Board of Governors of the Federal Reserve System

Good morning. It's a great pleasure to be with you, and I want to thank the Real Property section for inviting me to speak at the Advanced Real Property Institute. I have been immersed at the Federal Reserve in economic analysis, but my first professional language is the law, so it's a special treat to be before this audience of lawyers today where I can speak in my native tongue. I'm also excited to be with you this morning because I began my time as commissioner of financial regulation for the state of Maryland in the heat of the financial crisis, and I am keenly aware that Maryland's lawyers came together in a show of massive pro bono strength to help homeowners confronting unprecedented numbers of foreclosures in the state. I want to recognize the impressive, timely, and trend-setting work that was performed during the foreclosure crisis here in Maryland. This work continues in various ways, including most recently the announcement that the state of Maryland will be participating in the launch of HOPE Loan Port, the country's first web-based foreclosure mediation portal. Lawyers in Maryland consistently have been on the forefront in supporting mediation systems to resolve issues between homeowners and servicers.

During my first week as a Federal Reserve governor, the set of mortgage servicing problems and inequities related to robo-signing hit the national news. While many were surprised by the news, this is something that we had been dealing with here in Maryland more than three years ago; as many of you know from having worked on these issues, we had identified serious servicing problems as early as 2008 in Maryland, had analyzed them quite extensively, and had crafted some effective legal, administrative, and legislative early responses. Being rather dismayed that the federal folks believed this to be a new problem, and concerned that problems with servicing nationwide would hinder the revival of the housing market and our recovery from recession, I gave a speech advocating a series of measures at the national and state levels to contain the damage to homeowners in November 2010. To my dismay, here we are in 2011, with a recovery that is still being dragged down by serious housing problems that will require not just economic talent--but, significantly, legal talent--to address.

I want to talk this morning about a couple of those legal challenges. Of course, these remarks are intended to express my own views, and they do not necessarily reflect the opinions of my colleagues on the Federal Reserve Board or the Federal Open Market Committee.

Housing Market Conditions
Six years after house prices first began to fall, we are still dealing with the aftermath of the housing bubble. Nationally, house prices have fallen by nearly one-third since their peak in the first quarter of 2006, and total homeowners' equity in the United States has shrunk by more than one-half--a loss of more than $7 trillion. This is a shocking and enormous decline.

The drop in house prices has had far-reaching effects on families and the economy, in part because so many American families--nearly 70 percent--own their homes.¹ In contrast, only about one-half of American families hold any stocks, either directly or through accounts such as Individual Retirement Accounts, 401(k)s, or mutual funds.² The fall in house prices has caused families to cut back on their spending and has prevented families from using their home equity to fund education expenses or start small businesses. An estimated 3 million families are not able to refinance their mortgages at today's historically low interest rates because they are underwater on their mortgages.³ And many borrowers have lost their homes to foreclosure: More than 4 percent of all mortgages in the United States were in foreclosure, and an additional 3 percent or so were delinquent by 90 days or more in the second quarter of this year. Initially, when I was commissioner of financial regulation, foreclosures were concentrated among borrowers with subprime mortgages. But today, because of the magnitude of the economic downturn and the high level of unemployment, the typical borrower in foreclosure has a prime or Federal Housing Administration mortgage.

As all of you know, the drop in house prices has put considerable pressure on the legal system. Of course, the huge wave of foreclosures has strained and sometimes overwhelmed the courts, particularly in states with judicial foreclosure processes. But in a deeper sense, the standard contracts that govern mortgage lending, servicing, and securitization did not account for a potential collapse of this magnitude in the housing sector. As the bubble expanded, securitization contracts did not provide enough incentive to discourage shoddy underwriting practices. In the aftermath, the inadequacies of servicing contracts have contributed to the agonizingly slow pace of mortgage modifications and repeated breakdowns in the foreclosure process.

