III. THE U.S. GOVERNMENT’S ROLE
IN FOSTERING THE GROWTH OF
THE NTM MARKET

The preceding section of this dissenting statement described the damage that was done to the financial system by the unprecedented number of defaults and delinquencies that occurred among the 27 million NTMs that were present there in 2008. Given the damage they caused, the most important question about the financial crisis is why so many low quality mortgages were created. Another way to state this question is to ask why mortgage standards declined so substantially before and during the 1997-2007 bubble, allowing so many NTMs to be created. This massive and unprecedented change in underwriting standards had to have a cause—some factor that was present during the 1990s and thereafter that was not present in any earlier period. Part III addresses this fundamental question.

The conventional explanation for the financial crisis is the one given by Fed Chairman Bernanke in the same speech at Morehouse College quoted at the outset of Part II:

Saving inflows from abroad can be beneficial if the country that receives those inflows invests them well. Unfortunately, that was not always the case in the United States and some other countries. Financial institutions reacted to the surplus of available funds by competing aggressively for borrowers, and, in the years leading up to the crisis, credit to both households and businesses became relatively cheap and easy to obtain. One important consequence was a housing boom in the United States, a boom that was fueled in large part by a rapid expansion of mortgage lending. Unfortunately, much of this lending was poorly done, involving, for example, little or no down payment by the borrower or insufficient consideration by the lender of the borrower’s ability to make the monthly payments. Lenders may have become careless because they, like many people at the time, expected that house prices would continue to rise--thereby allowing borrowers to build up equity in their homes--and that credit would remain easily available, so that borrowers would be able to refinance if necessary. Regulators did not do enough to prevent poor lending, in part because many of the worst loans were made by firms subject to little or no federal regulation. [Emphasis supplied]⁵⁹

In other words, the liquidity in the world financial market caused U.S. banks to compete for borrowers by lowering their underwriting standards for mortgages and other loans. Lenders became careless. Regulators failed. Unregulated originators made bad loans. One has to ask: is it plausible that banks would compete for borrowers by lowering their mortgage standards? Mortgage originators—whether S&Ls, commercial banks, mortgage banks or unregulated brokers—have been competing for 100 years. That competition involved offering the lowest rates and the most benefits to potential borrowers. It did not, however, generally result in or involve the weakening of underwriting standards. Those standards—what made up the traditional U.S. mortgage—were generally 15 or 30 year amortizing loans to homebuyers who could provide a downpayment of at least 10-to-20 percent and had good credit records, jobs and steady incomes. Because of its inherent quality, this loan was known as a prime mortgage.

There were subprime loans and subprime lenders, but in the early 1990s subprime lenders were generally niche players that made loans to people who could not get traditional mortgage loans; the number of loans they generated was relatively small and bore higher than normal interest rates to compensate for the risks of default. In addition, mortgage bankers and others relied on FHA insurance for loans with low downpayments, impaired credit and high debt ratios. Until the 1990s, these NTMs were never more than a fraction of the total number of mortgages outstanding. The reason that low underwriting standards were not generally used is simple. Low standards would result in large losses when these mortgages defaulted, and very few lenders wanted to hold such mortgages. In addition, Fannie and Freddie were the buyers for most middle class mortgages in the United States, and they were conservative in their approach. Unless an originator made a traditional mortgage it was unlikely that Fannie or Freddie or another secondary market buyer could be found for it.

This is common sense. If you produce an inferior product—whether it’s a household cleaner, an automobile, or a loan—people soon recognize the lack of quality and you are out of business. This was not the experience with mortgages, which became weaker and riskier as the 1990s and 2000s progressed. Why did this happen?

In its report, the Commission majority seemed to assume that originators of mortgages controlled the quality of mortgages. Much is made in the majority’s report of the so-called “originate to distribute” idea, where an originator is not supposed to care about the quality of the mortgages because they would eventually be sold off. The originator, it is said, has no “skin in the game.” The motivation for making poor quality mortgages in this telling is to earn fees, not only on the origination but in each of the subsequent steps in the securitization process.

This theory turns the mortgage market upside down. Mortgage originators could make all the low quality mortgages they wanted, but they wouldn’t earn a dime unless there was a buyer. The real question, then, is why there were buyers for inferior mortgages and this, as it turns out, is the same as asking why mortgage underwriting standards, beginning in the early 1990s, deteriorated so badly. As Professor Raghuram Rajan notes in Fault Lines, “[A]s brokers came to know that someone out there was willing to buy subprime mortgage-backed securities without asking too many questions, they rushed to originate loans without checking the borrowers’ creditworthiness, and credit quality deteriorated. But for a while, the problems were hidden by growing house prices and low defaults—easy credit masked the problems caused by easy credit—until house prices stopped rising and the flood of defaults burst forth.”⁶⁰

Who were these buyers? Table 1, reporting the number of NTMs outstanding on June 30, 2008, identified government agencies and private organizations required by the government to acquire, hold or securitize NTMs as responsible for two-thirds of these mortgages, about 19 million. The table also identifies the private sector as the securitizer of the remaining one-third, about 7.8 million loans. In other words, if we are looking for the buyer of the NTMs that were being created by originators at the local level, the government’s policies would seem to be the most likely culprit. The private sector certainly played a role, but it was a subordinate one. Moreover, what the private sector did was respond to demand—that’s what the private sector does—but the government’s role involved deliberate policy, an entirely different matter. Of its own volition, it created a demand that would not otherwise have been there.

The deterioration in mortgage standards did not occur—contrary to the Commission majority’s apparent view—because banks and other originators suddenly started to make deficient loans; nor was it because of insufficient regulation at the originator level. The record shows unambiguously that government regulations made FHA, Fannie and Freddie, mortgage banks and insured banks of all kinds into competing buyers. All of them needed NTMs in order to meet various government requirements. Fannie and Freddie were subject to increasingly stringent affordable housing requirements; FHA was tasked with insuring loans to low-income borrowers that would not be made unless insured; banks and S&Ls were required by CRA to show that they were also making loans to the same group of borrowers; mortgage bankers who signed up for the HUD Best Practices Initiative and the Clinton administration’s National Homeownership Strategy were required to make the same kind of loans. Profit had nothing to do with the motivations of these firms; they were responding to government direction. Under these circumstances, it should be no surprise that underwriting standards declined, as all of these organizations scrambled to acquire the same low quality mortgages.

1. HUD’s Central Role

In testimony before the House Financial Services Committee on April 14, 2010, Shaun Donovan, Secretary of Housing and Urban Development, said in reference to the GSEs: “Seeing their market share decline [between 2004 and 2006] as a result of [a] change of demand, the GSEs made the decision to widen their focus from safer prime loans and begin chasing the non-prime market, loosening long-standing underwriting and risk management standards along the way. This would be a fateful decision that not only proved disastrous for the companies themselves –but ultimately also for the American taxpayer.”

Earlier, in a “Report to Congress on the Root Causes of the Foreclosure Crisis,” in January 2010, HUD declared “The serious financial troubles of the GSEs that led to their being placed into conservatorship by the Federal government provides strong testament to the fact that the GSEs were, indeed, overexposed to unduly risky mortgage investments. However, the evidence suggests that the GSEs’ decisions to purchase or guarantee non-prime loans was motivated much more by efforts to chase market share and profits than by the need to satisfy federal regulators.”⁶¹ [emphasis supplied]

Finger-pointing in Washington is endemic when problems occur, and agencies and individuals are constantly trying to find scapegoats for their own bad decisions, but HUD’s effort to blame Fannie and Freddie for the decline in underwriting standards sets a new standard for running from responsibility. Contrast the 2010 statement quoted above with this statement by HUD in 2000, when it was significantly increasing Fannie and Freddie’s affordable housing goals:

Lower-income and minority families have made major gains in access to the mortgage market in the 1990s. A variety of reasons have accounted for these gains, including improved housing affordability, enhanced enforcement of the Community Reinvestment Act, more flexible mortgage underwriting, and stepped-up enforcement of the Fair Housing Act. But most industry observers believe that one factor behind these gains has been the improved performance of Fannie Mae and Freddie Mac under HUD’s affordable lending goals. HUD’s recent increases in the goals for 2001-03 will encourage the GSEs to further step up their support for affordable lending.⁶² [emphasis supplied]

Or this statement in 2004, when HUD was again increasing the affordable housing goals for Fannie and Freddie:

Millions of Americans with less than perfect credit or who cannot meet some of the tougher underwriting requirements of the prime market for reasons such as inadequate income documentation, limited downpayment or cash reserves, or the desire to take more cash out in a refinancing than conventional loans allow, rely on subprime lenders for access to mortgage financing. If the GSEs reach deeper into the subprime market, more borrowers will benefit from the advantages that greater stability and standardization create.⁶³[emphasis supplied]

Or, finally, this statement in a 2005 report commissioned by HUD:

More liberal mortgage financing has contributed to the increase in demand for housing. During the 1990s, lenders have been encouraged by HUD and banking regulators to increase lending to low-income and minority households. The Community Reinvestment Act (CRA), Home Mortgage Disclosure Act (HMDA), government-sponsored enterprises (GSE) housing goals and fair lending laws have strongly encouraged mortgage brokers and lenders to market to low-income and minority borrowers. Sometimes these borrowers are higher risk, with blemished credit histories and high debt or simply little savings for a down payment. Lenders have responded with low down payment loan products and automated underwriting, which has allowed them to more carefully determine the risk of the loan.⁶⁴ [emphasis supplied]

