokers paid almost $36,000 less over the life of a loan than their counterparts who went through brokers.86

There were other pecuniary incentives for loan offi cers to steer borrowers who applied for prime loans to subprime products. According to Elizabeth Jacobson, formerly at Wells Fargo, loan offi cers in the prime division made more money if they referred

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 33

prime-eligible borrowers to loan offi cers handling subprime products. They would persuade borrowers to make the switch from prime to subprime saying that the pro- cessing time was shorter and required less documentation and no down payment. Others advised borrowers not to make down payments, which then made them ineli- gible for prime loans.87

Commissions were often based on a percentage of the loan principal, which meant that the larger the borrower’s loan, the larger the commission. This led loan origina- tors to encourage borrowers to take on more debt than they had originally requested. Brokers and loan offi cers could also generate fees by packing in products like credit life insurance.

All these opportunities for compensation were pernicious, not just because bor- rowers paid more than they should have. By infl ating the interest rates, the size of the loans, and the various fees borrowers had to pay, the commissions substantially increased the likelihood that subprime loans would default and go into foreclosure. This increased risk was not insubstantial. Economists have calculated that for every 1 percent increase in the initial interest rate of a home mortgage, the chance that a loan will go into default rises by 16 percent a year.88

THE RACE TO THE BOTTOM

Starting in 2003, the competition for loans to securitize got fi erce. Everyone who wanted to refi nance had done so, and there were more and bigger lenders fl ocking to the market. As one insider put it, “We ran out of borrowers. . . . Everybody that could qualify, anybody that could fog a mirror . . . had basically been refi nanced once, twice, three, sometimes four times.”89 At the same time, the spread between the interest rate lenders paid to borrow money and the interest rate they could charge borrowers was narrowing because of increased competition.

The year 2003 was when interest rates started to rise. Borrowers were increasingly shut out of the mortgage market because of higher home prices and higher interest rates. The once-booming mortgage business stalled, leading one reporter to write, “Fear is rampant that the housing boom is over. . . . [W]ill the fi nancial companies that rode the rocket fi zzle along with their best product, mortgages?”90

Lenders were desperate for new sources of mortgages. The “solution” was the expansion of the market through two techniques: risky new products called “non- traditional mortgages” and relaxed underwriting standards and loan terms, both of which were designed to qualify more borrowers. Most of the new, nontraditional mortgages combined artificially low initial payments with eye-popping monthly payments after a few years. This allowed lenders to qualify borrowers for loans based solely on the lower initial payments, without regard to whether the borrow- ers could make higher monthly payments later on. The assumption was that the borrowers could always refinance out of these loans if and when their payments became unaffordable because housing prices would keep going up. In describ- ing a nontraditional loan known as a hybrid ARM, New Century personnel laid out the risks: “Inevitably, the borrower lacks enough equity to continue this cycle (absent rapidly rising property values) and ends up having to sell the house or face foreclosure.”91

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

34 • PART I THE SUBPRIME MARKET TAKES OFF

Nontraditional Mortgages Lenders peddled a cornucopia of risky nontraditional mortgage products to borrowers during the housing bubble: hybrid ARMs, interest-only loans, pay-option ARMs, and loans with negative amortization. The emergence of these products marked a new phase in the subprime market. Historically, subprime had referred to features of bor- rowers. Typically, subprime borrowers had been people with blemished credit histo- ries, often with signifi cant amounts of equity in their homes. In the second iteration of subprime lending, the word subprime shifted to describe the type of loan, not the fea- tures of borrowers. These new products were also referred to as Alt-A or nonprime loans; prime loans were called A loans.

Nontraditional mortgages experienced a meteoric rise. Of these, hybrid ARMs, interest-only mortgages, and pay-option ARMs accounted for a growing share. Pay- option ARMs and interest-only mortgages went from 3 percent of all nonprime orig- inations in 2002 to well over 50 percent by 2005.92

The most common nontraditional mortgages were hybrid ARMs. By 2004 and continuing through 2006, hybrid ARMs represented about three-fourths of the loans in subprime securitizations. Hybrid ARMs had fi xed initial rates that reset into adjust- able rate mortgages in a set number of years. Often they were called 2/28s or 3/27s; the two numbers referred to the respective lengths of the fi xed-rate and adjustable-rate periods. For example, a 2/28 had a fi xed rate for two years and then converted to an adjustable rate for the next 28 years, with the rate usually adjusting every six months. The adjustable rate was calculated by adding a “margin” to an abstruse index such as the London Interbank Offered Rate (LIBOR), which is the rate at which London banks lend to each other. For example, say that a hybrid ARM had a margin of 4 percent and LIBOR was currently 5 percent; the adjustable rate would be 9 percent.93

Hybrid ARMs contained a hidden risk of payment shock—the risk that monthly payments would rise dramatically upon rate reset. The potential payment shock was worse than with traditional ARMs, which had lower reset rates and manageable life- time caps. Indeed, with hybrid ARMs, the only way interest rates could go was up. During the housing bubble, many subprime hybrid ARMs had initial rate resets of three percentage points, resulting in increased monthly payments of as much as 50 percent.94

Even more dangerous were interest-only ARMs. With these loans, borrowers only paid interest for an initial period of anywhere from six months to fi ve years. Once the introductory period expired, the borrowers’ payments went up, often substantially, for the same reason as hybrid ARMs, plus two more. First, after the initial period, the loan began to amortize and borrowers had to pay principal as well as interest. In addition, the principal payments were higher than they would be under a fully amortizing loan because there were fewer years left to pay off the principal. As a result, the payment shock on interest-only ARMs was often worse than on hybrid ARMs.95

The most toxic nontraditional mortgages, however, were “pay-option” or “pick-a- pay” ARMs. These loans allowed borrowers to select among three options each month: (1) paying the full principal and interest; (2) paying only the interest; or (3) paying some set amount that was less than the interest payment. When borrowers chose the third option, the principal balance of their loans actually grew over time until the principal reached a set limit, usually around 120 percent of the original loan balance.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 35

Seventy percent of borrowers with pay-option ARMs picked the minimum pay- ment option until they hit their limit, at which point their loans were “recast.” Then they faced much higher monthly payments that amortized the principal, including the interest that had been deferred, plus interest. The monthly payments went up even more because the loan was amortized over the remaining loan period, not the full loan term. In the process, borrowers’ payments easily doubled or tripled overnight.96

The loan disclosures for pay-option ARMs typically only showed what the pay- ments would be if borrowers made the minimum payments before recasting. With respect to possible, later higher payments, the disclosures often only provided a hypo- thetical involving a $10,000 loan and let the borrowers do the math.97 Not surpris- ingly, with their woefully defi cient disclosures and their high-risk profi le, pay-option ARMs were “the most likely” of all nontraditional mortgages “to default.”98

Lenders deliberately marketed pay-option ARMs to borrowers to obfuscate the back-end risk. For instance, Countrywide reportedly required the loan offi cers in its Full Spectrum Lending Division to memorize the following script: “Which would you rather have, a long-term fi xed payment or a short-term one that may allow you to realize several hundred dollars a month in savings? I am able to help many of my clients lower their monthly payments and it only takes a few minutes over the phone to get started.”99 Edward Marini, a disabled veteran who had a pay-option ARM with Countrywide, was left believing that his mortgage had low payments for fi ve years. To his surprise, three years after the closing, he learned that his monthly payment was about to triple from $1,300 to $3,800. The new payment amount exceeded his monthly income of $3,250.100

Lenders knew they were potentially putting pay-option ARM borrowers in unten- able positions with these products that were untested in the mass market. Coun- trywide’s chief executive offi cer, Angelo Mozilo, was quoted as saying that lenders were “fl ying blind on how [they would] perform in a stressed environment of higher unemployment, reduced values and slowing home sales.”101 That did not deter his company from making almost $750 billion worth of pay-option ARMs between 2004 and 2007.102

Easy Terms and Loose Underwriting Standards During the housing bubble, the other way lenders kept up volume was to loosen underwriting standards to qualify more borrowers for loans. If a down-payment requirement or low equity was a concern, no problem: the lender just reduced the minimum required equity in the home. If the borrower’s income was an issue, no prob- lem: the lender would just do a stated-income loan.

Loans with high loan-to-value (LTV ) ratios reduced the need for borrowers to come up with large down payments when buying homes. Similarly, high LTV loans made it possible for borrowers to refi nance with little or no equity. As a result, between 2001 and 2006, the average LTV among subprime loans increased from 79.4 per- cent to 85.9 percent. Some loans even had LTVs of over 100 percent, meaning they exceeded the value of the homes that served as the collateral. Not surprisingly, loans with high loan-to-value ratios were more likely to default.103

Piggyback loans were another technique to reduce down-payment or minimum equity requirements. These loans had the added attraction of obviating the need for

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

36 • PART I THE SUBPRIME MARKET TAKES OFF

borrowers to pay private mortgage insurance (PMI). In the past, if borrowers did not make down payments of at least 20 percent, they had to buy PMI, which would pay off the lenders if the borrowers defaulted. During the lending boom, lenders skirted these requirements by using piggyback loans.

Piggyback loans worked like this: The lender gave the borrower a fi rst mortgage for 80 percent of the value of the home. Then it granted a second loan covering all or part of the difference between the fi rst 80 percent and the remaining value of the home. Piggyback loans and other loans with combined LTVs of 100 percent grew at exponential rates, reaching almost 40 percent of all subprime loan originations by 2006.104

Piggyback loans layered another level of risk on top of subprime mortgages. When piggyback loans left borrowers with no equity, any depreciation in their homes put their loans “under water.” And if the fi rst mortgage was sold through securitization, the loan’s new owner often did not know that there was a “silent second” loan or that the home was 100 percent leveraged. Countrywide’s Angelo Mozilo wrote in an inter- nal email that he had “never seen a more toxic product.”105

Lenders further dropped their underwriting standards in response to the fact that housing prices were rapidly outstripping workers’ stagnant wages. They did this by lowering or eliminating their documentation requirements for income, assets, and jobs. In a “stated-income” or low-documentation loan, applicants reported their income, but did not provide proof. In a no-documentation loan, the loan was under- written with no information at all on the loan applicant’s income, stated or otherwise. With a NINA (no income, no assets) loan, the income and asset fi elds were left blank on loan applications. The only thing worse were NINJA (no income, no job, no assets) loans, which lenders made without any information on loan applicants’ income, assets, or jobs.

