SOURCE: Center for American Progress Action Fund
Download the full testimony (pdf)
The financial crisis has demonstrated just how central housing finance is to both our economy and to the lives of American families. The crisis forces us to step back and consider anew first principles—what are the goals of federal housing policy—and what system of housing finance will best accomplish these goals.
The testimony I submit today first describes the traditional goals of the system and argues they remain the right objectives. The missteps that led to the recent crisis represent, not the failure of this vision, but a failure to keep these objectives paramount. History suggests that the private market alone will not achieve these objectives.
It then looks backwards, before it looks forward. An assessment of the past is an important first step in designing the system of the future, as we must make sure we have learned the right lessons from the crisis about how to achieve the system’s goals. So this testimony lays out in some detail, first, an assessment of the origins of the crisis, a tale of failure by regulators to put the brakes on an unregulated system that was demanding the indiscriminate production of unsustainable mortgages and, second, a pointed rebuttal to some common assertions about the origins of the crisis that the evidence shows are unfounded.
Lastly, the testimony offers a caution to those who would act too precipitously or critique the administration for its deliberate step-by-step management of housing markets through the crisis.
About the Mortgage Finance Working Group (MFWG)
This testimony benefits from 18 months of conversations with the Mortgage Finance Working Group, sponsored by the Center for American Progress with the generous support of The Ford Foundation and Living Cities. CAP first assembled the MFWG members in 2008 in response to the housing crisis. These affordable housing finance experts each sought to strengthen their understanding of the causes of the crisis and possible options for public policy responses through discussion and shared learning. Immediately after the conservatorship of the housing GSEs, the group began exploring the options for the future of the U.S. mortgage markets. The members of the working group include academics, former government officials, representatives of housing nonprofit groups, private lenders and developers of affordable housing, and others. I am grateful for all I have learned from these colleagues and the ideas we have formed together, but of course I speak only for myself in the views expressed here. I offer my special thanks to CAP’s Associate Director for Financial Markets Policy for his assistance in preparing this testimony.
The working group has produced to date two pieces:
1. The goals of the housing finance system should include liquidity, stability, and affordability. These objectives served us well for almost three quarters of a century. The missteps that led to the recent crisis represent, not the failure of this vision, but a failure to keep these objectives paramount. Key features of a system to achieve these goals include: transparency, standardization, risk management, regulatory oversight, affordable and sustainable homeownership, long-term fixed rate prepayable mortgages, and access to credit for underserved communities. The system also must support a balanced housing policy that focuses on affordable rental housing, as well as sustainable homeownership, with a goal of having affordable options that are appropriate to the different circumstances of different individuals and families. History strongly suggests that the private market alone will not achieve these objectives.
2. The housing and economic crises were the result of the rapid and unchecked growth of a “shadow banking system” of unregulated and irrationally-priced private label mortgage-backed securities, or PLS. As investor demand for PLS grew, issuers in turn demanded more subprime loans than good lending practices would yield, driving down standards and distorting efficient markets for consumers, originators, issuers, and investors. The system of the future must learn the lesson from this experience. We must not reproduce a bifurcated system in which unregulated capital in one part of the market drove a “race to the bottom” in underwriting and highly leveraged risk. In the future, all mortgage backed securities, or MBS, whether or not backed by the government, must be subject to regulation. This is a key distinguishing feature of the draft proposal on which CAP’s Mortgage Finance Working Group is working—it subjects the private markets for mortgage backed securities to regulation comparable (albeit not identical) to that applied to any portion of the market benefiting from public support.
3. The Housing government-sponsored enterprises, Fannie Mae and Freddie Mac, made poor decisions with extremely costly consequences for taxpayers. They came to the party late, drawn into the subprime market in an attempt to regain lost market share and chase what seemed to be high rates of returns. As others left, they stayed and inexplicably “doubled down” as credit quality collapsed. Their regulators also made significant errors in how they exercised their oversight authority, most egregiously in giving them goals credit for subprime purchases without regard to whether the loans were sustainable. While both GSE decisions and failures of GSE regulatory oversight contributed to making problems worse, neither was the primary cause of either the flood of poorly underwritten subprime mortgage nor the larger global financial meltdown—a distinction that belongs to the failure of regulators to control the PLS market.
4. Neither was the crisis the result of lending to low- and moderate-income borrowers and minority homebuyers. Nor was it the result policies like the Community Reinvestment Act and the GSE affordable housing goals that encouraged certain institutions to provide those credit-worthy borrowers with access to credit. Misaligned incentives drove poor lending practices—not public policy goals.
5. While much of the lending to low- and moderate-income borrowers during subprime frenzy was, on net, more detriment than benefit to these families, in fact, we know how to do affordable homeownership right. In the years bad money chased out good, a range of policies and programs effectively offered sustainable, affordable homeownership. Close analysis by academics shows that borrowers benefitting from these sound lending practices were much more likely to sustain homeownership than comparable borrowers in subprime loans, even as economic conditions worsened. The secondary market system of the future should support rather than hinder the development of sound and sustainable affordable lending practices.