Despite the recent effort by HUD to deny its own role in fostering the growth of subprime and other high risk mortgage lending, there is strong—indeed irrefutable—evidence that, beginning in the early 1990s, HUD led an ultimately successful effort to lower underwriting standards in every area of the mortgage market where HUD had or could obtain influence. With support in congressional legislation, the policy was launched in the Clinton administration and extended almost to the end of the Bush administration. It involved FHA, which was under the direct control of HUD; Fannie Mae and Freddie Mac, which were subject to HUD’s affordable housing regulations; and the mortgage banking industry, which— while not subject to HUD’s legal jurisdiction—apparently agreed to pursue HUD’s policies out of fear that they would be brought under the Community Reinvestment Act through legislation.⁶⁵ In addition, although not subject to HUD’s jurisdiction, the new tighter CRA regulations that became effective in 1995 led to a process in which community groups could obtain commitments for substantial amounts of CRA-qualifying mortgages and other loans to subprime borrowers when banks were applying for merger approvals.⁶⁶

By 2004, HUD believed it had achieved the “revolution” it was looking for:

Over the past ten years, there has been a ‘revolution in affordable lending’ that has extended homeownership opportunities to historically underserved households. Fannie Mae and Freddie Mac have been a substantial part of this ‘revolution in affordable lending’. During the mid-to-late 1990s, they added flexibility to their underwriting guidelines, introduced new low-downpayment products, and worked to expand the use of automated underwriting in evaluating the creditworthiness of loan applicants. HMDA data suggest that the industry and GSE initiatives are increasing the flow of credit to underserved borrowers. Between 1993 and 2003, conventional loans to low income and minority families increased at much faster rates than loans to upper-income and nonminority families.⁶⁷[emphasis supplied]

This turned out to be an immense error of policy. By 2010, even the strongest supporters of affordable housing as enforced by HUD had recognized their error. In an interview on Larry Kudlow’s CNBC television program in late August, Representative Barney Frank (D-Mass.)—the chair of the House Financial Services Committee and previously the strongest congressional advocate for affordable housing—conceded that he had erred: “I hope by next year we’ll have abolished Fannie and Freddie . . . it was a great mistake to push lower-income people into housing they couldn’t afford and couldn’t really handle once they had it.” He then added, “I had been too sanguine about Fannie and Freddie.”⁶⁸

2. The Decline of Mortgage
Underwriting Standards

Before the enactment of the GSE Act in 1992, and HUD’s adoption of a policy thereafter to reduce underwriting standards, the GSEs followed conservative underwriting practices. For example, in a random review by Fannie Mae of 25,804 loans from October 1988 to January 1992, over 78 percent had LTV ratios of 80 percent or less, while only 5.75 percent had LTV ratios of 91 to 95 percent.⁶⁹ High risk lending was confined primarily to FHA (which was controlled by HUD) and specialized subprime lenders who often sold the mortgages they originated to FHA. What caused these conservative standards to decline? The Commission majority, echoing Chairman Bernanke, seems to believe that the impetus was competition among the banks, irresponsibility among originators, and the desire for profit. The majority’s report offers no other explanation.

However, there is no difficulty finding the source of the reductions in mortgage underwriting standards for Fannie and Freddie, or for the originators for whom they were the buyers. HUD made clear in numerous statements that its policy—in order to make credit available to low-income borrowers—was specifically intended to reduce underwriting standards. The GSE Act enabled HUD to put Fannie and Freddie into competition with FHA, and vice versa, creating what became a contest to lower mortgage standards. As the Fannie Mae Foundation noted in a 2000 report, “FHA loans constituted the largest share of Countrywide’s [subprime lending] activity, until Fannie Mae and Freddie Mac began accepting loans with higher LTVs [loan-to-value ratios] and greater underwriting flexibilities.”⁷⁰

Under the GSE Act, the HUD Secretary was authorized to establish affordable housing goals for Fannie and Freddie. Congress required that these goals include a low and moderate income goal and a special affordable goal (discussed below), both of which could be adjusted in the future. Among the factors the secretary was to consider in establishing the goals were national housing needs and “the ability of the enterprises [Fannie and Freddie] to lead the industry in making mortgage credit available for low-and moderate-income families.” The Act also established an interim affordable housing goal of 30 percent for the two-year period beginning January 1, 1993. Under this requirement, 30 percent of the GSEs’ mortgage purchases had to be affordable housing loans, defined as loans to borrowers at or below the AMI.⁷¹

Further, the Act established a “special affordable” goal to meet the “unaddressed needs of, and affordable to, low-income families in low-income areas and very low-income families.” This category was defined as follows: “(i) 45 percent shall be mortgages of low-income families who live in census tracts in which the median income does not exceed 80 percent of the area median income; and (ii) 55 percent shall be mortgages of very low income families,” which were later defined as 60 percent of AMI.⁷² Although the GSE Act initially required that the GSEs spend on special affordable mortgages “not less than 1 percent of the dollar amount of the mortgage purchases by the [GSEs] for the previous year,” HUD raised this requirement substantially in later years. Ultimately, it became the most difficult affordable housing AH burden for Fannie and Freddie to meet.

Finally, the GSEs were directed to: “(A) assist primary lenders to make housing credit available in areas with low-income and minority families; and (B) assist insured depository institutions to meet their obligations under the Community Reinvestment Act of 1977.”⁷³ There will be more on the CRA and its effect on the quality of mortgages later in this section.

Congress also made clear in the act that its intention was to call into question the high quality underwriting guidelines of the time. It did so by directing Fannie and Freddie to “examine—

(1) The extent to which the underwriting guidelines prevent or inhibit the purchase or securitization of mortgages for houses in mixed-use, urban center, and predominantly minority neighborhoods and for housing for low-and moderate-income families;

(2) The standards employed by private mortgage insurers and the extent to which such standards inhibit the purchase and securitization by the enterprises of mortgages described in paragraph (1); and

(3) The implications of implementing underwriting standards that— (A) establish a downpayment requirement for mortgagors of 5 percent or less; (B) allow the use of cash on hand as a source of downpayments; and (C) approve borrowers who have a credit history of delinquencies if the borrower can demonstrate a satisfactory credit history for at least the 12-month period ending on the date of the application for the mortgage.”⁷⁴

I could not find a record of reports by Fannie and Freddie required under this section of the act, but it would have been fairly clear to both companies, and to HUD, what Congress wanted in asking for these studies. Prevailing underwriting standards were inhibiting mortgage financing for low and moderate income (LMI) families, and would have to be substantially relaxed in order to meet the goals of the Act. Whatever the motivation, HUD set out to assure that downpayment requirements were substantially reduced (eventually they reached zero) and past credit history became a much less important issue when mortgages were made (permitting subprime mortgages to become far more common).

Until 1995, HUD enforced the temporary AH goals originally put in place by the GSE Act. With the exception of the special affordable requirements, which were small at this point, these goals were not burdensome. In the ordinary course of their business, the GSEs seem to have bought enough mortgages made to borrowers below the AMI to qualify for the 30 percent AH goal. In 1995, however, HUD raised the LMI goal to 40 percent, applicable to 1996, and to 42 percent for subsequent years. In terms of its effect on Fannie and Freddie, HUD’s most important move at this time was to set a Special Affordable goal (low and very low income borrowers) of 12 percent, which increased to 14 percent in 1997. Efforts to find loans to low or very low income borrowers (80 percent and 60 percent of AMI, respectively) that did not involve high risks would prove difficult. As early as November 1995, even before the effect of these new and higher goals, Fannie’s staff had already recognized that Fannie’s Community Homebuyer Program (CHBP), which featured a 97 percent loan-to-value (LTV) ratio—i.e., 3 percent downpayment⁷⁵—was showing significant rates of serious delinquency that exceeded Fannie’s expected rates by 26percent in origination year 1992, 93 percent in 1993 and 57 percent in 1994.⁷⁶

In 1995, continuing its efforts to erode underwriting standards in order to increase homeownership, HUD issued a policy statement entitled “The National Homeownership Strategy: Partners in the American Dream.” The Strategy was prepared by HUD, “under the direction of Secretary Henry G. Cisneros, in response to a request from President Clinton.”⁷⁷ The first paragraph of Chapter 1 stated: “The purpose of the National Homeownership Strategy is to achieve an all-time high level of homeownership in America within the next 6 years through an unprecedented collaboration of public and private housing industry organizations.”

The Strategy paper then noted that “industry representatives agreed to the formation of working groups to help develop the National Homeownership Strategy” and made clear that one of its purposes was to increase homeownership by reducing downpayments: “Lending institutions, secondary market investors, mortgage insurers, and other members of the partnership should work collaboratively to reduce homebuyer downpayment requirements. Mortgage financing with high loan-to-value ratios should generally be associated with enhanced homebuyer counseling and, where available, supplemental sources of downpayment assistance.”⁷⁸ According to a HUD summary, the purpose of the Strategy was to make financing “more available, affordable, and flexible.”⁷⁹ [emphasis supplied] It continued:

The inability (either real or perceived) of many younger families to qualify for a mortgage is widely recognized as a very serious barrier to homeownership. The National Homeownership Strategy commits both government and the mortgage industry to a number of initiatives designed to:

Cut transaction costs through streamlined regulations and technological and procedural efficiencies.