At the outset, stated-income loans were limited to prime, fi xed-rate loans requiring higher credit scores to qualify. This changed over time as lenders abandoned their underwriting requirements. Estimates are that at the top of the bubble, 80 percent of Alt-A loans and almost 40 percent of subprime loans were low-doc or no-doc loans. Low-doc and no-doc loans were most prevalent in rapidly appreciating markets like California, Nevada, Florida, and Arizona, where high prices made it hard for borrow- ers to qualify for loans. And like high LTV loans, stated-income and no-doc loans were more likely to default than full-documentation loans.106

Low-doc and no-doc loans were particularly noxious because they invited decep- tion. They soon became known as “liar’s loans.” Borrowers could put whatever fi gure they wanted down for their income and not back it up with tax returns or pay stubs. Likewise, examples a bound of brokers or loan offi cers who entered false income, asset, and job information on loan applications without borrowers’ knowledge. World Savings Bank, which Wachovia Corporation later purchased, allegedly qualifi ed a widow for a loan based on her deceased husband’s income with the knowledge that he was no longer alive.107 According to one study, close to 60 percent of applications for stated-income loans infl ated borrowers’ incomes by at least 50 percent above the amounts reported to the IRS.108 Another study of loans that went into default shortly after they were originated found that up to 70 percent of the loan applications in ques- tion contained false information.109

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 37

Lenders could have verifi ed borrowers’ income, but few bothered to do so. Not verifying borrowers’ income and assets made underwriting faster and cheaper. Wash- ington Mutual allegedly promoted low-doc and no-doc loans with a fl ier stating that “a thin fi le is a good fi le.”110 Countrywide called its low-doc product “Fast and Easy” because Countrywide could issue loan approvals without having to wait for pay stubs or income tax returns from applicants.111

Brokers and loan offi cers often earned higher commissions on low-doc and no-doc loans. Fees on a $300,000 stated-income loan went as high as $15,000. A compara- ble fi xed-rate, full-documentation loan would generate less than $5,000 in fees. No wonder brokers and loan offi cers often tried to steer borrowers to reduced documenta- tion products. Borrowers often fell for the ruse, either because brokers or loan offi cers convinced them that they had to close the deal quickly before rates went up or because they liked the convenience of low-doc and no-doc loans.112

Another way to cut corners was to qualify borrowers based on their monthly pay- ments for principal and interest without escrowing for homeowner’s insurance and property taxes. Historically, lenders had required borrowers to escrow insurance and property taxes as part of their monthly mortgage payments. Lenders would disperse the escrow funds when the tax and insurance bills came due. By not escrowing these expenses, lenders could make unaffordable loans appear affordable. The danger, how- ever, was that when the bills for taxes and insurance came in the mail, borrowers would not have the money to pay them.113

As if the risk of any one of these practices or products was not enough, lenders began layering multiple risks. For example, a lender might combine an interest-only ARM with a piggyback loan and dispense with verifying the borrower’s income. Each of these features increased the risk of the loan. A 2008 study found that stated-income loans with 100 percent loan-to-value ratios were particularly treacherous and had the highest rates of default.114

FIGURE 2.2.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

38 • PART I THE SUBPRIME MARKET TAKES OFF

All told, the new loan products spurred a sixfold increase in nonprime lending from 2000 through 2005. By January 2009, outstanding subprime mortgage debt, including Alt-A mortgages, was close to $2 trillion.115

Crowd-Out Effect ARMs offered lower initial monthly payments than fi xed-rate, fully amortizing loans, which allowed irresponsible lenders to outcompete safe lenders. Soon, banks and other conservative lenders came to realize that it didn’t pay to compete on fi xed-rate prime loans, so they expanded into the nonprime market as well. That is why hybrid sub- prime ARMs, interest-only mortgages, and pay option ARMs captured a growing part of the market during the housing bubble.116

Lenders were under constant pressure to increase production “by hunting down more borrowers, selling more loans, and processing loans as quickly as possible.”117

This pressure led to problems at every stage of loan production. Lenders competed fi ercely for brokers’ allegiance. To entice brokers to throw loans their way, lenders scrapped their documentation requirements and slashed their approval times. One company, NovaStar Financial, reportedly sent a brochure to brokers trumpeting, “Did You Know NovaStar Offers to Completely Ignore Consumer Credit!”118

Lenders also turned a blind eye to wrongdoing by brokers. One WaMu senior underwriter reported noticing that a broker was targeting seniors and minorities and fl ipping loans. He informed WaMu’s senior management and said he was going to decline the next loan submitted by the broker. To his dismay, management report- edly rejected his decision on grounds that “the broker gave WaMu a lot of loans.”119

Another executive complained that the subprime lender People’s Choice knew about broker fraud, but “calculated that it would have been too complicated and expensive to go after [it].”120

Lenders set up incentive systems for loan offi cers, rewarding those who met min- imum production goals with lavish vacations and bonuses. One former Ameriquest employee described “the drive to close deals and grab six-fi gure salaries” as lead- ing “many Ameriquest employees astray. They forged documents, hyped customers’ creditworthiness and ‘juiced’ mortgages with hidden rates and fees.” A loan offi cer from Ameriquest reported coming across “co-workers using a brightly lighted Coke machine as a tracing board, copying borrowers’ signatures on an unsigned piece of paper.”121 A former employee at WaMu said “she coached brokers to leave parts of applications blank to avoid prompting verifi cation if the borrower’s job or income was sketchy.”122 In Los Angeles, the FBI uncovered a document forger who created false W-2s, pay stubs, and other documents verifying borrowers’ credit, employment, and identifi cation for use by over 100 mortgage professionals.123

In the frenzy, lenders “lost sight of the basic tenets of underwriting and risk.” A quality assurance manager at one lender described his company as “going through the motions” of risk management during the “free for all to approve loans by the thou- sands.” The most graphic description of the breakdown in risk management came from a former WaMu underwriter, who said, “If you had a pulse, WaMu would give you a loan.”124

It was loan underwriters’ job to review loan fi les for creditworthiness and rule out fraud, but they rarely had enough time to thoroughly review fi les, given the constant

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 39

pressure to approve loans. In addition, lenders increasingly overrode underwriters’ loan decisions by making “exceptions” to the underwriting criteria. Underwriters claimed that when they nixed loans as too risky, senior management would reverse their deci- sions and approve the loans as exceptions. Sometimes refusal to approve a loan would be grounds for disciplinary action or retaliation against an underwriter.125 At New Century, one underwriter told of a salesman who hit her desk with a baseball bat when she “cut” his deal. The same underwriter reported that she was “constantly told ‘If you look the other way and let an additional three to four loans in a day that would mean millions more in revenue for New Century over the course of the week.’”126 Lenders used carrots as well as sticks. At some companies, underwriters who met or exceeded their targets for loan approvals earned bonuses as high as $5,000 per month.127

Some subprime lenders outsourced loan underwriting to contract underwriters for as little as $10 per loan application. With such low fees, it was not cost-effective to verify incomes and evaluate credit risk carefully. As a result, contract underwriters had economic incentives to dispense with careful verifi cation of borrowers’ eligibility for credit and deceive participants down the line about the risks of subprime loans.

In a 2006 review of a sample of loans with early defaults, Fitch Ratings uncovered compelling evidence of a “race to the bottom” in underwriting. Fitch concluded that “in many instances, misrepresentations and altered documentation [were] evident in the physical fi les.” It found that “loan fi les of borrowers with very high [credit] scores showed little evidence of a sound credit history but rather the borrowers appeared as ‘authorized’ users of someone else’s credit.” In other fi les, they found errors in the calculation of borrowers’ debt-to-income ratios, incomes reported in low-doc and no- doc loans that were “unreasonable,” and “substantial numbers of fi rst-time homebuyers with questionable credit/income.” One fi le included an admission by the borrower that he was a “ ‘straw buyer’ in a property fl ipping scheme.” Fitch concluded that “poor underwriting processes did not identify and prevent and, therefore, in effect, allowed willful misrepresentations by parties to the transactions, which . . . exacerbated the effects of declining home prices and lax program guidelines.”128

A 2009 study by the General Accountability Offi ce confi rmed the deterioration in loan quality. With every vintage, the percentage of subprime loans defaulting within three years got worse. For loans originated in 2004, the three-year default rate for loans was 5 percent. For loans originated in 2005 that rate was 8 percent, and for 2006 loans, the rate rose to 16 percent.129

The crowd-out effect also drove out affordable loans designed for borrowers with weak credit. Throughout the subprime boom, the Federal Housing Administration (FHA) and Veterans Affairs (VA) offered safer and cheaper alternatives to subprime loans. However, as subprime loan products proliferated, FHA and VA loans lost favor. From 1999 through 2006, for example, FHA’s market share by dollar volume fell from 7.96 percent to 1.75 percent. VA loans experienced a similar drop. In 2005 alone, the volume of FHA loans fell 39 percent, and VA loans fell 30 percent.130

There were many reasons for this decline. Government loan programs required detailed documentation and down payments and capped the size of eligible loans. FHA and VA loans were slower to process because of paperwork requirements and mandatory inspections. For some borrowers, the loans may have seemed like a hassle. Brokers also had lots to gain from putting borrowers in subprime loans with their high

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

40 • PART I THE SUBPRIME MARKET TAKES OFF

fees and reduced documentation, and lots to lose if they offered FHA and VA loans. The suppression of safe substitutes for subprime loans during the housing bubble was one more indication that subprime loans had crowded out prime loans.131

The crowd-out effect also tilted Fannie Mae and Freddie Mac toward buying increasingly dangerous loans. In 2003, the GSEs dominated the issuance of mort- gage-backed securities, with 78.4 percent of the market. Just two years later, in 2005, their combined market share had plummeted to 44.7 percent, a drop of 42 percent, as lenders sold more and more loosely underwritten loans to the private-label market. In response to this shift and under pressure from Congress to make more loans to low- and moderate-income borrowers, the GSEs made a conscious decision to push deeper into the subprime market, buying ever riskier loans for securitization, including stated-income loans, balloon loans, and loans with high LTVs.132 Fannie Mae also purchased Countrywide’s Fast and Easy low-doc loans. Initially, Countrywide had required Fast and Easy borrowers to have high FICO scores and down payments of at least 10 percent. Over time, however, Countrywide relaxed its requirements for low- doc loans and, despite these changes, Fannie Mae continued to buy them.133 In the words of the former loan-servicing director for Fannie Mae, the company “didn’t know what [it was] buying . . . The system was designed for plain vanilla loans, and we were trying to push chocolate sundaes through the gears.”134

PASS THE TRASH

In theory, lenders and brokers should have shied away from making abusive loans because their reputations and their solvency were at stake. Yet none of these con- cerns proved to be roadblocks. Brokers exploited borrowers and left town when homeowners or enforcement agencies were on their trail. States lacked any strong requirement that brokers be capitalized or bonded, which effectively made brokers judgment-proof.