What’s more, we have a responsibility to ensure that the system of the future helps to repair the damage done to communities stripped of equity by subprime lending and the foreclosure epidemic. Rebuilding these communities will be impossible without access to capital in the form of fair and sustainable loans. It would be obscene if we first failed to prevent harmful subprime lending and then denied the communities hardest hit the credit needed to recover.
6. Managing the housing markets through the transition. We need a new system with new institutions to arise from the ashes, once some kind of normalcy has returned. The current situation, in which the federal government, through the GSEs or FHA-insured loans in Ginnie Mae guarantied MBS, backstops almost 90 percent of the market for home mortgages, is not desirable or sustainable.No one seeks to preserve the government’s greatly expanded role longer than necessary. We need to gradually reduce the federal role to one focused on serving the historical objectives of liquidity, stability, and affordability and concentrate a federal backstop on a smaller portion of the market that best serves public purposes.
However, even simple pronouncements by policymakers about what the future might bring could move markets and could unleash further deterioration of home values, threaten the fragile economic recovery, and make domestic and overseas investors wary of so-called “agency securities,” which represent trillions of dollars of investment in the U.S. economy. What is more, the extent of taxpayer exposure to loss from its existing backstop obligations through the GSEs and FHA, which are keeping credit flowing to the housing market today, would be increased by turmoil in the housing markets. Similarly, taxpayer losses can be mitigated by careful housing market management.
As this committee knows well, policymakers have a heavy responsibility to move ahead carefully when considering housing finance reform. There is great value in having a robust public conversation outside the government to inform policymakers before proposals are made and action is taken. That is why this series of preliminary hearings is so important to begin the debate.
1. Goals of the housing finance system
Since the Great Depression, U.S. housing finance policy has rested on three enduring objectives:
a. Liquidity: The system should provide sufficient credit liquidity to meet demand across all market segments and cycles. Transparency and standardization have proven necessary to ensure consistent, broad, and deep secondary markets necessary for liquidity.
b. Stability: The system should work to reduce swings in value and the resulting effects on the local, national, and global economy. But intermediation between the needs of short-term investors and long-term borrowers is vulnerable to bubble-bust cycles and systemic losses, so appropriate risk management and regulatory oversight are necessary to reduce, to the extent possible, these wealth-destroying cycles.
c. Affordability: The system should work to promote affordable and sustainable homeownership, broad availability of long-term, prepayable, fixed rate loans, finance for affordable multifamily housing, and access to credit for underserved communities.
While some have criticized these principles—and more generally the involvement of government in the housing finance markets—they have served this country well through many generations, and should continue to be the basis for U.S. housing finance policy going forward. As detailed later in this testimony, the problems that drove the recent housing crisis stemmed from policy makers and regulators who, enamored with the elegance of free market theory, allowed a relatively unregulated private securitization market to run amok, creating a massive credit bubble driven by unsustainable (and some even fraudulent) lending. Far from the historical goals of housing policy, it was the divergence from these principles that led our economy astray. These same underlying objectives can guide us again as we build a new and improved system of housing finance.
The U.S. residential housing market is the largest single credit market in the world, with nearly $12 trillion in total outstanding debt. To meet the mortgage needs of Americans, a tremendous amount of credit liquidity is required. And to ensure that U.S. housing markets are relatively stable, this credit liquidity must be relatively constant over time, including during economic and financial downturns.
To meet the enormous financing needs of U.S, residential housing, intermediation between the needs of investors, who are typically seeking safe, short-term, liquid assets, and the needs of borrowers, who are typically seeking risky, long-term, illiquid loans, is required. To put it simply, investors are unlikely to commit capital to borrowers for periods as long as 30 years at a fixed rate of return for even a small fraction the market at reasonable rates of return. Securitization is the primary mechanism for such intermediation, whether by the GSEs, by lenders issuing MBS with a Ginnie Mae wrap (government guarantee), or through private securitization channels. (The Federal Home Loan Banks serve a somewhat similar function, but their role is beyond the scope of this testimony.)
The alternative is a financial system that predominantly provides short-term, nonamortizing home loans, such as the ARMs that proliferated during the past decade or the short-term bullet loans that dominated during the pre-New Deal era. We have learned the dangers to family and community stability from short-term adjustable or ballooning debt. Predictable and stable housing debt has largely been a successful way for American families to acquire equity that has helped to finance the educations, small business startups, and retirements of millions. Any proposed mortgage finance system must also be judged by whether it results in the availability of long-term fixed rate credit.
Unfortunately, the process of financial intermediation is inherently pro-cyclical. During good times, financial intermediaries tend to lend too freely, taking on bad credit risks. During downturns, these same financial intermediaries face impaired balance sheets and generally become more risk averse, so they tend to constrain credit too much. In the absence of some source of countercyclical liquidity, this dynamic can severely exacerbate economic downturns, as a lack of credit suffocates an already weakened economy. The Great Depression was an example of the extreme economic deterioration that can occur when a lack of countercyclical credit is paired with an economic decline. The ability of the federal government to provide countercyclical liquidity in the most recent crisis helped to keep our economy from repeating the Depression-era experience. Retaining the ability to provide countercyclical liquidity should be a strong consideration of policy makers as they consider how to rebuild the U.S. housing finance system.