Prior to the crisis, the nation's mortgage finance system developed into an extensive set of specialized industry players with a complex set of linkages operating not just at the lender and borrower level, but also with an extensive secondary market that was intended to diffuse risk. All of these relationships were governed by contracts. The promissory note, for example, lays out the terms under which the borrower will repay the loan, while the mortgage reflects the lender's security interest in the property. But there are also contracts between the borrower and third parties and between the lender and third parties. For example, the borrower likely will enter agreements for title, flood, or private mortgage insurance, while the lender may enter an agreement with another institution to service the loan. Add mortgage securitization to the mix, and the number of contractual relationships snowballs. For a private-label securitization, a crucial contract is the pooling and servicing agreement, or PSA, which specifies matters such as the characteristics of the mortgages in the pool, how these mortgages will be serviced, and how the money generated from the loans will be distributed to investors. At nearly every linkage of our mortgage finance system there is a contractual relationship, and at nearly every contractual linkage there has been legal challenge and litigation.

For example, let's just zero in on one such contractual relationship and examine how the legal structure set forth in this relationship has held up. First, let's look at the provisions that govern mortgage servicing, and second, the provisions that lay out the representations and warranties about the features and underwriting quality of the loans included in the securitization. Both sets of provisions exist in the contract that governs the securitization, the PSA.

Mortgage Servicing
The problems in mortgage servicing have been well documented. For example,  the federal banking regulators found significant problems when we looked at mortgage servicing and foreclosure processing at 14 federally regulated mortgage servicers, which collectively represent more than two-thirds of all servicing volume nationally. These problems include critical weaknesses in servicers' foreclosure governance processes, foreclosure document preparation processes, and oversight and monitoring of third-party law firms and other vendors. These problems indicate the existence of unsafe and unsound banking practices and violations of federal and state laws, as well as demonstrated patterns of misconduct and negligence on the part of servicers. Individual consumers have been harmed by this negligence, including the negligence of law firms. In addition, the imperative for servicers to fix their systems, review past decisions, and put in place the internal systems and controls that they should have had all along has impeded the repair of the overall housing market.

The burden of addressing these issues of repair, review, and implementation of systems and controls rests with the mortgage servicers, their parent financial institutions, and the law firms that mishandled the foreclosures. But what do the terms of the PSA, which is the standard contract governing this work, require in this context? The standard servicing contract provides disincentives for servicers to act in the best interests of investors and borrowers. This misalignment of incentives has more profound consequences when defaults are high.

Under the standard contract, servicers receive a flat fee per loan that they service, usually 1/4 to 1/2 percent of the unpaid loan balance annually, depending on the type of loan. The flat fee is considerably more than is required to service a performing loan and considerably less than is required to service a delinquent loan. The expectation is that, on average, the fee will cover the servicer's expenses. Servicers are also reimbursed for some expenses, such as the cost of a title search when pursuing a foreclosure, but generally not for unanticipated overhead or labor costs.

The PSA aligns the incentives of borrowers, servicers, and investors reasonably well when mortgage defaults are low. Under normal circumstances, mortgage servicing is profitable. A servicer primarily processes mortgage payments and minimizes labor and overhead costs by harnessing economies of scale and automating the process as much as possible.

The PSA does not perform nearly as well under stressed circumstances. When mortgage delinquencies are high, mortgage servicing is not profitable, and servicers may feel extra pressure to cut costs as much as possible. In addition, servicers may not be properly motivated to perform loan modifications even when such modifications are in the best interests of borrowers and investors. Servicer compensation is not generally tied to the performance of the loan, and in most cases a servicer receives no extra payment for preventing a default. Further, loan modifications are labor intensive, and this extra labor cost is not reimbursed under the contract. Instead, the PSA provides for the reimbursement of some foreclosure expenses.