Reduce downpayment requirements and interest costs by making terms more flexible, providing subsidies to low- and moderate-income families, and creating incentives to save for homeownership.

Increase the availability of alternative financing products in housing markets throughout the country.⁸⁰ [emphasis supplied]

Reductions in downpayments, the area on which HUD particularly concentrated in pursuing its AH goals and the National Homeownership Strategy, are especially important in weakening underwriting standards. Table 4, below, based on a large sample of loans from the 1990s, shows the risk relationships between downpayments and mortgage risks. It is particularly instructive to note that when low downpayments (i.e., high LTVs) are combined with low FICO scores (subprime loans) the expected delinquencies and defaults are multiplied several fold. For example, when a loan with a FICO score below 620 is combined with a downpayment of five percent, the risk of default is 4.2 times greater than it would be if the downpayment were 25 percent.

Table 4.⁸¹ High LTVs enhance the risk of low FICO scores

Despite these obvious dangers, HUD saw the erosion of downpayment requirements imposed by the private sector as one of the keys to the success of its strategy to increase home ownership through the “partnership” it had established with the mortgage financing community: “The amount of borrower equity is an important factor in assessing mortgage loan quality. However, many low-income families do not have access to sufficient funds for a downpayment. While members of the partnership have already made significant strides in reducing this barrier to home purchase, more must be done. In 1989 only 7 percent of home mortgages were made with less than 10 percent downpayment. By August 1994, low downpayment mortgage loans had increased to 29 percent.”⁸² [emphasis supplied]

HUD’s policy was highly successful in achieving the goals it sought. In 1989, only one in 230 homebuyers bought a home with a downpayment of 3 percent or less, but by 2003 one in seven buyers was providing a downpayment at that level, and by 2007 the number was less than one in three. The gradual increase in LTVs and CLTVs (first and second loans combined to produce a lower downpayment) under HUD’s policies is shown in Figure 4. Note the date (1992) when HUD began to have some influence over the downpayments that the GSEs would accept.

That HUD’s AH goals were the reason Fannie increased its high LTV (low downpayment) lending is clearly described in a Fannie presentation to HUD assistant secretary Albert Trevino on January 10, 2003: “Analyses of the market demonstrate the greatest barrier to home ownership for most renters are related to wealth—the lack of money for a downpayment…our low-downpayment lending— negligible until 1994—has grown considerably. It is a key part of our strategy to serve low-income and minority borrowers.” The figure that accompanied that statement showed that Fannie’s home purchase loans over 95 percent LTV had increased from one percent in 1994 to 7.9 percent in 2001.⁸³

Figure 4. Estimated Percentage of Home Purchase Volume with an LTV or CLTV >=97%
(Includes FHA and Conventional Loans*)
and Combined Foreclosure Start Rate for Conventional and Government Loans

* Fannie‘s percentage of home purchase loans with an LTV or CLTV >=97% used as the proxy for conventional loans.
Sources: FHA 2009 Actuarial Study, and HUD”s Office of Policy Development and Research - Profiles of GSE Mortgage Purchases in 1999 and 2000, in 2001-2004, and in 2005-2007, and Fannie’s 2007 10-K. Compiled by Edward Pinto.

The close relationship between low downpayments and delinquencies and defaults on mortgages is shown in Figure 5, which compares the increase in FHA 97 percent (or greater) CLTV or LTV mortgages to the increase in the foreclosure start rate on all loans published by the Mortgage Bankers Association.

Figure 5. Relationship between low downpayments and delinquencies or defaults on mortgages

Sources: MBA National Delinquency Survey, FHA 2009 Actuarial Study, and HUD’s Office of Policy Development and Research -Profiles of GSE Mortgage Purchases in 1999 and 2000, in 2001-2004, and in 2005-2007, SMR’s “Piggyback Mortgage Lending,” and Fannie’s 2007 10-K. Fannie is used as the proxy on the conventional market. Compiled by Edward Pinto.

In 1995, HUD also ruled that Fannie and Freddie could get AH credit for buying PMBS that were backed by loans to low-income borrowers.⁸⁴ This provided an opportunity for subprime lenders to create pools of subprime mortgages that were likely to be AH goals-rich. These were then sold through Wall Street underwriters to Fannie and Freddie, which became the largest buyers of these high risk PMBS between 2002 and 2005.⁸⁵ These PMBS pools were not bought for profit. As Adolfo Marzol, Fannie’s Chief Credit Officer, noted to Fannie CEO Dan Mudd in a 2005 memorandum, “large 2004 private label [PMBS] volumes were necessary to achieve challenging minority lending goals and housing goals.”⁸⁶ There is a strong possibility that by creating a market for PMBS backed by NTMs Fannie and Freddie enabled Wall Street—which had previously focused on securitizing prime jumbo loans—to get its start in developing an underwriting business in PMBS based on NTMs.

HUD pursued these policies throughout the balance of the Clinton administration and into the administration of George W. Bush. Ultimately, they would lead to the mortgage meltdown in 2007, as vast numbers of mortgages with low or no downpayments and other non-traditional features suffused the financial system. But in June, 1995, the dangers in HUD’s policies were not recognized. As President Clinton said in a 1995 speech, “Our homeownership strategy will not cost the taxpayers one extra red cent. It will not require legislation. It will not add more federal programs or grow the Federal bureaucracy.”⁸⁷ The lesson here is that the government can accomplish a lot of its goals without growing, as long as it has the power to enlist the private sector. That does not mean, however, as we have all now learned, that the taxpayers will not ultimately be faced with the costs.

The next significant move in the AH goals was made under HUD Secretary Andrew Cuomo, and it was a major step. On July 29, 1999, HUD issued a press release with the heading “Cuomo Announces Action to Provide $2.4 trillion in Mortgages for Affordable Housing for 28.1 Million Families.”⁸⁸ The release began: “Housing and Urban Development Secretary Andrew Cuomo today announced a policy to require the nation’s two largest housing finance companies to buy $2.4 trillion in mortgages over the next 10 years to provide affordable housing for about 28.1 million low-and moderate-income families.” This was followed by a quote from President Clinton to emphasize the importance of the initiative: “During the last six and a half years, my Administration has put tremendous emphasis on promoting homeownership and making housing more affordable for all Americans…Today, the homeownership rate is at an all-time high, with more than 66 percent of all American families owning their homes. Today, we take another significant step.”

The release then pointed out that the AH goals would be substantially raised and that “[u]nder the higher goals, Fannie Mae and Freddie Mac will buy an additional $488.3 billion in mortgages that will be used to provide affordable housing for 7 million more low-and moderate-income families over the next 10 years. Those new mortgages and families are over and above the $1.9 trillion in mortgages for 21.1 million families that would have been generated if the current goals had been retained.” The release also noted that “Fannie Mae Chairman Franklin D. Raines joined Cuomo at the news conference in which Cuomo announced the HUD action. Raines committed Fannie Mae to reaching HUD’s increased Affordable Housing goals.”

The policy behind this substantial increase in the AH goals was expressed in HUD’s discussion of the rule-making: “To fulfill the intent of [the GSE Act], the GSEs should lead the industry in ensuring that access to mortgage credit is made available for very low-, low- and moderate-income families and residents of underserved areas. HUD recognizes that, to lead the mortgage industry over time, the GSEs will have to stretch to reach certain goals and close the gap between the secondary mortgage market and the primary mortgage market. This approach is consistent with Congress’ recognition that ‘the enterprises will need to stretch their efforts to achieve’ the goals.”⁸⁹ [emphasis supplied]

The new AH goals announced in 1999 were not finally issued until October 2000. Their specifics were stunning and drove Fannie and Freddie into a new and far more challenging era. The basic goal, an LMI requirement of 42 percent, was raised to 50 percent, and the special affordable goal was raised from 14 percent to 20 percent. As a result, 75 percent of the increase in goals was concentrated in the low- and very-low income category—where the risks were the greatest. A HUD memo summarized the new rules:⁹⁰

For each year from 2001 through 2003, the goals are:

• Low- and moderate-income goal. At least 50 percent of the dwelling units financed by each GSE’s mortgage purchases should be for families with incomes no greater than area median income (AMI), defined as median income for the metropolitan area or nonmetropolitan county. The corresponding goal was 42 percent for 1997-2000.
• Special affordable goal. At least 20 percent of the dwelling units financed by each GSE’s mortgage purchases should be for very low-income families (those with incomes no greater than 60 percent of AMI) or for low-income families (those with incomes no greater than 80 percent of LMI) in low-income areas. The corresponding goal was 14 percent for 1997-2000.
• Underserved areas goal. At least 31 percent of the dwelling units financed by each GSE’s mortgage purchases should be for units located in underserved areas. Research by HUD and others has demonstrated that low-income and high-minority census tracts have high mortgage denial rates and low mortgage origination rates, and this forms the basis for HUD’s definition of underserved areas. The corresponding goal was 24 percent for 1997-2000.