Liquidity exists only so long as there are investors who want to invest. Deeper markets result when investors have confidence, there is transparency, and standardized investment vehicles. The larger the market for securities and the more homogenized the products, the greater the liquidity the market will provide.
Both the investors who financed the U.S. mortgage market and the borrowers obtaining credit all suffered from the housing bubble and its rapid deflation. A major goal of policy should be to avoid these cycles, which have historically plagued other kinds of housing finance systems, and the large social costs they impose. Stability should continue to be a key objective.
Systemic stability is threatened by poor risk assessment and bad underwriting practices, which can be created or exacerbated by misaligned incentives. A lack of standardization and transparency also increase the likelihood of mispricing risk, as investors have less ability to independently assess risk, thus reducing market discipline.
In short, a key goal for any mortgage finance system must be to encourage the best possible risk management which requires discipline in both loan origination and intermediation. This means that pro-cyclical tendencies must be monitored and mitigated and risk must be appropriately understood and priced at all levels of the lending channel. Special attention must be paid to any systemic risks to the taxpayer and larger economy.
There is a strong social interest in providing broad access to affordable mortgage credit on fair, nondiscriminatory, and sustainable terms. Homeownership has been historically one of the primary ways most Americans accumulate wealth, allowing them to save for education, retirement, and small business formation, and climb the socioeconomic ladder. A system that does not provide access to credit to credit-worthy low- and moderate-income borrowers is therefore inconsistent with our traditions and values.
All borrowers benefit to the extent that a secondary market system of housing finance system more efficiently allocates credit to borrowers. And a higher homeownership rate, if stemming from more homebuyers for whom homeownership is appropriate, yields community benefits and social cohesion as well. But government intervention and assumption of risk cannot be justified merely to lower the cost of homeownership for middle- and upper-income borrowers.
Access to credit for communities devastated by the foreclosure crisis
In the wake of the foreclosure crisis, lenders will be tempted to limit credit availability to only the strongest borrowers. But as housing markets normalize, we must not go back to the old days where entire communities were shut out from access to the best financing. Homeowners at the higher end of the socioeconomic ladder already enjoy significant governmental subsidies. And there is ample evidence that many households that may not fit the perfect mortgage model for private lenders—“20 percent down, established credit, 31 percent debt-to-income ratio”—can become successful, long-term homeowners, when well underwritten and given access to affordable, fixed-rate financing.
Policy makers and regulators opened a Pandora’s box of unregulated predatory and unsustainable lending that had devastating consequences for low- and moderate-income communities, particularly minority communities. While some too eagerly joined the speculative furor, millions of Americans thought they were playing by the rules—work hard, buy a home, pay your mortgage—only to find that the game had been rigged against them. It would be simply obscene if, as a result of the crisis, these foreclosure-impacted communities are now deprived of the credit they need to rebuild and restore home values for everyone. The housing finance system of the future must continue to ensure there is fair, nondiscriminatory, access to sustainable lending products for credit-worthy borrowers.
Rental housing finance
The housing finance system also must provide capital to support affordable rental housing. In recent years, our implicit national housing policy disproportionately emphasized homeownership. When homeownership is done right, it can be an important tool for economic mobility and opportunity for families, as well as providing social, psychological, and societal benefits. But for some, homeownership will never be appropriate and for others, it will only be appropriate at certain times in their lives.
Affordable rental housing is particularly important given the fallout from the foreclosure crisis. As households transition out of homeownership, many with badly damaged credit, the demand for quality rental housing will grow. Demographic trends also suggest rising demand and a continued gap between incomes and the rents those incomes can support. The housing finance system must support the production and preservation of housing stock to meet the full spectrum of housing needs in America.
In recent years, the housing GSEs were a dominant source of both equity and debt for the production and preservation of the multifamily units that house most renters. With unemployment so high and incomes constrained, financiers of rental housing are now facing rising defaults akin to the earlier wave of troubled single family loans. The multifamily finance market will go through a major restructuring. We tend to think about the housing finance system predominantly in terms of homeownership. But a balanced housing policy would give explicit consideration to the design of a system that works for the finance of rental housing as well.
2. The “Shadow Banking System” of PLS
Design of the system of the future must be informed by rigorous analysis of what worked and what did not work in the housing system in recent years.It is thus worthwhile repeating the history of the crisis for the lessons it offers for the design of the future system.
Until recently, there were effectively four home mortgage lending channels in the United States:
- Loans held in portfolio by depository institutions
- Loans originated with government insurance (FHA and VA) and sold to investors in MBS with a Ginnie Mae guaranty
- Loans originated for sale to the GSEs, which then sold them to investors in the form of MBS (or held them in portfolio)
- Loans originated for sale to investors in the form of private-label MBS (PLS)
Private securitization arose in the 1980s and became a popular way to access secondary market finance for nonconforming (not eligible for GSE-securitization) mortgages, subprime, and other niche products. This channel grew as a share of mortgage originations dramatically from 2002 to 2007. This discussion describes the business practices that we saw in that period.
