Moving Forward

Brookings Institution Press

Contents

Chapter One

ERIC S. BELSKY AND NELA RICHARDSON

The cycle of boom and bust in nonprime and nontraditional mortgage lending in the United States is without precedent. The factors that fueled the boom and the way it unfolded sowed the seeds for the bust that, in hindsight, appears to have been inevitable. The amount of risk in the system ballooned as a result of changes in lending practices. At the same time that credit was opened up to borrowers who previously had been denied loans because of past problems repaying their debts, many other underwriting standards were loosened. In addition, products with heavy payment reset risks proliferated in both the prime and nonprime markets. This layering of risk at or near the peak of an overheated housing market proved very deleterious to loan performance.

Yet few predicted that performance in the nonprime mortgage market and the way these loans were packaged, sold, and referenced in the global capital markets would cause a loss of investor confidence so profound that it would spark a severe global financial crisis. It was not until August 2007 that the Federal Reserve decided that the rapidly eroding performance of subprime mortgage loans—and evaporating demand for the securities they backed—was enough of a threat to the broader economy to warrant easing monetary policy. In an unusual move, the Fed lowered the discount rate for borrowing from the Federal Reserve in between regularly scheduled meetings of the Federal Open Market Committee. Although the committee held the more important federal funds rate target constant until its September meeting, lowering the discount rate signaled both its concern and willingness to take action to contain the damage from the deteriorating subprime residential mortgage market.

These interventions would prove inadequate. A little more than a year later, and within the span of less than two weeks, the government helped to rescue Bear Stearns and Merrill Lynch from collapse, allowed Lehman Brothers to fail, and bailed out insurance giant AIG. Credit markets froze nearly solid in the fall of 2008, the stock market went into a freefall, and job losses accelerated sharply. The interconnectedness of the global financial system became apparent as problems emanating from residential debt in the United States and in the derivatives used to hedge and trade mortgage risk prompted a global credit crisis.

Uncovering the causes of the nonprime boom and bust is essential to formulating effective government and business responses to the crisis. At stake are not only the safety and soundness of the financial system the next time that excess global liquidity creates pressure to relax underwriting standards and raise leverage, but also the access that Americans will have to mortgage credit, on what terms, and at what cost. Access to mortgage credit is vital to building assets through homeownership and opens up avenues to finance consumption and investment on terms that are generally more favorable than those for consumer credit. Government cannot easily back away from it without risking great economic dislocations. Especially at a time when the share of U.S. households with credit problems has soared, how credit-impaired borrowers are treated will shape asset-building opportunities during the next economic expansion for millions of Americans. And while the recent housing bust has underscored the risky nature of investing in residential real estate, it also has created the conditions—ratios of house price to income in some locations at or near lows not seen since the early 1990s—that could make homeownership very attractive for years to come.

Understanding the boom and bust in nonprime and nontraditional lending first requires a brief discussion of the evolution of the housing finance system in the United States from the 1970s onward.

Evolution of the Modern U.S. Housing Finance System

Throughout the 1970s, 1980s, and 1990s, mortgage lending was conducted with a limited number of mortgage products that dominated the market and were underwritten, with few exceptions, to long-accepted common standards. These relatively stringent underwriting criteria formed the backbone of a single "prime" market in which credit was allocated by adhering to these tight standards and charging all who were able to qualify for mortgages a similar interest rate. Only loans with higher ratios of loan to value commonly incurred the additional payment of a private or government mortgage insurance premium to offset the greater risk associated with lower-down-payment lending. As such, the system of credit allocation was considered a rationing system rather than a risk-based pricing system. This began to change around the early to middle 1990s, when—haltingly at first—individuals with previous problems repaying their debts were allowed to qualify for a loan but were charged a higher interest rate to cover the expected higher risk of default. Thus began a period of pricing for risk in a nonprime market rather than allocating mortgage credit at a common price in a single prime market. Around the same time, the number of mortgage products available to borrowers slowly began to increase, first with more widespread use of "hybrid" adjustable rates (which had an initial adjustment period of more than one year followed by conversion to a fixed rate). But by the mid-2000s nonprime lending had taken off, and nontraditional products had proliferated.