HUD’s new and more stringent AH goal requirements immediately stimulated strong interest at the GSEs for CRA loans, substantial portions of which were likely to be goals-qualifying. This is evident in a speech by Fannie’s Vice Chair, Jamie Gorelick, to an American Bankers Association conference on October 30, 2000, just aft er HUD announced the latest increase in the AH goals for the GSEs:

Your CRA business is very important to us. Since 1997, we have done nearly $7 billion in specially targeted CRA business—all with depositories like yours. But that is just the beginning. Before the decade is over, Fannie Mae is committed to finance over $20 billion in specially targeted CRA business and over $500 billion in CRA business altogether…

We want your CRA loans because they help us meet our housing goals… We will buy them from your portfolios, or package them into securities… We will also purchase CRA mortgages you make right at the point of origination... You can originate CRA loans for our purchase with one of our CRA-friendly products, like our 3 percent down Fannie 97. Or we have special community lending products with flexible underwriting and special financing… Our approach is “CRA your way”.⁹¹

The 50 percent level in the new HUD regulations was a turning point. Fannie and Freddie had to stretch a bit to reach the previous goal of 42 percent, but 50 percent was a significant challenge. As Dan Mudd told the Commission,

Fannie Mae’s mission regulator, HUD, imposed ever-higher housing goals that were very difficult to meet during my tenure as CEO [2005-2008]. The HUD goals greatly impacted Fannie Mae’s business, as a great deal of time, resources, energy, and personnel were dedicated to finding ways to meet these goals. HUD increased the goals aggressively over time to the point where they exceeded the 50% mark, requiring Fannie Mae to place greater emphasis on purchasing loans to underserved areas. Fannie Mae had to devote a great deal of resources to running its business to satisfy HUD’s goals and subgoals.⁹²

Mudd’s point can be illustrated with simple arithmetic. At the 50 percent level, for every mortgage acquired that was not goal-qualifying, Fannie and Freddie had to acquire a goal-qualifying loan. Although about 30 percent of prime loans were likely to be goal-qualifying in any event (because they were made to borrowers at or below the applicable AMI), most prime loans were not. Subprime and other NTM loans were goals-rich, but not every such loan was goal-qualifying. Accordingly, in order to meet a 50 percent goal, the GSEs had to purchase ever larger amounts of goals-rich NTMs in order to acquire sufficient quantities of goals-qualifying loans.

Thus, in a presentation to HUD in 2004, Fannie argued that to meet a 57 percent LMI goal (which was under consideration by HUD at the time) it would have to acquire 151.5 percent more subprime loans than the goal in order to capture enough goal-qualifying loans.⁹³ Moreover, with the special affordable category at 20 percent in 2004, the GSEs had to acquire large numbers of NTM loans from borrowers who were at or below 60 percent of the AMI. This requirement drove Fannie and Freddie even further into risk territory in search of loans that would meet this subgoal.

Most of what was going on here was under the radar, even for specialists in the housing finance field, but not everyone missed it. In a paper published in 2001,⁹⁴ financial analyst Josh Rosner recognized the deterioration in mortgage standards although he did not recognize how many loans were subject to this problem:

Over the past decade Fannie Mae and Freddie Mac have reduced required down payments on loans that they purchase in the secondary market. Those requirements have declined from 10% to 5% to 3% and in the past few months Fannie Mae announced that it would follow Freddie Mac’s recent move into the 0% down payment mortgage market. Although they are buying low down payment loans, those loans must be insured with ‘private mortgage insurance’ (PMI). On homes with PMI, even the closing costs can now be borrowed through unsecured loans, gifts or subsidies. This means that not only can the buyer put zero dollars down to purchase a new house but also that the mortgage can finance the closing costs….

[I]t appears a large portion of the housing sector’s growth in the 1990’s came from the easing of the credit underwriting process….The virtuous cycle of increasing homeownership due to greater leverage has the potential to become a vicious cycle of lower home prices due to an accelerating rate of foreclosures.⁹⁵[emphasis supplied]

The last increase in the AH goals occurred in 2004, when HUD raised the LMI goal to 52 percent for 2005, 53 percent for 2006, 55 percent for 2007 and 56 percent for 2008. Again, the percentage increases in the special affordable category outstripped the general LMI goal, putting added pressure on Fannie and Freddie to acquire additional risky NTMs. This category increased from 20 percent to 27 percent over the period. In the release that accompanied the increases, HUD declared:

Millions of Americans with less than perfect credit or who cannot meet some of the tougher underwriting requirements of the prime market for reasons such as inadequate income documentation, limited downpayment or cash reserves, or the desire to take more cash out in a refinancing than conventional loans allow, rely on subprime lenders for access to mortgage financing. If the GSEs reach deeper into the subprime market, more borrowers will benefit from the advantages that greater stability and standardization create.⁹⁶ [emphasis supplied]

Fannie did indeed reach deeper into the subprime market, confirming in a March 2003 presentation to HUD, “Higher goals force us deeper into FHA and subprime.”⁹⁷ According to HUD data, as a result of the AH goals Fannie Mae’s acquisitions of goal-qualifying loans (which were primarily subprime and Alt-A) increased (i) for very low income borrowers from 5.2 percent of their acquisitions in 1993 to 12.2 percent in 2007; (ii) for special affordable borrowers from 6.4 percent in 1993 to 15.2 percent in 2007; and (iii) for less than median income borrowers (which includes the other two categories) from 29.2 percent in 1993 to 41.5 percent in 2007.⁹⁸

By 2004, Fannie and Freddie were sufficiently in need of subprime loans to meet the AH goals that their CEOs, as the following account shows, went to a meeting of mortgage bankers to ask for more subprime loan production:

The top executives of Freddie Mac and Fannie Mae [Richard Syron and Franklin Raines] made no bones about their interest in buying loans made to borrowers formerly considered the province of nonprime and other niche lenders. …Fannie Mae Chairman and [CEO] Franklin Raines told mortgage bankers in San Francisco that his company’s lender-customers ‘need to learn the best from the subprime market and bring the best from the prime market into [the subprime market].’ He offered praise for nonprime lenders that, he said, ‘are some of the best marketers in financial services.’… We have to push products and opportunities to people who have lesser credit quality,” he said.⁹⁹ [emphasis supplied]

Accordingly, by 2004, when HUD put new and tougher AH goals into effect, Fannie and Freddie were using every available resource to meet the goals, including subprime loans, Alt-A loans and the purchase of PMBS. Some observers, including the Commission’s majority, have claimed that the GSEs bought NTM loans and PMBS for profit—that these instruments did not assist Fannie and Freddie in meeting the AH goals and therefore must have been acquired because they were profitable. However, the statement by Adolfo Marzol reported above, and the data in Table 5 furnished to the Commission by Fannie Mae shows that all three categories of NTMs—subprime loans (i.e., loans to borrowers with FICO scores less than 660), Alt-A loans and PMBS (called PLS for “Private Label Securities” in the table)— fulfilled the AH goals or subgoals for the years and in the percentages shown below. (Bolded numbers exceeded the applicable goal.) Table 5 also shows, significantly, that the gradual increase in Fannie’s purchases of these NTMs closely followed the gradual increase in the goals between 1996 and 2008.

Table 5.¹⁰⁰ Nontraditional Mortgages and the Affordable Housing Goals

* % of unit financed that qualified for base goals.
^ Not included in housing goals scoring in 2004

Table 5 also shows that ordinary subprime loans, Alt-A loans and PMBS backed by subprime loans were not always sufficient to meet the AH goals. For this reason, Fannie developed special categories of loans in which the firm waived some of its regular underwriting requirements in order to supplement what they were getting from higher quality NTMs. The two principal categories were My Community Mortgage (MCM) and Expanded Approval (EA). In many cases, these two categories enabled Fannie to meet the AH goals, but at the cost of much higher delinquency rates than occurred among higher quality NTMs they acquired. As the years progressed and the AH goals increased, Fannie had to acquire increasing numbers of loans in these categories, and as shown in Table 6 these increasing numbers also exhibited increasing delinquency rates:

Table 6.¹⁰¹ Higher Risk Loans Produced Higher Delinquency Rates at Fannie Mae

Just how desperate Fannie and Freddie were to meet their AH goals is revealed by Fannie’s behavior in 2004. As reported in the American Banker on May 13, 2005, “A House Financial Services Committee report shared with lawmakers Thursday accused Fannie Mae and Freddie Mac of engaging over several years in a series of dubious transactions to meet their affordable-housing goals…The report cited several large transactions entered into by Fannie under which sellers were allowed to repurchase loans without recourse. For example, it said that in September 2003, Fannie bought the option to buy up to $12 billion of multifamily mortgage loans from Washington Mutual, Inc., for a fee of $2 million, the report said. Under the agreement, the GSE permitted WaMu to repurchase the loans…’ This was the largest multifamily transaction ever undertaken by Fannie Mae and was critical for Fannie Mae to reach the affordable-housing goals, the report said.”¹⁰²

A clearer statement of what happened here is contained in WaMu’s 10-K for 2003. Freddie had engaged in a similar but larger transaction with WaMu in 2003, reported as follows in WaMu’s 10-K dated December 31, 2003:

Other noninterest income increased in 2003 compared with 2002 partially due to fees paid to the Company [WaMu] by the Federal Home Loan Mortgage Corporation (“FHLMC” or Freddie Mac”). The Company received $100 million in nonrefundable fees to induce the Company to swap approximately $6 billion of multi-family loans for 100% of the beneficial interest in those loans in the form of mortgage-backed securities issued by Freddie Mac. Since the Company has the unilateral right to collapse the securities after one year, the Company has effectively retained control over the loans. Accordingly, the assets continue to be accounted for and reported as loans. This transaction was undertaken by Freddie Mac in order to facilitate fulfilling its 2003 affordable housing goals as set by the Department of Housing and Urban Development.