Like other lending channels, private securitization of mortgages intermediated between the needs of investors seeking safe and liquid investments and borrowers seeking riskier and illiquid loans. It used an “originate to distribute” model, in which lenders (banks and nonbanks) originated mortgages with the intention of reselling them to issuers of MBS. These issuers, which were typically organized as conduits (with no other assets or liabilities other than those related to the securitization of loans), pooled the mortgages, and issued bonds based on the cash flows (principal plus interest). In theory, this model distributed risk to the investors best able to bear it. In practice, however, it increased the distance between lender and borrower, made the investment more complex and opaque and risk harder to assess, and created perverse incentives for all the intermediaries paid for their role in the process without regard to the performance of the loan or investment over time.
Key to the development of private securitization as a mainstream lending channel was its ability to produce investment-grade securities, which were greatly in demand. Institutional investors (such as pension funds, mutual funds, and money market funds) and central banks and sovereign wealth funds (from export-heavy countries such as China and the OPEC nations with large trade surpluses) had growing assets, creating an enormous demand for dollar-denominated investment-grade bonds, both for direct investment and for use as collateral in a variety of transactions, including commercial paper, the repo market, and credit default swaps.
Private securitization was able to create investment-grade paper out of subprime mortgages through two main mechanisms: (1) a structure of tranches that theoretically left the senior bond holders heavily overcollateralized against credit losses and (2) third party insurance arrangements. In the first, securities were issued with different levels of seniority, with a “cascading” stream of payment as obligations to the more senior tranches were satisfied. The most senior tranches were typically investment grade (AAA or AA) rated and were the first to get paid. Only when they were paid in full, would the lower tranches get paid. Consequently, the lower tranches were higher risk and received higher coupons – rates of return. The core idea was that, with several tranches that would cumulatively absorb a high level of losses (typically between 20-50 percent), this structure could create a seemingly high quality, safe investment security (the senior tranche) out of a pool of relatively risky loans.
Private securitization also relied heavily upon the use of third-party credit guarantees, including mono-line insurance and credit default swaps (CDS), to achieve investment-grade ratings for its bonds. When the senior tranche of a PLS issue did not have sufficient overcollateralization against loss to justify an investment-grade rating, the securitization’s sponsor would often purchase third-party insurance or CDS—effectively a promise to pay the investor in the event that their bond was hit by credit losses. Because the insurers (such as Ambac) or CDS issuers (such as AIG) were typically AAA-rated credit risks, their promise to repay in the event of a default translated into a AAA rating for the bond they were guaranteeing.
Of course, the entire process relied on the assumption that the rating agencies could accurately assess risk to the investor. But increasingly, they became more focused on the structure of the transaction than on the quality of the underlying loan assets. Little attention was given to the changing characteristics of the mortgages upon which these securities rested: the credit-worthiness of the borrower and the risk that the collateral (the home) might decline in value. Past performance of similarly transactions gave rating agencies and investors a false sense of confidence, while the asset quality of the underlying mortgages fell dramatically as demand for mortgages to feed the PLS market grew.
Subprime loans quickly saturated the market, with fewer and fewer borrowers available who had comparable risk characteristics to prior-era subprime borrowers. As demand for PLS offering high yields mounted, the PLS markets adapted by broadening the criteria for loans eligible securitization, while the rating agencies continued to give the senior tranches investment grade ratings.
An analogy is helpful. Imagine that there was suddenly great demand for hamburgers in the U.S., as health experts began to extol their benefits. But the beef industry would face a shortage of beef satisfying USDA criteria for Grade A meat to sell for human consumption. To satisfy restaurant and grocery demand, the beef industry might try to convince the USDA that older, less healthy cows should receive the Grade A designation. If USDA inspectors were dependant for their income on those whose meat they graded, they might feel pressure to change the criteria. And if there was no need to publish to the public the grading standards and submit for public comment changes, it might be some time before we realized that what went into hamburgers had changed. This is basically what happened with the PLS markets. Investors were eating horse meat.
PLS markets accepted a broad array of new loan types, which were untested but high yield and high risk, and ignored serious problems with underwriting, accepting a high level of “no doc” or “low doc” loans. As Figure 1 demonstrates below (see pdf), subprime credit quality dropped precipitously, with early delinquencies rising from just above 5 percent in 2003 to over 25 percent in 2007.