Other language in the contract also does little to encourage loan modifications, particularly in the case of securitized mortgages. Many PSAs provide minimal guidance about modifications beyond specifying that servicers should apply "usual and normal" servicing standards or the same standards that they apply to the loans held in portfolio. With such vague guidance, servicers assert that they are worried about litigation risk if they employ servicing approaches that have not been widely adopted throughout the industry. Furthermore, even when the PSAs provide guidance about loan modifications, it can differ widely across PSAs, often preventing servicers from designing a uniform modification program for their entire portfolio.

To borrow from the title of a popular marital advice column, Can this contract be saved? I think the answer is no--at least not in its current form. This contract might have been adequate if servicers had followed the age-old advice and used their profits from the fat years when defaults were low in order to build their capacity for lean years when defaults were high. But in practice, this does not appear to have happened--and perhaps even the most knowing servicer a few years ago might not have anticipated the subsequent level of mortgage defaults.

Given that failure, it is imperative to reconsider the compensation structure so that servicers have adequate incentives to perform payment processing efficiently on performing mortgages, and to perform effective loss mitigation on delinquent loans. After the compensation structure is reconsidered, the PSAs need to be amended or renegotiated in order to facilitate more workouts. Finally, PSAs should clarify the situations in which loan modifications and other mitigation strategies should be pursued. One tool that could aid in providing such clarity, and has received substantial attention over the last few years, is the net present value model. Requiring servicers to take mitigative actions that are net-present-value positive to the investor could encourage the fair and consistent treatment of borrowers. 

Investors also need tools that will allow them to better monitor servicer performance and take action accordingly. These tools should be developed and described in the contract. Currently, metrics that allow investors to measure servicers' execution are not widely available. Such metrics could include customer satisfaction ratings, delinquency and cure rates, the average time that a homeowner waits on the phone to talk with the servicer, and servicer error rates. Indeed, one can even imagine the development of a uniform servicer scorecard. To be meaningful, the ability to transfer servicing from low-performing servicers to high-performing ones would have to be enhanced. The creation of common back-office systems across servicers would make transfer less prone to error and less costly, and any contractual or legal barriers would need to be reduced to allow investors to "fire" a low-performing servicer. A new contractual regime designed along these lines would represent a significant change from the existing world of servicing, but it could help create a system in which servicers compete on the quality of their performance and are more accountable to both investors and consumers.

Representations and Warranties
I turn next to the problem of representations and warranties contained in the PSAs. Underwriting standards declined dramatically in the middle part of the previous decade as the housing bubble approached its peak. For example, the median combined loan-to-value ratio on subprime mortgages originated for home purchases rose from 90 percent in 2003 to 100 percent in 2005--meaning that more than one-half of borrowers who purchased homes with subprime mortgages put no money down. The share of mortgages in which borrowers did not document fully their income or assets also increased. In addition, the number of mortgages that defaulted in the first year after origination--commonly considered a gauge of poor underwriting--rose appreciably as the bubble approached its peak.⁴ 

The "originate to distribute" model of loan origination may have contributed to this decline in underwriting standards. Under this model, brokers or lenders sell the loan to a third party, typically a securitizer, who then issues securities collateralized by mortgage loans to investors. This model technically describes securities guaranteed by Fannie Mae and Freddie Mac as well as private-label securities sponsored by financial institutions such as commercial or investment banks, but problems were more acute in the private-label market.

Although the private-label originate-to-distribute model may have worked adequately at first, this discipline eroded as the securitization boom progressed. Originators and securitizers were generally able to complete securitization transactions while retaining very little or no economic stake in the performance of the loan after the sale, although some of these securities did ultimately make their way back onto the balance sheets of certain financial conglomerates, much to their dismay. As a result, originators and securitizers had an increasing incentive to weaken underwriting standards in order to increase the volume of loan origination, and thereby the volume of securitizations.

As securitization developed and became more prominent as a source of financing for mortgage lending, representations and warranties were one of the main tools market participants relied on to align incentives among originators and securitizers on the one hand and investors on the other. These provisions describe the underwriting standards and other matters with respect to the assets that are the subject of the securitization. The originator, in principle, is required to refund at par (less payments received) the value of the loan should it violate the originator's representations and warranties or in some cases, should it default within a specified time from origination. In practice, however, representations and warranties did not function as intended.