Segmentation of the Mortgage Market

During the 1970s, 1980s, and most of the 1990s, the mortgage market was essentially segmented into three parts. Two were parts of the "conventional" market, defined as the market for loans that did not have explicit federal guarantees against the loss of principal. The largest segment of the mortgage market was the "conforming" side of the conventional market. This was the market for loans purchased by or placed into mortgage-backed securities (MBSs) guaranteed by Fannie Mae and Freddie Mac. These two shareholder-owned corporations were chartered by Congress. Because of their unique charters and small lines of credit from the Department of Treasury, these government-sponsored enterprises (GSEs) were perceived by investors as having the implicit backing of the federal government. The two factors that made loans "conforming" is that they followed the exacting underwriting standards demanded by these two corporations and fell underneath loan size limits established by the federal government and benchmarked to a federal index. The other segment of the conventional market was the "jumbo" side. This was the market for loans above the conforming limits established for Fannie Mae and Freddie Mac or not acceptable because they deviated from the underwriting requirements of these two secondary mortgage market giants. While some small portion of these jumbo loans had more lenient standards or product features than Fannie Mae and Freddie Mac would accept, the overwhelming majority of loans simply were above the loan limits. Thus the conventional market for nearly the entire period also was the prime market. The third segment of the market was made up of loans or mortgage-backed securities explicitly guaranteed by the federal government. Loans insured by the government had more lenient down payment and debt-to-income requirements than conventional loans and were subject to relatively low mortgage limits established by Congress. The two major agencies insuring mortgages with the full faith and credit of the federal government were the Federal Housing Administration (FHA) and the Veterans Benefits Administration. Ginnie Mae was the agency that guaranteed the timely payment of principal and interest of securities backed by loans insured by the FHA. Figure 1-1 shows these shares over time.

Growth of the Secondary Mortgage Market

A hallmark of the evolution of the housing finance system since the 1970s has been the growth and development of the secondary market for mortgages, mostly in the form of securities they guaranteed. The secondary market developed rapidly over the course of the 1980s and continued to grow in the 1990s and 2000s. Initially, the movement away from holding whole loans in portfolios was precipitated by macroeconomic events in the 1980s and a series of laws that deregulated mortgage lending and supported the secondary market in general and Fannie Mae and Freddie Mac in particular. These included the Depository Institutions Deregulation and Monetary Control Act, the Alternative Transactions Mortgages Parity Act, and the Secondary Mortgage Market Enhancement Act. The macroeconomic events that facilitated the secondary market included the deep recession of the early 1980s and the wild gyrations in interest rates that hurt banks and thrifts stuck holding thirty-year fixed-rate mortgages when interest rates first rose sharply (because the interest rates they had to offer to attract depositors were above the interest rates on the loans they held in portfolio) and then when rates plummeted (because borrowers prepaid their mortgages and returned principal at a time when mortgage rates were much lower).

These events led more and more lenders to seek ways to get the implicit backing of the federal government by selling loans to Fannie Mae and Freddie Mac or to get explicit backing by using FHA insurance and Ginnie Mae MBS guarantees. In addition, these events generated interest in finding ways to convert illiquid assets into more tradable and liquid homogeneous securities with implicit or explicit government backing. Financial engineering in the secondary market gave another reason: the capacity to buy different classes of multiple-class securities backed by a single pool, with each class having different priorities from principal and interest payments—and hence having different exposures to prepayment risks. This process was known as "structured" finance because classes were structured to appeal to a variety of different appetites for prepayment risk. Lastly, as banks and thrifts lost market share to pension funds and insurance companies and as appetite for dollar-denominated mortgage assets around the world increased, the demand for securities backed implicitly or explicitly by the federal government also grew. Investors who did not make the mortgage loans or specialize in mortgage lending could look past the credit risk that taking on such debt might pose when acquiring mortgage assets. Thus participants in the market were persuaded that credit risk was fully neutralized by the FHA and Ginnie Mae, Fannie Mae, or Freddie Mac guarantees (with private mortgage insurance playing a supporting role).

Industry Consolidation and the Originate-to-Distribute Model

Two other important developments in the 1980s and 1990s are also worth noting. One—the increasing reliance on an "originate-to-distribute" model—is directly related to growth in the secondary market. The other was consolidation within the industry.

As the secondary market developed, more and more mortgages were originated by brokers, mortgage bankers, and banks and thrifts that collected a fee for originating loans (and sometimes for retaining servicing rights and the fees associated with them) but then conveyed loans to issuers of "agency" mortgage-backed securities (as MBSs guaranteed by Fannie Mae, Freddie Mac, and Ginnie Mae were called). During the 1990s, brokers steadily increased their share of the originations from 18.8 percent in 1994 to 27.9 percent in 2000 and to a peak of 31.3 percent in 2005 (see figure 1-2). The issuers of securities backed by loans conveyed by these brokers then assumed the credit risk on the loans. Furthermore, servicing rights became actively traded so that the originating lender often did not end up servicing the loans it originated. This system was efficient in that it allowed some firms to specialize in retail originations, others in pooling and wholesaling loans, and others in issuing securities and managing and pricing credit risk. The increasing use of mortgage brokers also allowed lenders to rely on variable rather than fixed costs to source loans, which had great advantages in the volatile and cyclical mortgage business. However, it also created a system in which credit risk was concentrated in a small number of entities and in which loan originators had more incentive to produce high unit volumes to earn up-front fees than to concern themselves with the long-term performance of the loans originated.