These problems might have been checked by a different model in which major actors had an incentive tied to long-term performance, not simply volume of origination and issuance. In the “originate to distribute” model, the primary market drivers had no “skin in the game.” The credit risk of PLS was grossly understated and the PLS risk was seriously underpriced. With portfolio managers instructed to invest only in investment grade securities and PLS investments offering higher returns, PLS saw a huge surge in market share during the credit boom. As a result, more and more exotic and poorly underwritten mortgages originated for the PLS pipeline were pitched to consumers who might have chosen “plain vanilla” mortgages in another time. GSE market share dropped to less than 30 percent in 2006, down from over 50 percent in the 1990s
Another consequence of the tremendous demand for PLS and the subprime mortgages that fed them was pressure on the GSEs to maintain their collapsing market share. Origination channels that had typically delivered a large volume of loans to the GSEs, most famously Countrywide but many others as well, suddenly had greater leverage. They were able to get the GSEs to provide better pricing and lower credit standards, as they faced losing yet more business if they did not. The GSEs also began to buy the triple-A rated tranches of PLS for their own portfolio and convinced their regulators to credit these purchases toward affordable housing goals, inexplicably crediting goals-eligible loans without discerning whether they were sustainable. In 2007, as many investors became wary of these products, the GSEs stayed in the market for PLS longer than most, thus consuming a larger share of the shrinking pie.
As the performance of PLS backed by subprime loans began to deteriorate in 2007, financial institutions began to weigh their exposure to these instruments. Increasingly, investors in PLS (and the associated paper that utilized PLS as collateral, such as commercial paper and repo agreements) panicked. Investment banks and other institutions with large exposures were no longer trusted as counterparties and literally faced a “run on the bank” by the fall of 2008, bringing the financial system to the verge of collapse.
The system of the future must learn the lesson from this experience. We must not reproduce a bifurcated system in which unregulated capital in one part of the market drove a “race to the bottom” in underwriting and highly leveraged risk. In the future, all mortgage-backed securities (MBS), whether or not backed by the government, must be subject to regulation. This is a key distinguishing feature of the draft proposal on which CAP’s Mortgage Finance Working Group is working—it subjects the private markets for mortgage-backed securities to regulation comparable (albeit not identical) to that applied to any portion of the market benefiting from public support.
3. The role of the GSEs: Late to the party
Some argue that the GSEs were the “but-for” cause of the housing crisis. A close review of historical record shows that they made problems worse and regulators failed to step in when they might have, but their practices were not the origin of the crisis.
GSE-guaranteed MBS are based upon the cash flows from “conforming mortgages.” Investors in GSE MBS rely upon a guarantee from Fannie Mae or Freddie Mac of timely payment of principal and interest that protects the investor against credit losses, although the investor retains the interest rate risk represented by early prepayments. Their charters effectively require that borrowers obtain private mortgage insurance when the loan amount is more than 80 percent of the collateral value. Until recent years, the GSEs purchased and issued MBS based primarily on “prime” mortgages, with generally sound underwriting.
The GSEs also began in the late 1990s to issue greater amounts of debt and use the borrowed funds, not to securitize loans, but to hold whole mortgages in their retained portfolio, taking advantage of their lower cost of capital. The portfolio was especially helpful for investing in innovative and unusual loan products and to support the affordable multifamily rental market, where securitization was less common until recently. But it also became an opportunity to buy and hold PLS for the GSEs’ own account.
The relevant history of the GSEs in can be considered in four periods.
Pre-2002: GSEs dominated the mortgage markets. Default rates were generally low, housing appreciation was relatively predictable, and generally pegged to inflation, rents, and other factors (such as measured by Case-Shiller or other indices).
2002-2005: PLS experienced enormous growth, taking large market share from the GSEs. New forms of loan products, such as 2/28 interest-only ARMs, financed through PLS which were underpriced for the risk, resulted in a home lending boom, with rapid home price appreciation and high levels of home refinancing.
2005-2008: GSEs respond to PLS competition by successfully lobbying their regulator to allow them to purchase Alt-A and some subprime mortgages for securitization, as well as AAA-rated PLS and lower-quality whole loans for their portfolio.
Fall 2008 to the present: Following the failures of Lehman and AIG, and with housing-related losses soaring, the GSEs are placed into conservatorship by the federal government. PLS evaporate as a source of capital for housing finance and the GSEs become an essential source of countercyclical mortgage credit lending. They also help the Bush and then Obama administration’s to implement efforts to keep the housing markets from collapsing (such as the loan modification and refinance programs).
The GSEs biggest problems arose because they wanted to respond to the competitive threat of the PLS issuers, who were providing investment-grade PLS securities that were more attractive to investors, and subprime exotic mortgage products, that were coming to dominate the home lending market. Regulators failed to detect that the PLS market had disregarded and underpriced risk, and catastrophic consequences awaited all who followed the PLS issuers into the deep end. The GSEs’ purchases of Alt-A and subprime loans and PLS for their own portfolio certainly helped to sustain investor demand for these loans longer than if they had been precluded from their purchase. But it was competition from this underpriced market that undermined the GSE business model and drove them to take greater and greater risks.
The GSEs, like depository banks and unlike PLS, were a underprudential risk regulation regime. Bank and GSE regulators in the middle of the decade failed to intervene as the systemic risk from the PLS market infected our entire financial system. The lesson to be learned here is that competition from an unregulated channel can distort incentives in even regulated channels and the very opposite of an efficient market results.