Lenders should have worried even more that bad lending would hurt their business. After all, lenders had substantial assets that borrowers could go after and reputations to preserve. The risk of lawsuits and damage to reputation didn’t seem to sway sub- prime lenders, however. Until the mortgage crash, even lenders who were embroiled in massive, nationwide lawsuits were able to stay in business. CitiFinancial, Ameriquest, and other major subprime lenders survived well-publicized consumer litigation and managed to keep their lending shops open. And, despite the bad press, borrowers still fl ocked to them for loans, and capital markets continued to fi nance their lending.

The reason subprime lenders did not worry about their solvency was securitiza- tion. Securitization allowed lenders to shift most of the default risk on to investors, who bore the fi nancial brunt if the loans went belly-up. In other words, lenders could “pass the trash.” In the past, when most mortgage lenders were regulated banks and thrifts that held their loans in portfolio, lenders took care when underwriting loans. Back then, default was a serious fi nancial event. With securitization, lenders—both regulated and unregulated—could roll the loans off their balance sheets. As New Century’s bankruptcy examiner explained: “So long as investors continued to be will- ing to purchase New Century loans, New Century apparently did not believe it needed signifi cantly to improve loan quality.”135

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 2 THE EMERGENCE OF THE SUBPRIME MARKET • 41

Securitization also altered the compensation structure of subprime lending. Lend- ers made their money on upfront fees collected from borrowers and the cash proceeds from securitization offerings, not on the interest payments on loans. This gave lenders the security of being paid in advance, instead of having to wait for uncertain monthly payments over the life of loans. As a result, lenders had even less reason to care about how well their loans performed. Instead, all that mattered was generating fees and quickly getting the loans off their books.

Some lenders even had two sets of underwriting standards: high standards for loans they kept on their books and lax standards for ones that they securitized. Researchers have confi rmed the moral hazard problems created by securitization. When lenders sold loans to outside buyers, the loans were more likely to default. This suggests that lenders were more careful about the default risks if they had formal relationships with the purchasers of their loans, for example, if the lenders and purchasers were owned by the same holding company. One of the most compelling studies compared function- ally equivalent subprime loans, some of which fell just below the threshold for securi- tization (a FICO credit score of 620)136 and those with scores just above the threshold. These loans should have had similar default rates. Instead, the authors found that the loans right above the 620 threshold—the ones more likely to be securitized—had higher rates of default. These results were strongest for stated-income loans, where the borrowers’ true incomes were unknown.137 Another study compared loans that were sold through securitization with those held in banks’ portfolios. Overall, the loans that lenders sold on the secondary market did not perform as well.138 An obvious conclu- sion from these studies is that lenders were less concerned with carefully underwriting loans that they knew were going to be securitized.139

WHAT ABOUT THE BORROWERS?

Without a doubt, most borrowers with subprime loans would have been better off with loans on better terms or with no loans at all. It is natural, then, to ask why bor- rowers ever agreed to ridiculously expensive subprime loans.

For one thing, many borrowers lacked full information when choosing subprime loans. Subprime loans were known for their complexity. The loans had multiple mov- ing parts, from interest rates that depended on the interest rate that London banks charged each other to dozens of different fees.

Complexity made it diffi cult for borrowers to determine the true cost of loans, which made it hard to comparison-shop. Plus, lenders and brokers were unwilling to give fi rm price quotes when borrowers were shopping for subprime loans. Federal mortgage disclosures were too outmoded, complicated, and late to help consumers shop or to alert them to the back-end risks of toxic ARMs.140

People fell for abusive subprime loans for another reason: information overload. When people have to evaluate deals with multiple features, they tend to focus on one or two features. There is simply too much information to consider all the variables. In the case of subprime loans, people often concentrated on the initial interest rate or the initial monthly payment. Behavioral economists call these rules of thumb “choice heu- ristics.” Choice heuristics can lead consumers to make decisions that are not in their best interests. Subprime lenders and brokers exploited these behavioral tendencies by

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

42 • PART I THE SUBPRIME MARKET TAKES OFF

offering complex products to borrowers and then directing them to consider only the initial monthly payment: was it more or less than what they were paying and could afford? Many borrowers, suffering information overload, would latch on to the initial monthly payments for exotic ARMs in the mistaken belief that lenders and brokers were acting in their best interests.141 What they did not know, though, was that their monthly payments could ultimately soar.

Overoptimism was another reason that borrowers took out subprime loans. People tend to look on the bright side. This can lead them to misjudge their ability to afford future increases in their mortgage payments. People also place excessive importance on immediate gains while discounting potential future losses. In the mortgage context, empirical studies have found that when consumers look at a loan amount and payment schedule, they tend to underestimate the cost of the interest rate they will pay.142 This explains why people might have preferred loans that put cash in their pockets over no-cash-out loans with lower interest rates and fees that would have reduced the odds they would lose their homes.143

Whether you call it predatory, nontraditional, Alt-A, nonprime or subprime lend- ing, making loans with terms that borrowers could not understand or afford was a recipe for disaster. It was with these loans that the subprime virus took off.

FIGURE 2.3.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

3

a A Rolling Loan Gathers No Loss

If lenders had kept their subprime loans on their books, they probably would have made fewer loans and taken greater care with the ones they made. With securitiza- tion, however, they could write risky loans and shed them quickly for cash. The buyers, mostly Wall Street banks, converted the loans into securities and passed the risk onto investors. That risk then “went viral” with the creation of trillions of dollars in complex credit derivatives built on subprime loans. Commercial banks, investment banks, insurance companies, hedge funds, pension plans, and governments around the world bought subprime derivatives, which depended on one thing: the timely payment of U.S. subprime mortgages. When that edifi ce cracked, the structure of private-label securitization came tumbling down.

Securitization did not even appear in Webster’s dictionary until 1981.1 Although the word is still in its infancy, most people in the United States sense that securitiza- tion had something to do with the subprime crisis. This chapter introduces readers to the process of securitization and to the actors who helped convert mortgage loans into complex fi nancial instruments. It then explains how that process went haywire despite warnings to the market and the government years in advance.

A THUMBNAIL SKETCH OF SECURITIZATION

At its core, securitization is rather simple. Investors provide lenders with capital to make mortgages or other loans. In return, the investors receive bonds backed by the loans.

During the credit boom, there were two main branches of residential mortgage securitization: agency securitization and private-label securitization. Agency securi- tization refers to the securitization of loans by Fannie Mae and Freddie Mac. His- torically, loans meeting Fannie’s and Freddie’s requirements for purchase were called conforming loans. Private-label securitization fi nanced non-conforming loans, such as subprime loans and jumbo prime loans, and was orchestrated by Wall Street fi rms. In this chapter, we will primarily be concerned with private-label securitization.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

44 • PART I THE SUBPRIME MARKET TAKES OFF

When a lender made a subprime loan with an eye to securitization, typically it sold the loan to an investment bank, whose function was described as the arranger, sponsor, or underwriter. We use the term arranger in this book. The job of the arranger was to convert loans into securities.

In a typical securitization, the arranger bundled a group of loans into a pool and created bonds out of the loan pool. These bonds were called residential mortgage-backed securities (RMBS) because they were backed by collateral consisting of the loans in the loan pool. Rating agencies then rated the bonds based on the expected likelihood of default. Once the arranger had the ratings in hand, it priced the securities and read- ied them for sale. Ultimately, the loans were transferred to a special purpose vehicle (SPV ), which usually took the form of a trust. The SPV insulated the loans from seizure by the lender’s creditors in case the lender became insolvent. Then the trust issued the securities and sold them to investors through broker-dealers, who often were affi liates of the arranger. Other entities, the servicers, did the heavy lifting of collecting the payments from borrowers, paying taxes and insurance, forwarding funds for disbursement to investors, and managing the loans. A fi nal entity, the trustee for the trust, was typically some other Wall Street investment bank that sent the payments on the bonds to the investors.

Arrangers Arrangers of private-label securitizations were usually investment banks. In 2007, Lehman Brothers, Bear Stearns, Morgan Stanley, and JPMorgan Chase were the top

BORROWERS

SPV/TRUST

INVESTORS Sells bonds backed by the mortgages

LENDER

ARRANGER

Sells loans

SERVICER

TRUSTEE

Forwards principal and interest payments

Make monthly paymentsMakes loans in

exchange for notes

Passes along payments

FIGURE 3.1. Flowchart of securitization.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 45

four arrangers of private-label securitizations. In a few cases, such as Countrywide, lenders directly issued their own securitization offerings.

Arrangers needed a constant fl ow of loans to generate securitization fees. One way they achieved this was by buying subprime lenders outright. Another way was by pro- viding lenders with warehouse lines of credit to fund the loans they made. In return, the lenders granted the arrangers the right to purchase the loans. These warehouse lines were signifi cant. At its peak, New Century had over $14 billion in warehouse lines from banks and other sources.

Arrangers also struck “selling forward” deals in which they agreed to buy loans from lenders, subject to stipulations, before the loans were made. For example, a lender might agree on October 1 that on December 1 it would deliver 2,000 adjustable-rate mortgage loans with an average interest rate of 6 percent, of which half would be subject to a prepayment penalty. The sales price was usually the face value of the loans plus a few percentage points. Based on stipulations regarding the characteristics of the loans, an arranger might agree to pay 102.25 percent of the original loan principal. In other words, the purchaser would agree in advance to pay the amount of the principal plus 2.25 percent. If the loans actually delivered had a slightly higher or lower average interest rate, the stipulations would specify an adjustment to the premium that was added to the principal.

Lender Representations and Warranties When arrangers purchased loans, they required lenders to provide them with repre- sentations (reps) and warranties about the quality of the loans. These reps and war- ranties included assurances that the loans complied with state and federal laws and satisfi ed stated underwriting criteria. To further sweeten the deals, lenders agreed to recourse clauses, in which they promised to buy back loans that violated the reps and warranties or that went into early default. Recourse clauses lulled some investors into believing that they did not have to carefully review the actual loans. After all, the thinking went, if a loan violated the reps and warranties the lender would have to repurchase it.