Fannie and Freddie were both paying holders of mortgages to temporarily transfer to them possession of goal-qualifying loans that the GSEs could use to satisfy the AH goals for the year 2003. After the end of the year, the seller had an absolute right to reacquire these loans. There can be little doubt, then, that as early as 2003, Fannie and Freddie were under so much pressure to find the subprime or other loans that they needed to meet their affordable housing obligations that they were willing to pay substantial sums to window-dress their reports to HUD.

3. The Affordable Housing Goals were the Sole
Reason That the GSEs’ Acquired So Many NTMs

Up to this point, we have seen that HUD’s policy was to reduce underwriting standards in order to make mortgage credit more readily available to low-income borrowers, and that Fannie and Freddie not only took the AH goals seriously but were willing to go to extraordinary lengths to make sure that they met them. Nevertheless, it seems to have become an accepted idea in some quarters— including in the Commission majority’s report—that Fannie and Freddie bought large numbers of subprime and Alt-A loans between 2004 and 2007 in order to recover the market share they had lost to subprime lenders such as Countrywide or Wall Street, or to make profits. Although there is no evidence whatever for this belief—and a great deal of evidence to the contrary—it has become another urban myth, repeated so often in books, blogs and other media that it has attained a kind of reality.¹⁰³

The formulations of the idea vary a bit. As noted earlier, HUD has claimed— absurdly, in light of its earlier efforts to reduce mortgage underwriting standards— that the GSEs were “chasing the nonprime market” or “chasing market share and profits,” principally between 2004 and 2007. The inference, all too easily accepted, is that this is another example of private greed doing harm, but it is clear that HUD was simply trying to evade its own culpability for using the AH goals to degrade the GSEs’ mortgage underwriting standards over the 15 year period between 1992 and 2007. The Commission majority also adopted a version of this idea in its report, blaming the GSEs’ loosening of their underwriting standards on a desire to please stock market analysts and investors, as well as to increase management compensation. None of HUD’s statements about its efforts to reduce underwriting standards managed to make it into the Commission majority’s report, which relied entirely on the idea that the GSEs’ underwriting standards were reduced by their desire to “follow Wall Street and other lenders in [the] rush for fool’s gold.”

These claims place the blame for Fannie and Freddie’s insolvency—and the huge number of low quality mortgages in the U.S. financial system immediately prior to the financial crisis—on the firms’ managements. They absolve the government, particularly HUD, from responsibility. The GSEs’ managements made plenty of mistakes—and won’t be defended here—but taking risks to compete for market share was not something they actually did. Because of the AH goals, Fannie and Freddie were major buyers of NTMs well before Wall Street firms and the subprime lenders who came to dominate the business entered the subprime PMBS market in any significant way. Moreover, the GSEs did not (indeed, could not) appreciably increase their purchases of NTMs during the years 2005 and 2006, when they had lost market share to the real PMBS issuers, Countrywide and other subprime lenders.

The following discussion addresses each of the claims about the GSEs’ motives in turn, and in the end will show that the only plausible motive for their actions was their effort to comply with HUD’s AH goals.

Did the GSEs acquire NTMs to “compete for market share” with Wall Street or others?

The idea that Fannie and Freddie were newcomers to the purchase of NTMs between 2004 and 2007, and reduced their underwriting standards so they could compete for market share with Wall Street or others, is wrong. As shown in Table 7, the GSEs’ acquisition of subprime loans and other NTMs began in the 1990s, when they first became subject to the AH goals. Research shows that, in contravention of their earlier standards, the GSEs began to acquire high loan-to-value (LTV) mortgages in 1994, shortly after the enactment of the GSE Act and the imposition of the AH goals, and by 2001—before the PMBS market reached $100 billion in annual issuances—the GSEs had already acquired at least $700 billion in NTMs, including over $400 billion in subprime loans.¹⁰⁴ Far from following Wall Street or anyone else into subprime loans between 2004 and 2007, the GSEs had become the largest buyers of subprime and other NTMs many years before the PMBS market began to develop. Given these facts, it would be more accurate to say that Wall Street and the subprime lenders who later came to dominate the PMBS market followed the GSEs into subprime lending. Table 7 does not show any significant increase in the GSEs’ acquisition of NTMs from 2004 to 2007, and the amount of subprime PMBS they acquired during this period actually decreased. This is consistent with the fact—outlined below—that the GSEs did not make any special effort to compete for market share during these years.

Table 7.¹⁰⁵ GSE Purchases of Subprime and Alt-A loans

* Total purchases of PMBS for 1997-2001 are known. Subprime purchases for these years were estimated based upon the percentage that subprime PMBS constituted of total PMBS purchases in 2002 (57%).
** Loans where borrower’s FICO <660.
*** Fannie and Freddie used their various affordable housing programs and individual lender variance programs (many times in conjunction with their automated underwriting systems once these came into general use in the late-1990s) to approve loans with Alt-A characteristics. However, they generally did not classify these loans as Alt-A. Classification as Alt-A started in the early-1990s. There is an unknown number of additional loans that had higher debt ratios, reduced reserves, loosened credit requirements, expanded seller contributions, etc. The volume of these loans is not included.
**** Loans with an original LTV or original combined LTV >90% (given industry practices, this effectively means >=95%). Data to estimate loans with CLTV.>90% is unavailable prior to 2003. Amounts for 2003-2007 are grossed up by 60% to account for the impact of loans with a CLTV >90%. These estimates are based on disclosures by Fannie and Freddie that at the end of 2007 their total exposures to loans with an LTV or CLTV >90% was 50% and 75% percent respectively higher than their exposure to loans with an LTV >90%. Fannie reports on p. 128 of its 2007 10-K that 15% of its entire book had an original combined LTV >90%. Its Original LTV percentage >90% (without counting the impact of any 2nd mortgage simultaneously negotiated) is 9.9%. Freddie reports on p60 of its Q2:2008 10 Q that 14% of its portfolio had an original combined LTV >90%. Its OLTV percentage >90% (without counting any simultaneous 2nd) is 8%. While Fannie and Freddie purchased only the first mortgage, these loans had the same or higher incidence of default as a loan with an LTV of >90%.
***** Since loans may have more than one characteristic, they may appear in more than one category. Totals are not adjusted to take this into account.

The claim that the GSEs loosened their underwriting standards in order to compete specifically with “Wall Street” can be easily dismissed—unless the Commission majority and others who have made this statement are including Countrywide (which was based in California) or other subprime lenders in the term “Wall Street.” Assuming, however, that the Commission majority and other commentators have been using the term Wall Street to apply to the commercial and investment banks that operate in the financial markets of New York, the data shows that Wall Street was not a significant participant in the subprime PMBS market between 2004 and 2007 or at any time before or after those dates. The top five players in 2004 were subprime lenders Ameriquest ($55 billion) and Countrywide ($40 billion), followed by Lehman Brothers ($27 billion), GMAC RFC ($26 billion), and New Century ($22 billion). Other than Lehman, some other Wall Street firms were scattered through the list of the top 25, but were not significant players as a group.

In 2005, the biggest year for subprime issuances, the five leaders were the same, and the total for all Wall Street institutions was $137 billion, or about 27 percent of the $508 billion issued that year.¹⁰⁶ In 2006, Lehman had dropped out of the top five and Countrywide had taken over the leadership among the issuers, but Wall Street’s share had not significantly changed. By the middle of 2007, the PMBS market had declined to such a degree that the market share numbers were meaningless. However, in that year the GSEs’ market share in NTMs increased because they had to continue buying NTMs—even though others had defaulted or left the business—in order to comply with the AH goals. Accordingly, if Fannie had ever loosened its lending standards to compete with some group, that group was not Wall Street.

The next question is whether the GSEs loosened their underwriting standards to compete with Countrywide, Ameriquest and the other subprime lenders who were the dominant players in the PMBS market between 2004 and 2007. Again, the answer seems clearly to be no. The subprime PMBS market was very small until 2002, when for the first time it exceeded $100 billion and reached $134 billion in subprime PMBS issuances.¹⁰⁷ Yet, Table 7 shows that in 2002 alone the GSEs bought $206 billion in subprime loans, more than the total amount securitized by all the subprime lenders and others combined in that year.

The discussion of internal documents that follows will focus almost exclusively on Fannie Mae. The Commission concentrated its investigation on Fannie and it was from Fannie that the Commission received the most complete set of internal documents.

By the early 2000s, Countrywide had succeeded in creating an integrated system of mortgage distribution that included originating, packaging, issuing and underwriting NTMs through PMBS. Other subprime lenders, as noted above, were also major issuers, but they sold their PMBS through Wall Street firms that were functioning as underwriters.