The GSEs are now experiencing losses originating from two different sources: their traditional MBS guarantee business and the purchase of PLS and Alt-A loans for their portfolio.
First, as a mono-line business exclusively invested in housing assets, the companies experienced and continue to experience significant losses from their core business of guaranteeing MBS issued on pools of conventional conforming mortgages. The housing bubble first inflated house prices and then values fell as much as 30 percent nationwide. The stress tests that regulators applied to their book of business tested their capacity to survive two regional recessions, but nothing like the severe house price depreciation of the past three years. Regulators failed to check the PLS-driven bubble. The GSEs’ regulators failed also to judge how rapid house price appreciation exposed the GSEs to great risk of loss and allowed them to chase market share with declining credit standards on the guarantee business, especially as they moved into buying so-called “Alt-A” (typically low documentation) loans for their MBS business and to hold in portfolio.
GSE-conforming loans, however, which have historically performed well, have seen default rates that are a fraction of default rates for loans originated for the PLS market, even in this unprecedented housing downturn. As of Q2 2009, PLS made up 13 percent of all single-family first mortgages, but accounted for 35 percent of serious delinquencies. GSEs, on the other hand, held 57 percent of all such mortgages but accounted for only 26 percent of seriously delinquent mortgages.
The GSEs also have experienced losses for their portfolio, funded by issuing debt and using the proceeds to finance direct investments that they hold in portfolio rather than use to back MBS. These losses were largely accumulated from investing in the AAA-rated traunches of PLS and subprime and Alt-A loans.
In sum, the GSEs made poor decisions with extremely costly consequences for taxpayers. They came to the party late, drawn into the market in an attempt to regain lost market share and chase return. As others left, they stayed and inexplicably “doubled down” as credit quality collapsed. Their regulators also made significant errors in how they exercised their oversight authority, most egregiously in giving them goals credit for subprime purchases without regard to whether the loans were sustainable. While both GSE decisions and failures of GSE regulatory oversight contributed to making problems worse, neither was the primary cause of either the flood of poorly underwritten subprime mortgage nor the larger global financial meltdown. That distinction that belongs to the failure of regulators to control the PLS market.
4. Low-mod lending didn’t cause the crisis
Some also claim that the Community Reinvestment Act (CRA) or the affordable housing goals of the GSEs were the driving cause of the mortgage crisis, broadly claiming that government intervention overcame the markets’ ability to reach an efficient outcome in pricing risk. This narrative does little to explain how Bear Stearns or AIG became exposed to subprime mortgage risk or how Goldman Sachs and Morgan Stanley developed multitrillion dollar repo markets based on the use of AAA-rated subprime mortgage securities as collateral. It also is based in a misunderstanding of the scope and impact of CRA and the housing goals.
CRA was enacted in 1977 in response to widespread reports of redlining and other forms of discrimination. It requires covered banks to provide broad access to credit on nondiscriminatory terms in any communities in which it operates consistent with safety and soundness. It also only applies to chartered banks and thrifts. The private securitization pipeline largely bypassed these regulated institutions, using a network of nonbank lenders, such as Ameriquest and New Century, to originate loans. At the height of the subprime boom in 2006, only one of the top 25 lenders was directly subject to CRA. What’s more, CRA does not reach the bank holding company level. So the fact that Countrywide owned a bank does not mean that Countrywide Financial Corporation as a whole was subject to CRA, but only the small bank that it operated. Finally, CRA obligations only extend to communities in which a bank has a branch office. As a result, only a tiny fraction of loans could be reasonably attributed to the CRA. Indeed, CRA assessment-area lending accounted for only 9 percent of higher-priced loans to borrowers and neighborhoods potentially eligible for CRA credit.
In assessing the claim that CRA drove subprime lending, Former Comptroller of the Currency Gene Ludwig and co-authors reach this conclusion: “[I]t is apparent that the increase in subprime defaults did not result from the CRA inducing banks to reduce underwriting standards or undervalue risk. Rather, investors’ desire for higher investment yields and Wall Street’s response pulled the non-CRA, unregulated mortgage market in that direction.”
Loans originated for CRA purposes actually performed quite well, both before, during, and after the subprime bubble. As San Francisco Federal Reserve Bank President Janet Yellen has stated: “There has been a tendency to conflate the current problems in the subprime market with CRA-motivated lending, or with lending to low-income families in general. I believe it is very important to make a distinction between the two. Most of the loans made by depository institutions examined under the CRA have not been higher-priced loans, and studies have shown that the CRA has increased the volume of responsible lending to low- and moderate-income households.”
Yellen’s comments were supported by research from the Federal Reserve Bank of San Francisco, which found that, controlling for borrower and loan characteristics of more than 200,000 purchase money mortgages originated in California from 2004 to 2006 in low- and moderate- income census tracts, loans originated by CRA-regulated institutions had significantly lower likelihood of foreclosure than those originated by non-CRA regulated independent mortgage companies.