Reps and warranties were only as good as the promises they made. For example, failing to confi rm a borrower’s income did not violate the reps and warranties unless the lender specifi cally stated that it verifi ed borrowers’ incomes.

Due Diligence Arrangers did not rely on reps and warranties alone when buying loans. To assuage investors and ensure that the securities would receive the highest, investment grade ratings, arrangers commissioned due diligence reviews of loan pools. Generally, they hired outside due diligence fi rms that would review a sample of loans in a pool to confi rm that the loans met the lenders’ underwriting standards and procedures. Ideally, due diligence fi rms also verifi ed information on borrowers’ applications and made sure that all the loan documents were in order. These fi rms would then report their fi ndings to arrangers for use in securities disclosures. This due diligence review was the only time that individual loan fi les normally received outside scrutiny during the securitiza- tion process.

With the growth of subprime lending and exotic new loan products, arrangers should have intensifi ed their due diligence reviews. Instead, due diligence became

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

46 • PART I THE SUBPRIME MARKET TAKES OFF

perfunctory. During the subprime bubble, the quality of due diligence reviews declined, and so did the number of loans that underwent review. In 1995, due diligence reviewers sampled up to 30 percent of the loans in a loan pool. In 2005, arrangers were instructing due diligence fi rms to review only 5 percent.2

Other developments around the same time contributed to the decline in the quality of due diligence reviews. With the emergence of no-doc and low-doc loans, borrow- ers’ loan fi les often did not contain paystubs or tax returns proving their income. In those cases, due diligence fi rms could not verify borrowers’ stated incomes based on the loan fi les.

Due diligence also took on a new meaning. In the past, due diligence had focused on the quality of loans, including borrowers’ ability to repay their loans. As subprime lending accelerated, the inquiry narrowed to the single question: did the loan adhere to the lender’s guidelines? If the guidelines did not require a hard-nosed assessment of a borrower’s ability to repay, then the due diligence review did not look at that issue.3

Like others in the securitization food chain, the people who conducted due dili- gence reviews were under constant pressure to sign off on dubious loans. For example, one employee of the due diligence fi rm Watterson-Prime, whose chief client was the now-defunct Bear Stearns, claimed that she rejected loans only to be overruled by her supervisors. According to the employee, 75 percent of the loans she rejected ultimately made it into subprime loan pools. When she showed her supervisors loan fi les that lacked proof that the borrowers made the incomes reported on their applications, her supervisors responded, “Oh, it’s fi ne. Don’t worry about it.”4

A key feature of any due diligence report was information on the number of loans that fl unked the underwriting criteria specifi ed in the lender’s reps and warranties. These loans were called “exceptions” and were more likely to default for obvious reasons. By 2006 and 2007, in some securities offerings, the number of exceptions exceeded the number of loans that met the lender’s underwriting standards. In some deals, up to 80 percent of loans were exceptions.

Lenders came up with various justifi cations for deviating from their underwriting standards. For example, New Century allowed exceptions for people who demon- strated pride of ownership by keeping their homes in better shape than their neigh- bors.5 When due diligence fi rms brought high exception rates to arrangers’ attention, their concerns were often dismissed. One executive at a major due diligence fi rm com- plained that people at his company felt like “potted plants” as they watched investment banks scoop up mortgages that fl unked due diligence reviews.6

Lenders had strong incentives to stuff in exception loans with “good” loans that were destined for securitization, and so did arrangers. First, passing the trash made it easier for all concerned to satisfy investors’ demand for subprime RMBS. Otherwise, they had to fi nd a replacement for every loan they rejected from a pool. Another reason arrangers threw rotten apples in with the good had to do with their contracts with lenders. It was “a generally accepted practice” for arrangers and lenders to negotiate a cap on the percentage of loans that an arranger could return to a lender.7 New Century, for example, extracted promises from investment banks to reject no more than 2.5 percent of the loans they purchased.8 These agreements provided lenders like New Century with incentives to slip bad loans into pools.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 47

Arrangers counted on investors and rating agencies not to notice suspect loans. It helped that arrangers kept the due diligence reports mum. As we write this book, the New York attorney general, with the cooperation of a leading due diligence fi rm, is investigating reports that investment banks withheld damaging information in due diligence reports from rating agencies and investors. Allegedly, “Some Wall Street fi rms concealed information about exceptions . . . in a bid to bolster ratings of mort- gage securities and make them more attractive to buyers.”9

Structuring the Bonds

Purchasing and reviewing loans were just two of the tasks of arrangers. Their main job was carving the principal and interest payments from borrowers into tranches (which is French for slices). Each tranche had its own bond, with its own yield, maturity date, and level of risk. A single deal could have over twenty tranches. The top tranche was the safest, with the lowest interest rate, and was paid off fi rst. The tranche right below the senior tranche was paid off next and had a slightly higher risk with a slightly higher interest rate. And so it went down the line to the last tranche, the junior tranche or equity tranche. The equity tranche was the last to be paid, offered the highest inter- est, and was the fi rst to absorb losses if borrowers defaulted.

The rating agencies assigned ratings to each of the tranches, using a unique combi- nation of upper- and lower-case letters. For example, Standard & Poor’s usually gave the top tranche a AAA rating; Moody’s used Aaa. The next highest tranche received an AA, while the medium risk (mezzanine) tranches were ranked A to BBB. Any tranche rated BB or lower was below investment grade. The equity tranche was not rated at all. Often lenders owned the equity tranche. The theory was that if lenders held the riskiest tranches, they would exercise more caution when underwriting loans. However, as we discuss later on, lenders fi gured out how to hedge the equity tranches or sell them for resecuritization and unload that risk.10

Arrangers had other responsibilities, too. They had to create the trusts that would hold the assets and issue the securities. They drafted prospectuses and offering mem- oranda that informed investors about lenders’ underwriting criteria and the risks asso- ciated with the securities. They also made other needed fi lings with the Securities and Exchange Commission (SEC) and ensured that all of the necessary regulatory approvals were obtained.

Arrangers made millions of dollars on every deal they put together. Underwriting fees were usually at least 1 percent of the value of the collateral and deals were in the billions of dollars. To get a sense of the total magnitude of this compensation, arrangers underwrote $2.1 trillion in subprime mortgage-backed securities between 2000 and 2007. One percent of $2.1 trillion is a very large number.11

Rating Agencies Arrangers worked closely with the three big rating agencies, Standard & Poor’s (“S&P”), Moody’s, and Fitch, whose task it was to grade each tranche based on the credit risk associated with that security. In rating mortgage-backed securities, the agencies relied on data from arrangers about the loans and the borrowers. These data included borrowers’ credit scores, loan-to-value ratios, whether the borrowers docu- mented their incomes, whether the properties were owner-occupied, and whether the

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

48 • PART I THE SUBPRIME MARKET TAKES OFF

loans were used to refi nance an existing loan or to buy a home. The rating agencies also reviewed the lenders’ reps and warranties and their reputations. Using this informa- tion, the agencies calculated the default risk for each of the tranches.

In rating RMBS, the rating agencies lacked one key source of information: the due diligence reports. The agencies maintained that “due diligence duties belonged to the other parties in the process.”12 On some occasions, rating agencies did request due diligence reports from arrangers, but were turned down.13

The SEC designated all three rating agencies Nationally Recognized Statistical Rating Organizations (NRSROs). For arrangers, garnering top ratings on mortgage- backed securities from one or more of these NRSROs was critically important. They prized a rating agency’s NRSRO status because state and federal laws prohibited banks, insurance companies, and pension plans from investing in securities that did not have investment grade ratings from an NRSRO.14

The law conferred another important benefi t on rating agencies that other mar- ket participants like securities analysts and accounting fi rms did not enjoy. They had immunity from broad swaths of legal exposure under the First Amendment and fed- eral securities law. Practically, this meant that rating agencies could evaluate risks, give opinions on those risks through ratings, and not have to answer for their errors in judgment.15

No bond issue could earn a AAA rating without protections known as credit enhancements. Credit enhancements were supposed to minimize risk to investors by creating a cushion to absorb losses in the event of widespread defaults. These enhance- ments together with strong ratings were designed to give investors confi dence to invest in the securities. Credit enhancements included overcollateralization, where the total balance of the principal of the loans exceeded the outstanding principal balance on the securities, and excess spread, where the interest payments from borrowers exceeded the interest owed to bondholders.

Bond insurance was another common form of credit enhancement. Issued by mon- oline insurers, such as Ambac and MBIA, bond insurance kicked in if the mortgage payments on a pool of loans fell too low to pay the trust’s obligations to investors. The top bond insurers had vaunted AAA ratings and were considered “good for the money.” At least one major bond insurer reportedly never expected to have to pay out on its policies and did not review the quality of the mortgages backing the securities it insured, something that only came to light in 2009.16

For many years, rating agencies functioned in the shadow of subprime securitiza- tion. Their role in the crisis did not become apparent until journalists and regulators began to dig deeper. Startling evidence of poor judgment eventually came to light. For example, from 2000 through 2006, Standard & Poor’s took the position that piggy- back loans—those with simultaneous second mortgages—were just as safe as loans in which borrowers had equity in their homes. This defi ed common sense. When S&P later discovered that piggyback loans were 43 percent more likely to default than other loans, it still did not downgrade securities backed by piggyback loans, saying it “had to further monitor the performance of loans.”17

Speed was the name of the game at the peak of subprime. This encouraged less-than-careful work by rating agency analysts. Inundated with requests to assess new issues from arrangers, the agencies could not keep up. Sometimes the deadlines

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 49

were so tight that analysts only had one day to evaluate submissions from arrangers.18

According to the SEC, an analytical manager at one fi rm warned in an internal email: “We ran our staffi ng model assuming the analysts are working 60 hours a week and we are short on resources. . . . The analysts on average are working longer than this and we are burning them out.”19

The problems at the rating agencies were not limited to fl awed assumptions and overwork. An even bigger issue was their compensation system. In the past, rating agencies made money by selling their ratings to investors for use in their investment decisions. Having investors, not arrangers, pay for ratings was key to rating agencies’ objectivity because, as a former Moody’s vice president said in 1957, “We obviously cannot ask payment for rating a bond. To do so would attach a price to the process, and we could not escape the charge . . . that our ratings are for sale.”20

Despite this admonition, in the 1970s the SEC began allowing arrangers to pay the rating agencies for rating the securities they were underwriting. Suddenly, the agencies were beholden to Wall Street for revenue.21 An employee from Standard & Poor’s structured fi nance division captured this mentality in an email saying that a deal “could be structured by cows, and we would rate it.”22