The success of Countrywide and other subprime lenders as distributors of NTMs through PMBS was troubling to Fannie for two reasons. First, Countrywide had been Fannie’s largest supplier of subprime mortgages; the fact that it could now securitize mortgages it formerly sold to Fannie meant that Fannie would have more difficulty finding subprime mortgages that were AH goals-eligible. In addition, the GSEs knew that their support in Congress depended heavily on meeting the AH goals and “leading the market” in lending to low income borrowers. In 2005 and 2006, the Bush administration and a growing number of Republicans in Congress were calling for tighter regulation of Fannie and Freddie, and the GSEs needed allies in Congress to hold this off. The fact that subprime lenders were taking an increasing market share in these years—suggesting that the GSEs were no longer the most important sources of low income mortgage credit—was thus a matter of great concern to Fannie’s management. Without strong support among the Democrats in Congress, there was a significant chance that the Republican Congress would enact tougher regulatory legislation. This was expressed at Fannie as concern about a loss of “relevance,” and provoked wide-ranging consideration within the firm about how they could regain their leadership role in low-income lending.

Nevertheless, although Fannie had strong reasons for wanting to compete for market share with Countrywide and others, it did not have either the operational or financial capacity to do so. In the end, Fannie was unable to take any significant action during the key years 2005 and 2006 that would regain market share from the subprime lenders or anyone else. They reduced their underwriting standards to the degree necessary to keep pace with the increasing AH goals, but not to go significantly beyond those requirements.

In a key memo dated June 27, 2005 (the “Crossroads” memo), Tom Lund, Executive Vice President for Single Family Business, addressed the question of Fannie’s loss of market share and how this share position could be regained. The date of this memo is important. It shows that even in the middle of 2005 there was still a debate going on within Fannie about whether to compete for market share with Countrywide and the other subprime issuers. No such competition had actually begun. Lund starts the discussion in the memo by saying “We are at a strategic crossroad…[his ellipses] We face two stark choices: 1. Stay the Course [or] 2. Meet the Market Where the Market Is”. “Staying the course” meant trying to maintain the mortgage quality standards that Fannie had generally followed up to that point (except as necessary to meet HUD’s AH goals). “Meeting the market” meant competing with Countrywide and others not only by acquiring substantially more NTMs than the AH goals required, but also by acquiring much riskier mortgages than Fannie—which specialized in fixed rate mortgages—had been buying up to that time.

These riskier potential acquisitions would have included much larger numbers of Option ARMs (involving negative amortization) and other loans involving multiple (or “layered”) risks with which Fannie had no prior experience. Thus, Lund noted that to compete in this business Fannie lacked “capabilities and infrastructure…knowledge… willingness to compete on price..[and] a value proposition for subprime.” His conclusion was as stark as the choice: “Realistically, we are not in a position to ‘Meet the Market’ today.” “Therefore,” Lund continued, “we recommend that we: Pursue a ‘Stay the Course’ strategy and test whether market changes are cyclical vs secular.”¹⁰⁸ [emphasis supplied]

In the balance of the Crossroads memo, Lund notes that subprime and Alt-A loans are driving the “leakage” of “goals rich” products to PMBS issuers. He points out the severity of the loss of market share, but never suggests that this changes his view that Fannie was unequipped to compete with Countrywide and others at that time. According to an internal FCIC staff investigation, dated March 31, 2010, other senior officials—Robert Levin (Executive Vice President and Chief Business Officer), Kenneth Bacon (Executive Vice President for Housing and Community Development), and Pamela Johnson (Senior Vice President for Single Family Business)—all concurred that Fannie should follow Lund’s recommendation to “stay the course.”

There is no indication in any of Fannie’s documents after June 2005 that Lund’s “Stay the Course” recommendation was ever changed or challenged during 2005 or 2006—the period when Fannie and Freddie were supposed to have begun to acquire large numbers of NTMs (beyond what was required to meet the AH goals) in order to compete with Countrywide or (in some telling) Wall Street.

Thus, in June 2006, one year after the Lund Crossroads memo, Stephen B. Ashley, then the chairman of the board, told Fannie’s senior executives: “2006 is a transition year. To be sure, there are still issues to resolve. The consent order with OFHEO [among other things, the order raised capital requirements temporarily] is demanding. And from a strategy standpoint, it is clear that until we have eliminated operations and control weaknesses, taking on more risk or opening new lines of business will be viewed dimly by our regulators.”¹⁰⁹ [emphasis supplied] So, again, we have confirmation that Fannie’s top officials did not believe that the firm was in any position—in the middle of 2006—to take on the additional risk that would be necessary s to compete with Countrywide and other subprime lenders that were selling PMBS backed by subprime and other NTMs.

Moreover, there is very strong financially-based evidence that Fannie either never tried or was never financially able to compete for market share with Countrywide and other subprime lenders from 2004 to 2007. For example, set out below are Fannie’s key financial data, published by OFHEO, its former regulator, in early 2008.¹¹⁰

Table 8. Fannie Mae Financial Highlights

Table 8 shows that Fannie’s average guarantee fee increased during the period from 2003 to 2007. To understand the significance of this, it is necessary to understand the way the mortgage business works. Most of Fannie’s guarantee business—the business that competed with securitizations of PMBS by Countrywide and others—was done with wholesale sellers of mortgage pools. In these deals, the wholesaler or issuer, a Countrywide or a Wells Fargo, would assemble a pool of mortgages and look for a guaranty mechanism that would offer the best pricing. In the case of a Fannie MBS, the key issue was the GSEs’ guarantee fee, because that determined how much of the profit the issuer would be able to retain. In the case of a PMBS issue, it was the amount and cost of the credit enhancement needed to attain a AAA rating for a large percentage of the securities backed by the mortgage pool.

The issuer had a choice of securitizing through Fannie, Freddie or one of the Wall Street underwriters. Thus, if Fannie wanted to compete with the private issuers for subprime and other loans there was only one way to do it—by reducing its guarantee fees (called “G-fees” at Fannie and Freddie) and in this way making itself a more attractive outlet than using a Wall Street underwriter. The fact that Fannie does not appear to have done so is strong evidence that it never tried to compete for share with Countrywide and the other subprime issuers after the date of the Crossroads memo in June 2005.

The OFHEO financial summary also shows that Fannie in reality had very little flexibility to compete by lowering its G-fees. Its net income and its return on equity were all declining quickly during this period, and a cut in its G-fees would have hastened this decline.

Finally, there are Fannie’s own reports about its acquisitions of subprime loans. According to Fannie’s 10-K reports for 2004 (which, as restated, covered periods through 2006) and 2007, Fannie’s acquisition of subprime loans barely increased from 2004 through 2007. These are the numbers:

Table 9.¹¹¹ Fannie Mae’s Acquisition of Subprime Loans, 2004-2007

These percentages are consistent with Fannie’s effort to comply with the gradual increase in the AH goals during the years 2004 through 2007; they are not consistent with an effort to substantially increase its purchases of subprime mortgages in order to compete with firms like Countrywide that were growing their market share through securitizing subprime and other loans.

Finally, Fannie’s 2005 10-K (which, as restated and filed in May 2007, also covered 2005 and 2006), contains a statement similar to that made in 2006, confirming that the GSE made no effort to compete for subprime loans (except as necessary to meet the AH goals), and that in fact it lost market share by declining to do so in 2004, 2005 and 2006:

[I]n recent years, an increasing proportion of single-family mortgage loan originations has consisted of non-traditional mortgages such as interest-only mortgages, negative-amortizing mortgages and sub-prime mortgages, and demand for traditional 30-year fixed-rate mortgages has decreased. We did not participate in large amounts of these non-traditional mortgages in 2004, 2005 and 2006 because we determined that the pricing offered for these mortgages often offered insufficient compensation for the additional credit risk associated with these mortgages. These trends and our decision not to participate in large amounts of these non-traditional mortgages contributed to a significant loss in our share of new single-family mortgages-related securities issuances to private-label issuers during this period, with our market share decreasing from 45.0% in 2003 to 29.2% in 2004, 23.5% in 2005 and 23.7 in 2006.¹¹² [emphasis supplied]

Accordingly, despite losing market share to Countrywide and others in 2004, 2005 and 2006, Fannie did not attempt to acquire unusual numbers of subprime loans in order to regain this share. Instead, it continued to acquire only the subprime and other NTM loans that were necessary to meet the AH goals. That the AH goals were Fannie’s sole motive for acquiring NTMs is shown by the firm’s actions after the PMBS market collapsed in 2007. At that point, Fannie’s market share began to rise as Countrywide and others could not continue to issue PMBS. Nevertheless, despite the losses on subprime loans that were beginning to show up in the markets, Fannie continued to buy NTMs until they were taken over by the government in September 2008. The reason for this nearly reckless behavior is obvious—they were still subject to the AH goals, which were increasing through this period. If they had only acquired these NTMs to compete with Countrywide and others for market share the competition was already over; their competitors had abandoned the field. But the fact is that Fannie did not—or could not—increase its market share between 2004 and 2006 shows without question that market share was not the reason they had acquired so many NTMs by the time they failed in September 2008.