With respect to the claim that it was the affordable housing goals of the GSEs that caused the crisis, the basic argument appears to be that these goals forced the GSEs into buying up subprime securities, which single-handedly drove the market for subprime loans. While there is no doubt that the GSEs added to the total demand for such loans, this line of criticism overlooks the ample demand for such securities from the rest of the market. It also assumes that the affordable housing goals were the driving cause of the GSEs’ ill-conceived foray into AAA-rated subprime, rather than the relatively high rates of return these securities were offering over similarly rated instruments and the GSEs’ struggle to maintain market share.
There are some other important points to consider in assessing whether CRA and the affordable housing goals drove the crisis. First is the question of timing. CRA was enacted in 1977, and the GSE affordable housing goals were implemented in 1993. Why was it only in the mid-2000s that these initiatives would have caused major problems?
Furthermore, if these government mandates related to residential mortgage lending were the cause of the financial crisis, why did we see the exact same credit expansion and collapse pattern in commercial real estate, which did not have any parallel requirements to the affordable housing goals or CRA? As the chart below indicates, commercial real estate followed almost an identical bubble-bust cycle as that of residential real estate. Similar cycles can be seen in other credit markets in which private securitization played a major role.
5. Affordable homeownership done right
While much of the lending to low- and moderate-income borrowers during subprime frenzy was, on net, more detriment than benefit to these families, in fact,we know how to do affordable homeownership right. It would be unfortunate if, as a result of the crisis, we would now shy away from lending to those underserved borrowers for whom sustainable homeownership is possible with appropriate lending products and practices.
In the years before the capital markets fueled a subprime deluge and bad money chased out good, a range of policies and programs effectively created sustainable, affordable homeownership. Close analysis by academics at the Center for Community Capital at the University of North Carolina, the Urban Institute, the Federal Reserve Bank of San Francisco, and elsewhere shows that borrowers from private and nonprofit lenders using sound lending practices were much more likely to sustain homeownership than comparable borrowers in subprime loans, even as economic conditions worsened. Proven on the ground, these programs point a way forward that provides access to affordable homeownership for those who are ready for it. The secondary market system of the future should support rather than hinder the development of sound and sustainable lending practices informed by the record of affordable lending that worked.
The deluge of PLS-driven lending that targeted low- and moderate-income borrowers—particularly in minority communities—during the past decade was not actually affordable home lending. As discussed above, a robust PLS market created a strong demand for high-yield mortgage products. This was manifested in the form of strong financial incentives across the PLS “originate-to-distribute” pipeline to originate and securitize more high-cost, high-risk mortgages, such as interest-only or negative amortization adjustable-rate mortgages. The PLS market’s strong demand for high-yield mortgage products also resulted in a blind eye being turned to underwriting standards, creating the conditions for rampant fraud. Simply put, the incentives of the PLS pipeline were not to promote affordable or sustainable mortgages, but rather to promote higher-cost, higher-risk mortgages.
Thus, while a flood of cheap PLS financing poured into low- and moderate-income communities, this financing did not translate into affordable or sustainable mortgages. And it has been shown to have a negative impact on homeownership rates. We now have 2.6 million fewer homeowners than we did before the rise of subprime lending, a number that is almost certain to increase as the fallout from the foreclosure crisis continues. Because so much of the predatory lending targeted minority communities, the impacts among minority homeownership rates are just as profound. Among African Americans, the homeownership rate has dropped from 49 percent in 2004 to 46 percent at the end of 2009, a level not seen since 1999.
Lost amid the debate over how to prevent another subprime lending crisis, is the fact that subprime lending actually crowded out well-designed affordable homeownership programs that were actually working quite well. A 2009 examination of the foreclosure experiences of city-based affordable homeownership programs in five cities (Boston, Chicago, Los Angeles, New York, and San Francisco), found that, out of nearly 9,000 low-income families helped to purchase homes, the overall default rate was below 1 percent. More recently, in New York City, the housing agency reported only 13 foreclosures out of more than 20,000 subsidized homes sold to low-income buyers since 2004.
Rigorous research has confirmed that these are not just isolated successes. Researchers at the UNC Center for Community Capital compared the performance of loans from a large, national portfolio of affordable and community reinvestment act mortgages to that of loans made by the subprime market. When matching borrowers with similar profiles (for example, comparable borrower risk factors, down payment, and market conditions) the borrowers who obtained subprime loans were three to five times as likely to default as their counterparts who had received the prime, affordable mortgages instead. In this study, adjustable rate mortgage, prepayment penalty, and broker originations were features associated with increased risk of default, and layering of these features generally magnified default risk.
Perhaps more importantly, we have identified certain key features that enhance the likelihood of sustainable homeownership. While it is true that high down payments, higher wealth, and lower debt-to-income ratios are associated with lower default risk, these are by no means the only relevant factors for homeownership. Loans to LMI borrowers who lack high wealth or income are still highly sustainable when they are accompanied by flexible underwriting guidelines, combined with risk mitigation strategies, education and counseling, enhanced servicing, and default prevention strategies. Fixed rates, fully amortizing loan terms, full documentation of income, and demonstrated ability to pay are among the other important feature of affordable mortgages that work.