Consolidation in mortgage origination and servicing was fueled by the increasing commoditization of mortgages that secondary markets and advances in information technology allowed as well as by consolidation in the banking industry (which was, in turn, fueled by deregulation, such as the lifting of interstate banking restrictions). In 1996 the largest twenty-five lending institutions accounted for 40 percent of the $785 billion in home purchase and refinance originations. By 2008 their share had grown to more than 90 percent (see figure 1-3). Mortgage servicing also was consolidating (see figure 1-4). Like the originate-to-distribute model, consolidation had strengths and weaknesses. On the one hand, economies of scale were achieved that brought down the costs of originating and servicing. On the other, a small customer base for Fannie Mae and Freddie Mac grew to have increasing market power in negotiating with these corporations. A handful could produce large exposures to counterparty risks (the risk that counterparties will work against the interests of another or that the failure of one will cause significant harm to another).

Emergence and Rapid Growth of Nonprime Lending

Beginning in the 1990s, the segmentation of the market into conventional conforming, jumbo, and government-insured mortgages that had been in place since the 1970s began to break down, and the nonprime market emerged. Finance companies that funded their operations with corporate bond issues began to lend to borrowers with previous credit problems who could not meet the conforming market standards. The subprime industry was born largely in an effort to serve these borrowers. The early pioneers reasoned that they could lend to these borrowers if the borrowers had gotten behind on their credit payments because of temporary shocks to incomes or expenses that had largely passed and if the borrowers had substantial equity in their homes. The second condition offered lenders protection if the bet that these borrowers could repay their loans and rebuild their credit histories proved bad. Much of this lending was for second mortgages. These lenders charged higher interest rates on the loans to cover higher expected losses and found that borrowers were willing to pay them.

As a market emerged in which loans were priced based on risk, complaints began to arise that some lenders were preying on unsuspecting borrowers. Consumer advocates labeled a slew of practices as "predatory," and they lobbied at the federal, state, and local levels to protect consumers from these practices. Concerns also began to emerge that consumers could be discriminated against in the interest rate charged for the loan or the fees or conditions imposed rather than through loan rejection.

Around the same time, the prime market began to use statistical credit and mortgage scores to do a better job of discerning good from bad risks. Mortgage scores were based on modeling the performance of mortgage products that had been around for long periods of time and were based on performance under a range of interest rate environments and market conditions. These models were embedded in automated underwriting systems that increasingly supplanted manual underwriting. The advantage was that the automated systems were both better at modeling risk and less subjective, relying on colorblind models instead of on individual underwriters armed with detailed and large manuals that they could use to exercise discretion to approve variances from stricter underwriting and documentation standards. The large banks and Fannie Mae and Freddie Mac that developed these automated systems and scoring models found that the systems both drove down costs and allowed more borrowers to qualify for loans without adding to expected risk because the models allowed tighter underwriting on one standard to compensate for more lenient underwriting on another.

Armed with mortgage-scoring models and some limited experience with nonprime mortgage lending, nonprime lending expanded rapidly after 2000 and most especially around 2004. At first, most of the subprime loans originated were held in portfolio. Increasingly, though, these loans were sold and placed into securities issued by investment banks or by large specialist subprime lenders. At the peak of the subprime lending boom in 2005 about eight in ten—and by 2007 fully nine in ten—subprime loans were placed into securities, according to Inside Mortgage Finance. With the exception of $13 billion wrapped by Fannie Mae, these securities had neither an implicit government guarantee against loss of principal from Fannie Mae or Freddie Mac nor an explicit guarantee from the FHA and Ginnie Mae. These securities were therefore called private-label securities and were traded in what was called the asset-backed securities (ABSs) market. This market was separate and distinct from the "agency" market where securities had implicit or explicit federal guarantees against the loss of principal (also called credit risk). From 1985 to 1995, the private-label MBS market grew from just $3.9 billion a year to fully $69 billion in constant 2008 dollars and continued to grow rapidly after 1995. While total MBS issuance increased more than 70 percent in real terms from $449 billion in 1995 to $769 billion in 2000, nonagency MBS issuance increased more, from a real annual level of $69 billion in 1995 to $170 billion in 2000. This lifted its share of MBS issuance from 15 percent in 1995 to 22 percent in 2000.

(Continues...)

Excerpted from Moving Forward Copyright © 2011 by THE BROOKINGS INSTITUTION. Excerpted by permission of Brookings Institution Press. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.