Beleaguered by accounting problems, suffering diminished profitability, and lacking the capability to evaluate the risks of the new kinds of mortgages they would have to buy, Fannie had no option but to stay the course they had been following for 15 years. The NTMs they bought during the period from 2004 to 2007 were acquired to comply with the AH goals and not to increase their market share—as much as Fannie might have preferred to do so. Fannie’s market share finally did increase in 2007, when the asset-backed market collapsed, Countrywide weakened, and neither Countrywide nor anyone else could continue to securitize mortgages. In a report to the board of directors on October 16, 2007, Mudd reported that Fannie’s market share, which was 20 percent of the whole market at the beginning of 2007, had risen to 42 percent.¹¹³

That leaves one other possibility—that Fannie and Freddie were buying NTMs because they were profitable. That issue is addressed in the next section.

Did Fannie acquire NTMs because these loans were profitable?

From time to time, commentators on the GSEs have suggested that the GSEs’ real motive for acquiring NTMs was not that they had to comply with the AH goals, but that they were seeking the profits these risky loans produced. This could have been true in the 1990s, but after the major increase in the AH goals in 2000 Fannie began to recognize that complying with the goals was reducing the firm’s profitability. By 2007, Fannie was asking for relief from the goals.

The following table, drawn from a FHFA publication, shows the applicable AH goals over the period from 1996 through 2008 and the GSEs’ success in meeting them.

Table 10.¹¹⁴ GSEs’ Success in Meeting Affordable Housing Goals, 1996-2007

As the table shows, Fannie and Freddie exceeded the AH goals virtually each year, but not by significant margins. They simply kept pace with the increases in the goals as these requirements came into force over the years. This alone suggests that they did not increase their purchases in order to earn profits. If that was their purpose they would have substantially exceeded the goals, since their financial advantages (low financing costs and low capital requirements) allowed them to pay more for the mortgages they wanted than any of their competitors. As HUD noted in 2000: “Because the GSEs have a funding advantage over other market participants, they have the ability to underprice their competitors and increase their market share.”¹¹⁵

As early as 1999, there were clear concerns at Fannie about how the 50 percent LMI goal—which HUD had signaled as its next move—would be met. In a June 15, 1999, memorandum,¹¹⁶ four Fannie staff members proposed three categories of rules changes that would enable Fannie to meet the goals more easily: (i) persuade HUD to change the goals accounting (what goes into the numerator and denominator); (ii) enter other businesses where the pickings might be goals-rich, such as manufactured housing and, significantly, Alt-A and subprime (“Efforts to expand into Alt-A and A-markets (the highest grade of subprime lending) should also yield incremental business that will have a salutary effect on our low-and moderate-income score”); and (iii) persuade HUD to adopt different methods of goals scoring.

By 2000, Fannie was effectively in competition with banks that were required to make mortgage loans under CRA to roughly the same population of low-income borrowers targeted in HUD’s AH goals. Rather than selling their CRA loans to Fannie and Freddie, banks and S&Ls had begun to retain the loans in portfolio. In a presentation in November 2000, Barry Zigas, a Senior Vice President of Fannie, noted that “Our own anecdotal evidence suggests that this increase [in banks’ and S&Ls’ holding loans in portfolio] is due in part to below-market CRA products.”¹¹⁷ In other words, banks and S&Ls subject to CRA were making mortgage loans at below market interest rates, and thus could not sell them without taking losses. This was troubling for Fannie because it meant that in order to capture these loans they would have to increase what they were willing to pay for these loans. Doing so would underprice the risks they would be assuming.

It is important to recognize what was happening. Fannie, and the banks and S&Ls under CRA, were now competing for the same kinds of NTMs, and were doing so by lowering their mortgage underwriting standards and adding flexibilities and subsidies. Simply as a result of supply and demand, all of the participants in this competition were required to pay higher prices for these increasingly risky mortgages. The banks and S&Ls that acquired these loans could not sell them, without taking a loss, when market interest rates were higher than the rates on the mortgages. This is the first indication in the documents that the FCIC received from Fannie that competition for subprime loans among the GSEs, banks, S&Ls, and FHA was causing the underpricing of risk—one of the principal causes of the mortgage meltdown and thus the financial crisis.

In January 2003, Fannie began planning for how to confront HUD before the next round of increases in the AH goals, expected to occur in 2004. In an “Action Plan for the Housing Goals Rewrite,” dated January 22, 2003, Fannie staff reviewed a number of options, and concluded that “Fannie should strongly oppose: goals increases and new subgoals.” (Slide 35)¹¹⁸

In March 2003, as Fannie prepared for new increases in the AH goals, its staff prepared a presentation, perhaps for HUD or for policy defense in public forums. The apparent purpose was to show that the goals should not be increased significantly in 2004. Slide 5 stated:

In 2002, Fannie Mae exceeded all our goals for the 9th straight year. But it was probably the most challenging environment we’ve ever faced. Meeting the goals required heroic 4th quarter efforts on the part of many across the company. Vacations were cancelled. The midnight oil burned. Moreover, the challenge freaked out the business side of the house. Especially because the tenseness around meeting the goals meant that we considered not doing deals—not fulfilling our liquidity function—and did deals at risks and prices we would not have otherwise done.¹¹⁹ [emphasis supplied]

By September 2004, it was becoming clear that continuing increases in the AH goals were having a major adverse effect on Fannie’s profitability. In a memorandum to Brian Graham (another Fannie official), Paul Weech, Director of Market Research and Policy Development, wrote: “Meeting the goals in difficult markets imposes significant costs on the Company and potentially causes market-distorting behaviors. In 1998, 2002, and 2003 especially, the Company has had to pursue certain transactions as much for housing goals attainment as for the economics of the transaction.”¹²⁰

In a June 2005 presentation entitled “Costs and Benefits of Mission Activities,”¹²¹ the authors noted in slide 10 that AH goal costs had risen from $2,632,500 in 2000 to $13,447,500 in 2003. Slide 17 is entitled: “Meeting Future HUD Goals Appear Quite Daunting and Potentially Costly” and reports, “Based on 2003 experience where goal acquisition costs (relative to Fannie Mae model fees) cost between $65 per goals unit in the first quarter to $370 per unit in the fourth quarter, meeting the shortfall could cost the company $6.5-$36.5 million to purchase sufficient units.” The presentation concludes (slide 20): “Cost of mission activities— explicit and implicit—over the 2000-2004 period likely averaged approximately $200 million per year.”

Earlier, I noted the efforts of Fannie and Freddie to window-dress their records for HUD by temporarily acquiring loans that would comply with the AH goals, while giving the seller the option to reacquire the loans at a later time. In 2005, we begin to see efforts by Fannie’s staff to accomplish the same window-dressing in another way--delaying acquisitions of non-goal-eligible loans so Fannie can meet the AH goals in that year; we also see the first efforts to calculate systematically the effect of goal-compliance on Fannie’s profitability. In a presentation dated September 30, 2005, Barry Zigas, the key Fannie official on affordable housing, outlined a “business deferral option.” Under that initiative, Fannie would ask seven major lenders to defer until 2006 sending non-goal loans to Fannie for acquisition. This would reduce the denominator of the AH goal computation and thus bring Fannie nearer to goal compliance in the 4th quarter of 2005. The cost of the deferral alone was estimated at $30-$38 million.¹²²

In a presentation to HUD on October 31, 2005, entitled “Update on Fannie Mae’s Housing Goals Performance,”¹²³ Fannie noted several “Undesirable Tradeoffs Necessary to Meet Goals.” These included significant additional credit risk, and negative returns (“Deal economics are well below target returns; some deals are producing negative returns” and “G-fees may not cover expected losses”). One of the most noteworthy points was the following: “Liquidity to Questionable Products: Buying exotic product encourages continuation of risky lending; many products present with significant risk-layering; consumers are at risk of payment shock and loss of equity; potential need to waive our responsible lending policies to get goals business.”

Much of the narrative about the financial crisis posits that unscrupulous and unregulated mortgage originators tricked borrowers into taking on bad mortgages. The idea that predatory lending was a major source of the NTMs in the financial system in 2008 is a significant element of the Commission majority’s report, although the Commission was never able to provide any data to support this point. This Fannie slide suggests that loans later dubbed “predatory” might actually have been made to comply with the AH goals. This possibility is suggested, too, in a message sent in 2004 to Freddie’s CEO, Richard Syron, by Freddie’s chief risk manager, David Andrukonis, when Syron was considering whether to authorize a “Ninja” (no income/no jobs/no assets) product that he ultimately approved. Andrukonis argued against authorizing Freddie’s purchase: “The potential for the perception and reality of predatory lending, with this product is great.”¹²⁴ But the product was approved by Freddie, probably for the reason stated by another Freddie employee: “The Alt-A [(low doc/no doc)] business makes a contribution to our HUD goals.”¹²⁵

On May 5, 2006, a Fannie staff memo to the Single Family Business Credit Committee revealed the serious credit and financial problems Fannie was facing when acquiring subprime mortgages to meet the AH goals. The memo describes the competitive landscape, in which “product enhancements from Freddie Mac, FHA, Alt-A and subprime lenders have all contributed to increased competition for goals rich loans…On the issue of seller contributions [in which the seller of the home pays cash expenses for the buyer] even FHA has expanded their guidelines by allowing 6% contributions for LTVs up to 97% that can be used toward closing, prepaid expenses, discount points and other financing concessions.”¹²⁶

The memorandum is eye-opening for what it says about the credit risks Fannie had to take in order to get the goals-rich loans it needed to meet HUD’s AH requirements for 2006. Table 11 below shows the costs of NTMs in terms of the guarantee fee (G-fee) “gap.” (In order to determine whether a loan contributed to a return on equity, Fannie used a G-fee pricing model that took into account credit risk as well as a number of other factors; a G-fee “gap” was the difference between the G-fees required by the pricing model for a particular loan to contribute to a return on equity and a loan that did not.) The table in this memo shows the results for three subprime products under consideration, a 30 year fixed rate mortgage (FRM), a 5 year ARM, and 35 and 40 year fixed rate mortgages. For simplicity, this analysis will discuss only the 30 year fixed rate product. The table shows that the base product, the 30 year FRM, with a zero downpayment should be priced according to the model at a G-fee of 106 basis points. However, the memo reports that Fannie is actually buying loans like that at a price consistent with an annual fee of 37.50 basis points, producing a gap (or loss from the model) of 68.50 basis points. The reason the gap is so large is shown in the table: the anticipated default rate on that zero-down mortgage was 34 percent. The table then goes on to look at other possible loan alternatives, with the following results:

Table 11.¹²⁷ Fannie Mae Took Losses on Higher Risk Mortgages

From this report, it is clear that in order to meet the AH goals Fannie had to pay up for goals-rich mortgages, taking a huge credit risk along the way.