The rating of a mortgage-backed security garnered fees that were four or fi ve times the fees from rating a municipal bond offering. These fees propelled Moody’s average profi t margin to 53 percent.23 From 2000 to 2006, Moody’s had a 375 per- cent increase in profi ts, and its stock value increased fi vefold.24 The compensation

FIGURE 3.2.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

50 • PART I THE SUBPRIME MARKET TAKES OFF

structure, along with arrangers’ ability to shop for ratings, led to ratings infl ation. Eighty percent of securities backed by subprime mortgages were rated AAA, and 95 percent were rated A, AA or AAA.25 As one Moody’s employee reportedly wrote in an email, the ratings suggested incompetence or that the fi rm had “sold [its] soul to the devil for revenue.”26

Rating agencies snared business by satisfying their customers. In doing so, they did put their ratings up for sale. According to New York Times reporters, rating agen- cies were collaborators who worked “behind the scenes, with the underwriters that were putting . . . securities together.” The reporters explained that underwriters didn’t “assemble a security out of home loans and ship it off to the credit raters to see what grade” it would get. Rather, underwriters “work[ed] with rating companies while designing a mortgage bond or other security.”27 The rating analysts would advise the arrangers “how to structure the bonds to achieve maximum triple-A ratings.”28 The agencies also provided software that originators, investment banks, investors, mort- gage insurers, and other entities could use to get a sense of what it would take for a deal to receive an investment grade rating.29 Arrangers could plug various numbers into the software until they came up with a structure that would generate a AAA rating for the top tranche. This practice enabled investment banks to “game” the system.30

Arrangers wielded a great deal of power. The rating agencies needed their business, and most of that business came from a handful of arrangers. In the RMBS market, a dozen arrangers were involved in 90 percent of the deals.31 According to industry insiders, arrangers would shop among the rating agencies to see who would give them the best rating on their securities. This practice was known as “maximizing value” or “best execution.”32 In the drive to secure business, fi rewalls designed to insulate rating agency analysts from infl uence broke down. An SEC report revealed internal emails from one rating agency suggesting that analysts should consider the impact of changes to their ratings methodology on the fi rm’s market share. The same report told of ana- lysts being involved in discussing the agency’s fees with arrangers and sometimes even participating in fee negotiations. 33

Arrangers actively lobbied analysts, hoping to infl uence their ratings. Raymond McDaniel, the chief executive offi cer of Moody’s, reportedly described this phenom- enon to his board of directors in a memo discussing “ratings erosion by persuasion.” McDaniel explained that analysts and managing directors were “continually ‘pitched’ by bankers, issuers, investors . . . whose views can color credit judgment, sometimes improv- ing it, other times degrading it (we ‘drink the kool-aid’).”34 Drinking the Kool-Aid often meant “adjusting the criteria . . . because of the ongoing threat of losing deals.”35

To their credit, some rating analysts refused to succumb to pressure from arrangers. Such refusals could generate calls from irate arrangers who would ask the agencies to assign a new analyst to rate the deal. The Wall Street Journal reported that clients of Moody’s who complained about the conclusions Moody’s analysts reached on deals, at times, had their reviews transferred to other analysts. Countrywide had a reputation of complaining to Moody’s that its assessments of Countrywide’s issues were “too tough.” Moody’s response was to “soften[] its stance on Countrywide securities . . . even though no new and signifi cant information had come to light.”36

In 2008, the SEC investigated rating agencies’ processes and found numerous problems. The agencies used ratings criteria that they had never published. They failed

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 51

to document their policies and procedures for rating subprime bonds, and had no method for detecting errors in their models. One of the more alarming fi ndings was that the agencies sometimes deviated from their own models in a practice known as “ ‘out of model’ adjustments” when issuing ratings. These adjustments tamped down the losses that the agencies’ models otherwise would have predicted. When the SEC asked about these deviations, the staff at the agencies was not always able to offer an explanation. The fate of billions of dollars turned on these inexplicable ratings.37

As authors of a New York Times op-ed put it, “In pursuit of their own short-term earnings,” the rating agencies “did exactly the opposite of what they were meant to do: rather than expose fi nancial risk they systematically disguised it.”38

Selling the Securities Once the arranger worked out the details of a securitization, it parked the loans in an SPV that was the actual issuer of the securities. Broker-dealers then stepped in and marketed the securities to investors. Selling mortgage-backed securities was a lucrative business, with revenues consisting of a cut of the sales proceeds in the form of dis- counts, concessions, or commissions.39

It was not unusual for broker-dealers to hold some of the securities, either as invest- ments or because there was no market for a particular tranche. Sometimes investors rejected the top tranche because the yield was too low. At other times, arrangers like Merrill Lynch held on to senior tranches in the mistaken belief that the tranches were “shielded from falls in the prices of mortgage securities.”40 Citigroup made the same mistake. In 2008, Citigroup ended up with $20 billion in senior mortgage-related securities on its books even after taking write-downs on some of those securities of up to 80 percent.41 Both companies suffered grievous losses as a result.

Servicers Servicers were key players in securitizations. They collected and processed borrowers’ mortgage payments and passed the monthly loan proceeds on to the trust after taking out their servicing fees and any escrow payments for real estate taxes and insurance. The trustee then distributed the principal and interest payments to the investors.

Loan servicers got a nice piece of the subprime pie. Some of their revenue consisted of a cut of the interest that borrowers paid on loans. For prime, fi xed-rate loans, this was usually 0.25 percent of the loan amount. For subprime loans, it was usually 0.50 percent. ARMs garnered even higher fees because servicers had to process the various rate adjustments over time. For FHA loans, VA loans, and subprime loans, the ser- vicers’ cut was also substantial because these loans were more likely to go into default and required more servicing.

Servicers had other sources of revenue as well. They were allowed to retain late fees and received additional compensation if they met performance goals for loss mitiga- tion. These fees were not benign. They created an incentive for servicers to delay post- ing payments and employ aggressive tactics to get borrowers to come up with money for their loan payments.42

CDOs and Their Cousins Residential mortgage-backed securities were simple compared to what came next. Wall Street began concocting collateralized debt obligations (CDOs), which involved

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

52 • PART I THE SUBPRIME MARKET TAKES OFF

pooling tranches of bonds and converting them into new securities. Drexel Burnham Lambert invented CDOs in the late 1980s as a way to unload corporate bonds. Only later did the model expand to mortgage-backed securities. In a process that paralleled the securitization of mortgages, arrangers constructed CDOs by taking lower-rated tranches of mortgage-backed securities (including junk RMBS), pooling those tranches, and dividing the pool into a new set of tranches for sale to investors. Alter- natively, arrangers built “synthetic” CDOs out of sellers’ obligations to pay on credit default swaps on subprime bonds.

Rating agencies reviewed the CDOs and gave the senior tranches AAA ratings. Just as with mortgage-backed securities, the top tranche carried the least risk and offered a lower yield than the junior tranches. For CDOs, the rating process was somewhat different than for RMBS. A CDO’s assets were actively managed and were constantly changing. As a result, rating fi rms could not base their ratings on the assets that were actually backing a CDO. Instead, they reviewed the restrictions on the col- lateral the CDO was permitted to hold and used this information to make judgments about the risks associated with each tranche of the CDO.43

The complexity ran riot with the resecuritization of CDO tranches. CDOs were pooled and tranched into CDOs2 and CDOs2 were resecuritized into CDOs3. The astonishing thing about CDOs (whether they were plain, squared, or cubed) was that the underlying bonds could be junk, yet the top tranche of any CDO could carry a AAA rating. Essentially, CDOs purported to make steak out of chicken. Arrangers pooled tranches of mortgage-backed securities with low or no ratings (the chicken) and sliced those pools into tranches, with the top tranche earning a AAA rating (the steak).

FIGURE 3.3. Drawing by Kagan McLeod.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 53

It was too good to be true. Just because a CDO had a AAA rating didn’t mean the assets backing it were top fl ight. Up to 80 percent of CDOs had AAA ratings, even though as many as 70 percent of the underlying RMBS had ratings below AAA.44 All a AAA rating meant was that the AAA tranche would get paid fi rst. Being fi rst in line is irrelevant, however, if there is nothing to get. That is exactly what happened. When droves of subprime mortgages went into default, the most junior mortgage-backed securities received no payments and neither did the CDOs containing these tranches, regardless whether those CDO tranches were rated AAA or D. CDOs had other problems. Some CDOs were formed and sold before arrangers had assembled all the collateral that would go into the pool. This meant that investors had to trust arrangers’ assurances about what would go into the pools, with no ability to verify the collateral before deciding to invest.45

Lack of diversifi cation was another issue. Some pools contained as few as twenty assets. And even CDOs with hundreds of different assets could lack diversity if most or all of the assets came from the same sector, such as residential mortgage-backed securities. This was not a hypothetical concern. According to one report, RMBS grew from making up 43.3 percent of CDO portfolios in 2003 to 71.3 percent in 2006. In some cases, RMBS represented 90 percent of CDOs’ portfolios. So much for CDOs as risk-diversifying instruments.46

Mortgage-related CDOs and their squared and cubed cousins were scarce until the mid-1990s, when JPMorgan Chase dove into the market. At the top of the market in 2006 and 2007, banks issued over $200 billion worth of CDOs backed by risky mortgage-backed securities.47 The underwriting fees were astronomical, with some arrangers garnering 2.5 percent of the value of a CDO offering. Citigroup alone issued $20 billion worth of CDOs in 2005. Merrill Lynch bested Citi in 2007, creating more than $30 billion in mortgage-backed CDOs in just seven months.48

Investors fl ocked to CDOs. During the housing bubble, hedge funds bought over 45 percent of CDOs; insurance companies, banks, and asset managers held most of the rest. As would be expected, banks held the top-rated tranches, while hedge funds preferred the equity tranches.49

SIVs Banks were not keen on keeping RMBS and CDOs on their balance sheets, so they found a way to off-load these risky holdings through entities called struc- tured investment vehicles (SIVs). With an SIV, a bank could sell its mortgage- backed securities to its SIV, which became the actual owner of the securities. To pay for the securities, the SIV issued commercial paper—a fancy way of saying it borrowed money, often from money market funds—at low rates and for short terms. The SIV ’s RMBS and CDOs served as the collateral for the loans. If the return on the SIV ’s assets exceeded the interest rate on its loans, the SIV made a profit.50