The experience of Self Help, based in North Carolina, illustrates these points. Since the early 1990s, Self Help has made more than 4,000 direct loans totaling over $318 million. It also created a Secondary Mortgage Market program, which has financed over 40,000 home purchase loans to low-income and minority borrowers in 48 states totaling more than $4 billion and made by nearly 40 lenders mostly between 1999 and 2005. These loans were made to low- and moderate-income borrowers (average income of $32,600), and featured minimal cash to close and high loan-to-value ratios, with more than half having an LTV of 97 percent or higher. By offering some of the loan features described above, these loans had performed quite well, with a delinquency rate well below that of subprime ARMs, subprime fixed, and even prime ARMs.
Another approach that has worked to address the wealth barrier facing lower income and minority families is shared equity. In brief, the shared equity approach bridges the gap between an affordably sized first mortgage loan not exceeding 80 percent of purchase price, and the borrower’s limited savings. Public or nonprofit supplied funding provides down payment assistance. This down payment assistance, however, is treated as an investment that creates in effect a partnership between the individual homebuyer and the public-nonprofit support. Shared equity fairly returns to the public-nonprofit its share of the investment through the creation of a long-term affordable asset, while returning to the homeowner a reasonable increase in personal wealth. Approaches such as the community land trust and deed restricted resales embody these policies. One recent study found that the foreclosure rate among community land trust homeowners was less than 0.2 percent—one-sixth of the national average and an even smaller fraction of the average among the lower-income homeowners that these groups serve.
Savings programs targeted to lower-income people, such as Individual Development Accounts, also appear to create more stable homeownership. For example, a soon to be released study sponsored by the Corporation for Enterprise Development, or CFED, examined the incidence of foreclosure among a sample of 831 IDA participants who purchased homes between 2001 and early 2008. Roughly 68 percent of IDA buyers were minority households, and roughly 75 percent were headed by women. But only 3 percent of the IDA borrowers entered foreclosure between 2001 and April 2009. This is contrasted to an overall foreclosure rate in the same communities for all loans originated over the same time period of 6.3 percent, and a nearly 9 percent foreclosure rate for low-income individuals who purchased similarly priced homes over the same time period.
6. Managing the housing market through the transition
The administration deserves credit for its deliberate step-by-step management of housing markets through the crisis. If overseas investors had lost confidence in the mortgage-backed securities that finance most home mortgages (especially as bank capital was squeezed), credit for home loans would have become virtually impossible to obtain. Home values and consumer confidence could have fallen far further if credit availability had not been maintained. I do not minimize pain felt by the millions of Americans who have lost or face losing their home or have seen their home equity erode. We have been critical of the administration for not taking some additional steps to do more to stem foreclosures. Still, they faced the risk of significantly worse than we have experienced. If they had not been as successful at keeping credit flowing, the consequences for the consumer would have been extreme.
The current situation, in which the federal government, through FHA-Ginnie Mae or the GSEs, backstops almost 90 percent of the market for home mortgages, is not desirable or sustainable. No one seeks to preserve the government’s expanded role one moment longer than necessary. But there could yet be severe consequences from acting too precipitously to disrupt the unfortunate status quo. Policymakers must move carefully and avoid destabilizing action that could unleash further deterioration of home values, threaten the fragile economic recovery, and make domestic and overseas investors wary of so-called “agency securities,” which represent trillions of dollars of investment in the U.S. economy. We have learned that the larger global economy is deeply entangled with the U.S. housing market. It would be irresponsible for policymakers to make dramatic pronouncements about what the future may bring without carefully considering the ways such pronouncements could affect domestic and international market responses and the consequences of those responses to the assets, expenses, and economic opportunities of American families.
How policymakers manage the transition from the status quo to the future will determine the ultimate taxpayer cost. The taxpayers will take significant losses from their exposure on the GSEs and are at risk as well through FHA if housing markets decline significantly further. The extent of those losses may be reduced by prudent management of housing markets by policymakers and they could be greatly exacerbated by additional home value deterioration or new threats to our very fragile economy. The wrong steps that too quickly constrain access to credit could result in far greater taxpayer losses from our existing exposure to the housing market.
As this committee knows well, simple statements by policymakers about their views on policy direction can move markets. As a result, there is great value in having a robust public conversation outside the government to inform policymakers before proposals are made and action is taken. That is why this series of preliminary hearings is so important to begin the debate.
Speaking only for myself, I do not purport to know yet the right answer to all of the complex questions that must be resolved by policymakers. This is one of the most difficult and complex set of financial problems this nation has faced. As we learned over the last few years, the consequences are great if we get it wrong.
Circumstances demand a robust but deliberate development of options and analysis of their consequences. We need to engage and learn from academics, think tank denizens, builders, lenders, investors and other market participants, community advocates, state and local government representatives, and consumer representatives, and many others before good policy choices can be confidently made.
Download the full testimony (pdf)
Sarah Rosen Wartell is the Executive Vice President at American Progress.