First, as in the case of servicing, contracts in private-label mortgage securitizations varied considerably from deal to deal, and there was no consistent language or standard for representations and warranties. Therefore, a poorly underwritten mortgage loan could breach the representations and warranties of one deal's contract but not be a violation of representations and warranties in another very similar deal; this discrepancy would only be apparent following a careful reading and nuanced comparison of the differences in drafting between the two contracts. Given the volume of transactions investors were considering, it is not surprising that they may not have conducted a thorough analysis of the different risks each set of representations and warranties presented. Second, the lack of standardization of representations and warranties gives rise to greater opportunity for parties in a failed securitization transaction to argue over the interpretation of contract terms, increasing legal uncertainty and making effective resolution of investor claims more difficult. Finally, in order to enforce their representations and warranties, investors may need to band together in groups representing 25 percent or more of the voting rights in order to make certain demands upon the securitization trustee. Investors may be dispersed and have difficulty coordinating their actions.

The flaws in the representations and warranties provisions of the PSAs appear to be widely recognized. For example, many key players in the securitization industry have worked through Project RESTART, an initiative of the American Securitization Forum (ASF), to create an industry standard for representations and warranties that covers prudent practices and is commonly understood by all market participants.⁵   The ASF has also published model principles for repurchasing mortgage loans that breach representation warranties to provide a standard industry PSA enforcement mechanism.⁶ I believe that these are useful developments and should promote greater consistency in governing contracts of mortgage securitizations in the future, thereby facilitating investor monitoring of underwriting standards and setting common expectations for resolution of breaches of representations and warranties when they occur.

Nevertheless, standard representations and warranties, no matter how well drafted, are unlikely on their own, or even in combination with a well-constructed enforcement standard, to go far enough to resolve the crucial problems of misaligned incentives and disparities in access to information in the securitization process. During lending and asset booms, there is a temptation for market participants to collectively allow underwriting standards to weaken in order to meet a market demand for increased lending. When a crisis then occurs, bankruptcies of originators and securitizers of mortgages may prevent investors in practice from obtaining restitution for breaches of representations and warranties. 

Thus, other contractual mechanisms are also required to promote a healthy and properly aligned securitization market. For instance, comprehensive disclosures about the mortgage loans being securitized are vital to ensure that investors are able to conduct meaningful due diligence prior to making a decision to invest in a securitization. They are also vital to investors' ongoing monitoring of the servicing and administration of securitization transactions. The work of the Securities and Exchange Commission to revise securitization disclosures is particularly important and should significantly contribute to improving the flow of information in securitization markets.⁷ 

Conclusion
The period since the onset of the financial crisis has been described as one of the worst economic episodes since the Great Depression, which was another time of tremendous strain on the housing market. House prices declined about 30 percent from peak to trough, and the homeownership rate, as recorded by the decennial census, fell from 48 percent in 1930 to 44 percent in 1940. Nonfarm mortgage foreclosures reached a thousand per day in 1933, and, in fact, foreclosures were so widespread that 28 states imposed foreclosure moratoriums. Mortgage lending was disrupted, and private mortgage insurance companies and private mortgage securitization collapsed.

One factor in the upheaval in mortgage lending was the mortgage contract. Before the Great Depression, the long-term fully amortized contract had not yet been widely adopted. Instead, one common contract featured a short maturity, generally five years or less, and often did not call for any regular payments on the principal. A borrower was likely to default at the end of the loan term if he was not able to roll over the loan or obtain a new loan to pay off the existing balance. This contract also made it difficult to assess the health of financial institutions, as lenders booked the loans as short‑term exposures when the risk was actually much longer term. One contemporary source called the contract "a menace."⁸ 

After the Great Depression, fully amortizing loans with maturities of 15 years or longer became the standard. Federal policy appears to have been a key factor in this transition. First, the Home Owners' Loan Corporation--a sort of precursor to today's Making Home Affordable program--purchased distressed mortgages and modified them into loans that amortized for 15 years or longer.⁹  Second, the newly created Federal Housing Administration sold insurance on 20-year fully amortizing loans. Third, savings and loans with new federal charters were required to make fully amortizing loans except when state law forbade it.