SIVs were a bright spot for a time. Citigroup fi rst introduced SIVs in the late 1980s, and by 2007, banks had created over twenty SIVs with total assets ranging from $350 to $400 billion. The business model of SIVs was inherently unstable. The SIVs’ securities had lengthy maturities, but their commercial paper came due in less than nine months and sometimes in only a few days. To keep SIVs going, managers had to

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

54 • PART I THE SUBPRIME MARKET TAKES OFF

constantly issue new commercial paper to refi nance the old. Consequently, SIVs’ suc- cess depended on their liquidity, and their liquidity depended on the strength of their assets. If they could not roll over their commercial paper because of faulty collateral, the SIVs would inevitably fail.51

Because SIVs were new on the block, there was no history to consult in predict- ing their future performance. This concerned investors who bought SIVs’ commercial paper. They worried that if an SIV’s RMBS and CDOs started defaulting, the SIV would not be able to pay back its loans. This was a risk that investors were unwilling to take. The solution was to extract informal promises from banks to make good on their SIVs’ obligations if the SIVs’ collateral failed to perform. The SIVs implemented these promises with guarantees, called “liquidity puts,” issued to their investors. Despite these promises, in a seemingly masterful but ultimately disastrous sleight of hand, the banks did not book the SIVs’ debts on their balance sheets when they made their promises.52

Credit Default Swaps The proliferation of credit derivatives did not stop with CDOs and SIVs. The risk that RMBS and CDOs might default spawned another product called credit default swaps (CDS). CDS were a tool for hedging default risk. A swap purchaser—usually a bank or other investor—would buy swap protection from a swap seller that would cover any losses if a bond covered by the swap defaulted. Investors liked swaps because swaps limited their exposure to default risk (so long as their swap seller was solvent if and when they needed to collect). Swap sellers were willing to provide swaps because the swap premiums were lucrative and sellers thought it was unlikely they would ever have to pay out on the swaps.53

In addition to their hedging function, swaps were used for speculation. Investors bought and sold swaps as bets on the performance of bonds. In cases of pure specula- tion, neither the buyer nor the seller actually owned the bond in question. Instead, they were betting on the performance of a bond held by someone else. Nothing limited the number of credit default swaps referencing a single security. For example, bonds valued at $1 million could be the basis for a hundred bets totaling $100 million on the performance of those bonds. The dominant seller of credit default swaps on mortgage bonds was American International Group (AIG).

We could go on to describe further complexities in fi nancial products. The point, however, is not to inundate readers with descriptions of fi nancial products, but to show how securitization, in the words of law professor Kurt Eggert, was “able to spin endless amounts of Wall Street gold . . . out of even the most suspect and speculative straw.”54

The Quants Mathematical wizards, known as “quants,” made subprime mortgage-backed securi- ties possible by developing complex algorithms that calculated the risk of RMBS, CDOs, CDOs2, and CDOs3 as well as swaps. Over time, modeling complex fi nancial instruments became a religion on Wall Street. As a “recovered” Wall Street model builder wrote, “Throw some epsilons and thetas on a paper, hoist a few Ph.D.’s behind your name, and now you’re an expert in divining the future.”55 These models helped arrangers fi gure out how to structure deals, assisted rating agencies in deciding what grade to assign each tranche and aided swap dealers in pricing swaps.56

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 55

Not everyone believed in the quants’ models. Long before the crash, Warren Buf- fett famously warned: “Beware of geeks bearing formulas.” Even earlier, after the col- lapse of Long-Term Capital Management, a 1998 Merrill Lynch memo reportedly cautioned that fi nancial models “may provide a greater sense of security than war- ranted; therefore reliance on these models should be limited.” Merrill Lynch later failed to heed its own advice. The skeptics were right: the geeks’ formulas were deeply fl awed.57

Quants often made mistakes in their assumptions when designing models. Some assumed that historically low mortgage default rates would continue. Many did not even consider the possibility that subprime loans might have higher default rates than prime loans. And, few models incorporated the possibility that a fi nancial cataclysm would increase default risk. The thinking was that because there had been no recent crises, the coming years would be free from crises too.58

Even when quants factored past market crashes into their risk models, the mod- els were not fail-safe. They did not always include extreme, “tail” events. As one risk consultant put it, “Historic[al] data only has rainstorms and then a tornado hits.”59 The tail event that many, including modelers at Citigroup, did not consider was the risk that houses could lose value nationwide. Economists know that even tiny mistakes in assumptions can dramatically skew the predictive power of mod- els. The risks from mistaken assumptions were particularly potent in the mortgage sector, where the effect of each mistake was compounded whenever a security was resecuritized.60

Another problem for the quants was the newness of subprime lending. Quants had plenty of information on the performance of prime loans but little information on the performance of subprime mortgages. As lenders devised new types of loans, the quants had to estimate new risks with no historical experience. No one who was honest could be confi dent about the quants’ predictions during the infancy of subprime. This did not prevent the rating agencies from awarding AAA ratings to securities backed by subprime mortgages.

As the subprime market matured, quants did have opportunities to revise their models to account for new information on subprime loan performance. That did not always happen. For example, rating agency quants reportedly used data from 2001 through 2003 (when losses were 6 percent) to predict the risk on subprime loans made in 2006. The rating agencies’ sloppiness could not have come at a worse time— just when new subprime products were entering the market and borrower quality was declining.61

Rating agencies also could have used default and other performance metrics to adjust their ratings on the securities that were already on the market. These adjustments—rating upgrades or downgrades—were infrequent and usually too late. In fact, agencies often neglected to monitor the performance of subprime securities unless they had some reason to know the securities were in trouble. As a consequence, investors did not have access to timely information about the quality of previously issued bonds.

Frank Raiter, the former managing director and head of the residential mortgage- backed ratings group at Standard & Poor’s, explained the situation in testimony before the House Committee on Oversight and Government Reform:

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

56 • PART I THE SUBPRIME MARKET TAKES OFF

The stress for profi ts and the desire to keep expenses low prevented us from in fact developing and implementing the appropriate methodology to keep track of the new products.

As a result, we didn’t have the data going forward in 2004 and 2005 to really track what was happening with the subprime products and some of the new alternative-payment type products. And we did not, therefore, have the ability to forecast when they started to go awry. As a result, we did not, by that time, have the support of management in order to implement the analytics that, in my opinion, might have forestalled some of the problems we’re experiencing today.62

Lastly, the quants’ models could be gamed. As we mentioned earlier, the arrang- ers ran potential deals through rating agency models to see what structures would generate the best rating at the least cost. Law professor Frank Partnoy wrote about this in the context of CDOs, saying, “The process of rating CDOs [became] a mathematical game that smart bankers [knew] that they [could] win. A person who under[stood] the details of the model [could] tweak the inputs, assumptions, and underlying assets to produce a CDO that appear[ed] to add value, though in reality it [did] not.”63

Arrangers as Market Makers It is easy to view investment banks and other arrangers as mechanics who simply oper- ated the machinery that linked lenders to capital markets. In reality, arrangers orches- trated subprime lending behind the scenes. Drawing on his experience as a former derivatives trader, Frank Partnoy wrote, “The driving force behind the explosion of subprime mortgage lending in the U.S. was neither lenders nor borrowers. It was the arrangers of CDOs. They were the ones supplying the cocaine. The lenders and bor- rowers were just mice pushing the button.”64

Behind the scenes, arrangers were the real ones pulling the strings of subprime lending, but their role received scant attention. One explanation for this omission is that the relationships between arrangers and lenders were opaque and diffi cult to dis- sect. Furthermore, many of the lenders who could have “talked” went out of business. On the investment banking side, the threat of personal liability may well have discour- aged people from coming forward with information.

The evidence that does exist comes from public documents and the few people who chose to spill the beans. One of these is William Dallas, the founder and former chief executive offi cer of a lender, Ownit. According to the New York Times, Dallas said that investment banks pressured his fi rm to make questionable loans for packaging into securities. Merrill Lynch explicitly told Dallas to increase the number of stated- income loans Ownit was producing. The message, Dallas said, was obvious: “You are leaving money on the table—do more [low-doc loans].”65

Publicly available documents echo this depiction. An annual report from Fremont General portrayed how Fremont changed its mix of loan products to satisfy demand from Wall Street:

The company [sought] to maximize the premiums on whole loan sales and secu- ritizations by closely monitoring the requirements of the various institutional

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 57

purchasers, investors and rating agencies, and focusing on originating the types of loans that met their criteria and for which higher premiums were more likely to be realized.66

In a 2008 lawsuit against Countrywide, the State of California made similar allegations:

In order to maximize the profi ts earned by the sale of its loans to the secondary market, Countrywide’s business model increasingly focused on fi nding ways to generate an ever larger volume of the types of loans most demanded by investors. For example, Countrywide developed and modifi ed loan products by discussing with investors the prices they would be willing to pay for loans with particular characteristics (or for securities backed by loans with particular characteristics), and also would receive requests from investors for pools of certain types of loans, or loans with particular characteristics. This enabled Countrywide to determine which loans were most likely to be sold on the secondary market for the highest premiums. . . .

The information regarding the premiums that particular loan products and terms could earn on the secondary market was forwarded to Countrywide’s pro- duction department, which was responsible for setting the prices at which loans were marketed to consumers.67

To ensure a constant supply of loans to feed their securitization machines, arrang- ers bought subprime lenders and made them captive. Ultimately, many investment banks had vertically-integrated production factories with lenders, servicers, and insur- ers. The only piece of the action that arrangers wanted no part of was the retail side of mortgage originations. They believed that by having a layer—mortgage brokers— between borrowers and their fi rms, they could eliminate their exposure to fair lending and consumer protection claims.68

Working in tandem, investment banks extended credit to their affi liates to make loans, purchased those loans for securitization, put securitization deals together, bought and sold the securities through their broker-dealer arms, and serviced the loans.69 Bear Stearns, for example, had a Web-based platform that allowed mort- gage brokers to search loan types and prices, submit loan applications, and obtain automated approvals. A Bear Stearns affi liate then took the loans and put them into securitization deals. The company’s broker-dealers sold the securities and Bear Stearns’s servicing arm, EMC Mortgage, collected the borrowers’ monthly payments.70

Bear Stearns was not alone. Lehman Brothers owned numerous wholesale mort- gage companies, including BNC Mortgage and Finance America, as well as a servicer, Aurora Loan Serving. In 2006, Morgan Stanley acquired Saxon Capital, a servicer and lender. Commercial banks also pursued vertical integration strategies, buying up and merging with loan servicers, originators, and broker-dealers. Even hedge funds, in a departure from their usual mode of operation, adopted vertical integration. For example, Cerberus Capital Management and Fortress Investment Group owned lend- ers and servicers.71

The following excerpt from a Morgan Stanley prospectus gives a sense of fi rms’ tentacle-like components:

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

58 • PART I THE SUBPRIME MARKET TAKES OFF

The sponsor is Morgan Stanley Mortgage Capital Holdings LLC, a New York limited liability company (“MSMCH”), successor-in-interest by merger to Morgan Stanley Mortgage Capital Inc. MSMCH is an affi liate, through com- mon parent ownership, of Morgan Stanley Capital Services Inc., the interest rate swap provider and interest rate cap provider, and Morgan Stanley & Co. Incorporated, the underwriter. MSMCH is also an affi liate of the depositor and a direct, wholly-owned subsidiary of Morgan Stanley (NYSE:MS). As a result of a merger, completed on December 4, 2006, between a subsidiary of MSMCH and Saxon Capital, Inc., MSMCH is an affi liate, through common parent ownership, of Saxon Mortgage Services, Inc., one of the servicers.72

At the end of the day, fi nancial institutions had their fi ngers in every piece of the pie and generated fees from origination through foreclosure.73 This prompted Senator Schumer to charge: “The bottom feeders of society, these predatory lenders, reach up to the highest economic titans in society, and the two work together.”74

In just a few years, mortgage-backed securitization and related services had become huge profi t centers, stepping into the breach after 9/11 and Enron, when the initial public offering market dried up. Countrywide reportedly issued $647 billion in mort- gage bonds in 2000 and $3.8 trillion in 2006.75 In 2006, the fi nancial services indus- try contributed over 8 percent of the country’s gross domestic product and employed over six million people.76 At the big investment banks, chief executive offi cers pulled down indecent amounts of pay. In 2006, the chief executive offi cer of Morgan Stanley received over $40 million in bonuses and his counterpart at Goldman Sachs scored even more—a $53.4 million bonus.77

INVESTORS

So far we have been discussing the fi nancial institutions that issued subprime bonds, not the investors who snapped up those bonds. Buyers of mortgage-backed securities hailed from all over the world—from small shires in Australia to major Chinese banks. University endowments, pension funds, insurance companies, banks, mutual funds, and municipalities all clamored for investment grade subprime RMBS and CDOs. More aggressive investors, like hedge funds, went for the riskier tranches with their higher yields.

Investors took to subprime mortgage-backed securities like fi sh to water. At their peak, subprime bonds were considered excellent investments. There were lots of rea- sons for investors to like them. One was the fact their yields exceeded those on con- ventional government and corporate bonds; some returns were even triple the yield on U.S. Treasuries. RMBS and CDOs also appeared to offer suffi cient diversifi cation to make the risk worth taking. The thinking was that, even if a few loans went bad, the pools were large enough and diverse enough to absorb an occasional default. Fur- thermore, rating agencies touted the top-rated subprime bonds—ranging from AAA down to A—as hardly ever defaulting. Other high-yield options, like bonds issued by countries with emerging economies, were considered substantially riskier.78

Mortgage-backed securities were appealing for another reason. Investors who were looking for highly rated securities had very few options. For example, “Only fi ve non- fi nancial companies and a few sovereigns had AAA ratings as of 2007.”79 In contrast,

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 59

most tranches of mortgage-backed securities had investment grade ratings. For the same reason, investors eagerly bought the asset-backed commercial paper issued by SIVs. With banks essentially guaranteeing the SIVs’ loans, investors felt protected from any defaults on the underlying RMBS and CDOs.

State and federal laws also spurred demand for investment grade mortgage- backed securities. Insurance companies, pension plans, and banks were all sub- ject to laws restricting their bond holdings to investment grade debt. Institutional investors who wanted to expand beyond cash and corporate bonds turned to highly rated RMBS.

Even Fannie Mae and Freddie Mac bought securities backed by subprime loans. Rules governing the GSEs restricted Fannie and Freddie from purchasing loans that did not comply with their anti-predatory lending rules. But no similar restriction applied to their purchases of subprime bonds. Indeed, the Department of Housing and Urban Development (HUD) permitted Fannie and Freddie to fulfi ll their afford- able lending obligations by purchasing securities backed by subprime loans. And so they did. Between 2004 and 2006, Fannie Mae and Freddie Mac purchased a total of $434 billion in subprime mortgage-backed securities, the bulk from Countrywide, New Century, and Ameriquest.80

Challenges Facing Investors in Subprime Bonds Investing in mortgage-backed securities was not a simple task. It was nearly impossi- ble for investors to grasp their potential exposure and nearly impossible to know if they were getting fair deals. Take just one security—a CDO2, which is a collection of tranches of CDOs. The CDOs in the pool underlying the CDO2 would themselves include tranches from RMBS, which in turn incorporated pools of subprime mort- gages. The word opaque does not begin to describe these products. Credit default swaps were even harder to value. As one investment bank manager reportedly said, “We can’t accurately price [credit default swaps], although we’re confi dent that we’re getting a good price for them.”81

Both Alan Greenspan and Ben Bernanke recognized that complexity made valuing securities challenging. In 2005, when discussing CDOs, Greenspan drew attention to a study that found that “understanding the credit risk profi les of CDO tranches poses challenges to even the most-sophisticated of market participants.” Greenspan went on to advise investors “not to rely on rating-agency assessments of credit risk.”82 Later, a reporter overheard Chairman Bernanke say in reference to mortgage-backed securities, “I would like to know what those damn things are worth.”83

Given that neither the former nor the current chairman of the Fed could harness the formidable resources at their fi ngertips to determine the value of mortgage-related securities, it is not surprising that investors couldn’t either. In active markets, like pub- licly traded stocks, prices have a semblance of reliability because markets are liquid and prices are publicly posted. That was not true for subprime bonds. RMBS and CDOs were not traded on an exchange; rather, they were traded on the over-the-counter (OTC) market. Dealers executed OTC trades with customers on an individual basis, without publicly posting the sales volume or the sales price. Because these bonds were so opaque, the volume of RMBS and CDOs that were resold was small. With little

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

60 • PART I THE SUBPRIME MARKET TAKES OFF

active trading and no public resale market, there was no reliable mechanism for the market to set the price and investors could not determine whether the subprime bonds were accurately valued.

There was one vehicle that indirectly tracked the value of subprime mortgage-backed securities: the ABX index. Some credit default swaps traded with frequency and the ABX index tracked those deals. But the ABX index wasn’t perfect. Critics complained that it only captured 3 percent of the mortgage market and overstated potential losses.84

Arguably, investors could have used their own mathematical tools and had their own quants appraise the value of subprime bonds that were offered for sale. This, however, was an expensive enterprise for most investors. Some commentators main- tained that investors chose not to investigate the bonds because they “didn’t want to spend the time and money required to be prudent investors at a time when low inter- est rates had everyone reaching for higher returns without contemplating the higher risks.”85

Whatever the reason, most investors did not conduct their own due diligence. Instead, they relied on issuers’ offering documents, ratings from the rating agencies, the structure of the deals, and assurances from broker-dealers. The offering documents often omitted critical information and sometimes were just wrong. According to the SEC, Countrywide misled investors by characterizing borrowers with FICO scores as low as 500 as having prime loans, even though the industry considered any borrower with a FICO score below 620 as subprime.86

FIGURE 3.4.

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 61

Whether through necessity or choice, it was easiest for investors to rely on the rat- ing agencies. If a bond had the Good Housekeeping seal of approval from Standard & Poor’s, Moody’s, or Fitch, investors considered the security a good investment without ever assessing the agencies’ methods or the information the agencies relied on to issue the ratings. This myopia was particularly alarming because the rating agencies claimed they had no duty to confi rm the truth of the information they received from arrangers. Too late in the game, investors learned that elevating rating agencies to “god-like status” had been a mistake.87

In the CDO market, the situation was even more worrisome. Investors would com- mit to purchasing CDOs before the arrangers could even tell them what collateral would back the bonds, “making a mockery of anyone who tried to do a fundamental analysis . . . before agreeing to buy.”88 To make matters worse, competition to purchase the securities meant investors had to make decisions on the fl y. Credit committees for investors often only had a couple of days or sometimes just hours to review an offering. In the rush, they often just asked for the price.89

Investors also rashly relied on the advice of broker-dealers, mistakenly believing that the middlemen were speaking the truth. This reliance blinded them to the need to carefully review prospectuses and independently evaluate the risks involved in transac- tions.90 Investment advisors had the strongest infl uence over pint-sized institutional investors like school districts and small municipalities. Whitefi sh Bay, Wisconsin, for example, listened to its investment banker and together with four other school districts bought $200 million worth of synthetic CDOs. An analyst in Chicago explained that “Selling these products to municipalities was pretty widespread. They tend to be less sophisticated. So bankers sell them products stuffed with junk.”91

THEY SAW IT COMING

Some people maintain that the subprime collapse was a surprise. The truth is, many saw it coming and others could have had they not been blinded by euphoria or greed. Starting in 2000, lenders were making increasingly risky loans, and by the end of 2005, “Degradation of the subprime market was apparent.”92 Even in 2004, lenders were experiencing a rise in early payment defaults, which are loans that default within a few months of origination. New Century’s rate of early payment defaults, for instance, was already 7.24 percent in 2004.93

Other evidence abounded. Starting in the 1990s, there was a constant drumbeat of congressional hearings about abuses in the subprime market. In 1997, for example, Margot Saunders of the National Consumer Law Center alerted members of the Senate Banking Committee that home foreclosures had tripled in the past fi fteen years, cautioning: “It does not help Americans to tantalize them with the dream of homeownership without providing the support to allow them to maintain that home- ownership.” Drake University law professor Cathy Lesser Mansfi eld advised the House Banking Committee in 2000 that high-cost subprime loans accounted for 22 percent of all foreclosures in 1998. In 2001, the Senate Banking Committee held a hearing where Allen Fishbein of the Center for Community Change testifi ed that the “‘dirty, rotten secret’ of predatory lending is that many of the worst abuses are not necessarily illegal under existing consumer protections.” In 2004, Norma Garcia from

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

62 • PART I THE SUBPRIME MARKET TAKES OFF

the Consumers Union told the House Banking Committee that “too many subprime loans” were “simply unaffordable and destined to fail.”94