The challenge for us today, as I see it, is to emulate our predecessors from the 1930s who dealt with a comparable crisis. We need to consider our current array of mortgage contracts with a dispassionate eye and open mind. If the contracts today are not working‑-and the evidence seems clear to me that, along some dimensions, they are not--we as lawyers should be working on ways to improve them. Without such improvements you may not be serving your clients, and from the perspective of the macroeconomy, none of us will be contributing to a reconstruction of the legal framework in the crucial context of mortgage finance. Lawyers have a singular responsibility and exceptional ability to help rebuild this framework, and I urge you to consider how you can contribute to the effort of renegotiating these relationships and how the terms of such a renegotiation might become memorialized in contracts. This is a massive challenge for lawyers who practice in the field of mortgage finance and your country needs you to tackle it with the same vigor and energy that you have brought to other legal challenges in this housing crisis.

Cicero said that "the safety of the people should be the highest law." Our people will not be financially safe until we re-examine and re-set the terms that define relationships in the housing market. So we have a serious challenge ahead of us. It's a challenge that I have confidence we can meet with your help.

Thank you, and I'm happy to take your questions and hear your comments.

 

---

1. See U.S. Census Bureau, "Housing Vacancies and Homeownership (CPS/HVS)," webpage. Return to text

2. See Brian K. Bucks, Arthur B. Kennickell, Traci L. Mach, and Kevin B. Moore (2009), "Changes in U.S. Family Finances from 2004 to 2007:  Evidence from the Survey of Consumer Finances (PDF)," Federal Reserve Bulletin, vol. 95 (February), pp. A1-A55. Return to text

3. This calculation is a Board staff estimate of the number of borrowers with fixed-rate mortgages with an interest rate of 4.75 or higher and a loan-to-value ratio greater than 100 percent. Return to text

4. See Christopher Mayer, Karen Pence, and Shane Sherlund (2009), "The Rise in Mortgage Defaults," Journal of Economic Perspectives, vol. 23 (Winter), pp. 27-50. Return to text

5. For model representations and warranties, see American Securitization Forum (2009), "ASF Project RESTART:  ASF Model RMBS Representations and Warranties (PDF)," paper, December 15. Return to text

6.  For model repurchase principles, see American Securitization Forum (2011), "ASF Project RESTART:  ASF Model RMBS Repurchase Principles Release (PDF)," press release, August 30. Return to text

7. See Disclosure of Asset-Backed Securities Required by Section 943 of the Dodd-Frank Wall Street Reform and Consumer Protection Act, 76 Fed. Reg. 4489 (Jan. 26, 2011) (to be codified at 17 C.F.R. pts. 229, 232, 240, and 249). In addition, the risk retention requirements of section 941 of the Dodd-Frank Wall Street Reform and Consumer Protection Act are also important to a comprehensive effort to align incentives in the securitization markets in order to promote sound underwriting. By requiring securitizers to retain a share of the economic risk in the mortgages they securitize from the outset of the transaction, the requirement creates incentives for securitizers to better monitor the credit quality of assets they securitize and ultimately discourages unsafe and unsound underwriting practices by originators.  Return to text

8. See Morton Bodfish and A.D. Theobald (1938), Savings and Loan Principles (New York:  Prentice-Hall), p. 175. Return to text

9. See Jonathan D. Rose (2011), "The Incredible HOLC?  Mortgage Relief during the Great Depression," Journal of Money, Credit and Banking, vol. 43 (6), pp. 1073-107. Return to text