Over that same period, there were consumer protection lawsuits and enforce- ment actions against lenders across the country, many settling for huge amounts of money. For example, in 1999 Lehman Brothers faced predatory lending allegations when making a bid to acquire a Delaware bank. Lehman resolved the problem by agreeing that the bank would not “engage in predatory pricing” and would adopt procedures to “identify predatory pricing practices.”95 In 2002, Citigroup settled a Federal Trade Commission predatory lending claim for $215 million.96 In 2003, the Federal Reserve Board fi ned CitiFinancial, Citibank’s subprime arm, $70 mil- lion for making abusive loans. Around the same time, Household Finance, owned by HSBC, paid $484 million to settle consumer claims by state attorneys general. Ameriquest followed suit in 2006, paying $325 million to resolve similar claims by 48 states.97

For a time, government agencies were actively tracking the problems with sub- prime loans. In 1998, HUD and the Federal Reserve Board issued a report on defi - ciencies in subprime mortgage disclosures.98 The Department of Treasury and HUD followed two years later with a major report on subprime abuses.99 The Offi ce of Thrift Supervision closed Superior Bank, FSB, in Chicago in 2001 after the bank made billions of dollars worth of loans to unqualifi ed subprime borrowers.100 Even President George W. Bush knew about the dangers of subprime loans. In 2006, top White House advisors warned him of a housing bubble and that the country would soon face a foreclosure crisis.101

The experience in individual states also provided evidence of the building storm. Beginning with North Carolina in 1999, states passed a succession of anti-predatory lending laws. These laws were a clear signal that states were contending with mount- ing problem loans.

Behind the scenes, investment banks knew that lenders were up to no good. That knowledge did not stop them from opening the money spigot to lenders. Nor did it stop them from buying loans for securitization. Take American Business Financial Services (ABFS), a subprime lender in Philadelphia. ABFS raised money by selling notes with high interest rates directly to individuals—mostly senior citizens—through newspaper ads. ABFS then used the proceeds to make high-cost mortgage loans. Wall Street fi rms greased ABFS’ operations by lending it money and by buying its loans for securitization. Eventually, when ABFS collapsed and went into bankruptcy in 2005, investors lost over $600 million. According to the Wall Street Journal, during the bankruptcy proceedings, emails surfaced showing that investment banks had known as early as 2001 that ABFS was exploiting investors and engaging in dicey lending practices.102

Investment banks were also aware that lenders were relaxing their underwriting standards in ways that increased the risk of default. They had forty years of evidence that highly leveraged borrowing went hand in hand with heightened defaults, yet they continued to fi nance and buy loans even when the borrowers had no equity in their homes.103 Wall Street fi rms also knew that mortgage-backed securities and related derivatives were spawning manic risks. As a former risk manager at Morgan Stanley told a New York Times reporter:

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 63

You absolutely could see it coming. You could see the risks rising. However, in the two years before the crisis hit, instead of preparing for it, the opposite took place to an extreme degree. The real trouble we got into today is because of things that took place in the two years before, when the risk measures were saying things were getting bad.104

In 2005, Fannie Mae’s chief risk offi cer wrote a memo about the subprime-backed bonds in Fannie’s portfolio, warning that the loans backing the securities would lose value if housing prices dropped. He also expressed concern that the rating agencies had not adequately assessed the risk in subprime and Alt-A loans.105 The previous year, Freddie Mac’s chief risk offi cer had advised his higher-ups that subprime loans “would likely pose an enormous fi nancial and reputational risk to the company and the country.” The response from the head of Freddie Mac was that the company “couldn’t afford to say no to anyone.”106 The same sentiment reigned at Citigroup, where Charles Prince, the former CEO, said, “As long as the music is playing, you’ve got to get up and dance.”107

Rating agencies were alert to the looming crisis, too. As early as 2003, a direc- tor at Fitch Ratings reported that his fi rm was “watching closely for a loosening in underwriting guidelines. . . . [I]f we start to see changes for the worse, moving down the credit scale, that would raise red fl ags.”108 By 2005, the rating agencies were fi eld- ing complaints that their ratings on mortgage-backed securities were too high and did not accurately refl ect default risk.109 In December 2006, a Standard & Poor’s employee described his fi rm’s ratings of CDOs as creating “an even bigger monster— the CDO market. Let’s hope we are all wealthy and retired by the time this house of cards falters.”110

AIG is another case in point. AIG stopped writing credit default swaps on subprime bonds in 2005 after consulting with Wall Street fi rms. This move was in response to concerns about deterioration in the quality of subprime loans.111 In a 2007 investor conference call, AIG explained its decision to exit the market:

We were seeing . . . through the many meetings that we held with everyone related to the market, from the managers, the originators, the servicers, the repackagers, we met all of them. And we came back from our trips thinking things are changing and they are clearly not changing for the better. So as a result, we stopped accepting the collateral and pulled out of the business.112

Perhaps the strongest evidence that players knew of the risks associated with sub- prime lending comes from history. The subprime mortgage crisis that began in 2006 was not the fi rst. During the 1990s, companies like Green Tree Financial were fi nanc- ing the purchase of manufactured homes—trailers and double-wides. Like subprime mortgages, these manufactured home loans had terms that borrowers often could not afford.113 At the end of each month, Green Tree’s underwriting was at its weakest as salespeople tried to meet quotas and bonus targets.114 A 2001 article reported: “The go-go years in manufactured homes were driven by loose accounting practices, infl ated reports to investors and high-pressure sales tactics, at the local level. . . . Many con- sumers who bought mobile homes looked only at the monthly payment.”115 Green Tree, which later became part of Conseco, sold the loans for securitization on Wall

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

64 • PART I THE SUBPRIME MARKET TAKES OFF

Street. Eventually, Green Tree brought down Conseco in 2002 and forced it into bankruptcy.116

During the heyday of manufactured home lending, there were also several good- sized subprime mortgage lenders plying high-risk loans that were ultimately securi- tized. In 1998 and 1999, several of these fi rms failed. At the time, investors complained that the investment banks had done a poor job structuring the deals and that the rating companies were incompetent.117

The subprime auto fi nance market tells a similar story. In the late 1990s risky subprime car loans prompted a spate of bankruptcies among auto fi nance companies. Cutthroat competition was one cause of this distress:

The environment of readily available credit resulted in many new entrants in the subprime industry. . . . As the number of subprime automobile fi nance com- panies increased exponentially, competition for market share intensifi ed. More competition caused credit quality to deteriorate, while increasing the pricing of loans.118

It’s a sadly familiar tale.

POLICING THE MARKET

What is hard to understand is why no one but consumers, their advocates, outside researchers, and a handful of politicians yelled “fi re” even though the fl ames were at the windows. After all, one would think that if lenders were making loans to borrowers who could not afford their monthly loan payments, the market would have shut them down. Why didn’t that happen? The answer is that all the various actors, from mort- gage brokers to securities brokers and every institution in between, believed they could make money on subprime and pass the risk down the subprime food chain. In the words of George W. Bush, “Wall Street got drunk.”119

The dominant ideology under the Bush administration was that the market would sniff out mortgage abuses and excess risk and police them. But market discipline of that sort did not happen. Instead, market participants blithely believed that if the mar- ket started to tank, they could protect themselves by selling any risky holdings. With no one caring about the harm to borrowers, to society, or even to themselves, subprime lending and subprime securitization descended into a Hobbesian nightmare.

Mortgage brokers originated high-risk subprime loans because they did not bear any credit risk and collected their fees at closing. Lenders made risky loans because they earned up-front fees while dumping the loans on investors by way of arrangers. Investment banks glossed over the risks of subprime loans because their earnings came from securitization. For all these entities, any check on abusive lending would have been bad for business. As Donna Tanoue, the former chairman of the Federal Deposit Insurance Corporation, warned: “The underwriter’s motivation appears to be to receive the highest price . . . on behalf of the issuer—not to help curb predatory loans.”120

At least investors should have cared about shoddy loans, even if the middlemen did not. After all, next to borrowers, investors had the most to lose from bad sub- prime lending. In reality, investors threw caution to the wind. They believed that they

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

CHAPTER 3 A ROLLING LOAN GATHERS NO LOSS • 65

were insulated from credit risk. The ratings were strong and investors received their interest payments consistently for years so they did not question the performance of the underlying loans. They also hedged that risk by buying swap protection on the underlying securities.

There were some potential sources of market discipline. For example, lenders were often required to retain the riskiest equity tranches of RMBS when they securitized their loans, which should have given them reason to care about the quality of the loans they made. That isn’t what happened. Instead, lenders often disposed of the risk by bundling and repackaging their equity tranches and resecuritizing them into CDOs. Wall Street liked to tout recourse clauses as another form of market discipline. As long as the vast majority of loans kept performing, however, investors had no reason to insist on recourse. In 2006, when borrowers began defaulting on their mortgages at high rates, so many subprime lenders wound up in bankruptcy court or disappeared altogether that recourse clauses became unenforceable. In short, recourse provisions were only as good as a lender’s solvency and clearly were not effective at curtailing high-risk lending.

Even legal judgments did not slow down the frenzied pace of subprime securiti- zations. A test case arose involving Lehman Brothers and First Alliance Mortgage Company (FAMCO). Lehman Brothers bought hundreds of millions of dollars of loans from FAMCO at the same time that states were publicly investigating FAM- CO’s lending practices and consumers were suing FAMCO for predatory lending.121

In one consumer class action lawsuit, consumers named Lehman Brothers as a defen- dant, claiming that the fi rm had aided and abetted FAMCO’s abusive lending. The evidence in support of the claim included a 1995 memo from a Lehman executive describing FAMCO as a “sweat shop” that used “high pressure sales for people who are in a weak state.”122 Ultimately, after protracted litigation that ended up in bank- ruptcy court, Lehman Brothers was found liable and had to pay fi ve million dollars in damages. That wasn’t much for a fi nancial giant and was certainly not enough to motivate Lehman Brothers or any other Wall Street fi rm to screen out abusive loans. If anything, it gave investment banks confi dence that their relationships with preda- tory lenders would not bring them down. All told, the saga of subprime securitization rendered true the industry mantra that “a rolling loan gathers no loss.”

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use

This page intentionally left blank

EBSCOhost - printed on 4/27/2020 5:40 PM via AMERICAN PUBLIC UNIV SYSTEM. All use subject to https://www.ebsco.com/terms-of-use