The dismal financial results that were developing at Fannie as a result of the AH goals were also described in Fannie’s 10-K report for 2006, which anticipated both losses of revenue and higher credit losses as a result of acquiring the mortgages required by the AH goals:

[W]e have made, and continue to make, significant adjustments to our mortgage loan sourcing and purchase strategies in an effort to meet HUD’s increased housing goals and new subgoals. These strategies include entering into some purchase and securitization transactions with lower expected economic returns than our typical transactions. We have also relaxed some of our underwriting criteria to obtain goals-qualifying mortgage loans and increased our investments in higher-risk mortgage loan products that are more likely to serve the borrowers targeted by HUD’s goals and subgoals, which could increase our credit losses. [emphasis supplied]¹²⁸

The underlying reasons for the “lower expected returns” were reported in February 2007 in a document the FCIC received from Fannie, which noted that for 2006 the “cash flow cost” of meeting the housing goals was $140 million while the “opportunity cost” was $470 million.¹²⁹ In a report to HUD on the AH goals, dated April 11, 2007, Fannie described these costs as follows: “The largest costs [of meeting the goals] are opportunity costs of foregone revenue. In 2006, opportunity cost was about $400 million, whereas the cash flow cost was about $134 million. If opportunity cost was $0, our shareholders would be indifferent to the deal. The cash flow cost is the implied out of pocket cost.”¹³⁰

By this time, “Alignment Meetings”—in which Fannie staff considered how they would meet the AH goals—were taking place almost monthly (according to the frequency with which presentations to Alignment meetings occur in the documentary record). In an Alignment Meeting on June 22, 2007, on a “Housing Goals Forecast,” three plans were considered for meeting the 2007 AH goals, even though half the year was already gone. One of the plans was forecast to result in opportunity costs of $767.7 million, while the other two plans resulted in opportunity costs of $817.1 million.¹³¹ In a Forecast meeting on July 27, 2007, a “Plan to Meet Base Goals,” which probably meant the topline LMI goal including all subgoals, was placed at $1.156 billion for 2007.¹³²

Finally, in a December 21, 2007, letter to Brian Montgomery, Assistant Secretary of Housing, Fannie CEO Daniel Mudd asked that, in light of the financial and economic conditions then prevailing in the country—particularly the absence of a PMBS market and the increasing number of mortgage delinquencies and defaults— HUD’s AH goals for 2007 be declared “infeasible.” He noted that HUD also has an obligation to “consider the financial condition of the enterprise when determining the feasibility of goals.” Then he continued: “Fannie Mae submits that the company took all reasonable actions to meet the subgoals that were both financially prudent and likely to contribute to the achievement of the subgoals….In 2006, Fannie Mae relaxed certain underwriting standards and purchased some higher risk mortgage loan products in an effort to meet the housing goals. The company continued to purchase higher risk loans into 2007, and believes these efforts to acquire goals-rich loans are partially responsible for increasing credit losses.”¹³³ [emphasis supplied]

This statement confirms two facts that are critical on the question of why Fannie (and Freddie) acquired so many high risk loans in 2006 and earlier years: first, the companies were trying to meet the AH goals established by HUD and not because these loans were profitable. It also shows that the efforts of HUD and others— including the Commission majority in its report—to blame the managements of Fannie and Freddie for purchasing the loans that ultimately dragged them to insolvency is misplaced.

Finally, in a July 2009 report, the Federal Housing Finance Agency (FHFA, the GSEs’ new regulator, replacing OFHEO), noted that Fannie and Freddie both followed the practice of cross-subsidizing the subprime and Alt-A loans that they acquired:

Although Fannie Mae and Freddie Mac consider model-derived estimates of cost in determining the single-family guarantee fees they charge, their pricing often subsidizes their guarantees on some mortgages using higher returns they expect to earn on guarantees of other loans. In both 2007 and 2008, cross-subsidization in single-family guarantee fees charged by the Enterprises was evident across product types, credit score categories, and LTV ratio categories. In each case, there were cross-subsidies from mortgages that posed lower credit risk on average to loans that posed higher credit risk. The greatest estimated subsidies generally went to the highest-risk mortgages.¹³⁴

The higher risk mortgages were the ones most needed by Fannie and Freddie to meet the AH goals. Needless to say, there is no need to cross-subsidize the G-fees of loans that are acquired because they are profitable.

Accordingly, both market share and profitability must be excluded as reasons that Fannie (and Freddie) acquired subprime and Alt-A loans between 2004 and 2007. Th e only remaining motive—and the valid one—was the eff ect of the AH goals imposed by HUD.

In 2008, after its takeover by the government, Fannie Mae finally published a credit supplement to its 2008 10-K, which contained an accounting of its subprime and Alt-A credit exposure. The table is reproduced below in order to provide a picture of the kinds of loans Fannie acquired in order to meet the AH goals. Loans may appear in more than one category, so the table does not reveal Fannie’s total net exposure to each category, nor does it include Fannie’s holdings of non-Fannie MBS or PMBS, for which it did not have loan level data. Note the reference to $8.4 billion in the column for subprime loans. As noted earlier, Fannie classified as subprime only those loans that it purchased from subprime lenders. However, Fannie included loans with FICO scores of less than 660 in the table, indicating that they are not prime loans but without classifying them formally as subprime.

In a later credit supplement, filed in August 2009, Fannie eliminated the duplications among the loans in Table 12, and reported that as of June 30, 2009, it held the credit risk on NTMs with a total unpaid principal amount of $2.7 trillion. The average loan amount was $151,000, for a total of 5.73 million NTM loans.¹³⁵ This number does not include Fannie’s holdings of subprime PMBS as to which it does not have loan level data.

Table 12. Fannie Mae Credit Profile by Key Product Features: Credit Characteristics of Single-Family Conventional Mortgage Credit Book of Business¹³⁶

⁽¹⁾ Alt-A, Subprime, and Jumbo Conforming Loans are calculated as a percentage of the single-family mortgage credit book of business, which includes government loans. Government loans are guaranteed or insured by the U.S. Government or its agencies, such as the Department of Veterans Affairs (VA), the Federal Housing Administration (FHA) or the Rural Housing and Community Facilities Program and the Department of Agriculture.
⁽²⁾ Unpaid principal balance of all loans with credit enhancement as a percentage of unpaid principal balance of single-family conventional mortgage credit book of business. Includes primary mortgage insurance, pool insurance, lender recourse and other credit enhancement.
⁽³⁾ Expressed as a percentage of credit losses for the single-family mortgage credit book of business. For information on total credit losses, refer to Fannie Mae’s 2008 Form 10-K.

Freddie Mac. As noted earlier, in its limited review of the role of the GSEs in the financial crisis, the Commission spent most of its time and staff resources on a review of Fannie Mae, and for that reason this dissent focuses primarily on documents received from Fannie. However, things were not substantially different at Freddie Mac. In a document dated June 4, 2009, entitled “Cost of Freddie Mac’s Affordable Housing Mission,” a report to the Business Risk Committee, of the Board of Directors,136 several points were made that show the experience of Freddie was no different than Fannie’s:

• Our housing goals compliance required little direct subsidy prior to 2003, but since then subsidies have averaged $200 million per year.
• Higher credit risk mortgages disproportionately tend to be goal-qualifying. Targeted affordable lending generally requires ‘accepting’ substantially higher credit risk.
• We charge more for targeted (and baseline) affordable single-family loans, but not enough to fully offset their higher incremental risk.
• Goal-qualifying single-family loans accounted for the disproportionate share of our 2008 realized losses that was predicted by our models. (slide 2)
• In 2007 Freddie Mac failed two subgoals, but compliance was subsequently deemed infeasible by the regulator due to economic conditions. In 2008 Freddie Mac failed six goals and subgoals, five of which were deemed infeasible. No enforcement action was taken regarding the sixth missed goal because of our financial condition. (slide 3)
• Goal-qualifying loans tend to be higher risk. Lower household income correlates with various risk factors such as less wealth, less employment stability, higher loan-to-value ratios, or lower credit scores. (slide 7)
• Targeted affordable loans have much higher expected default probabilities... Over one-half of targeted affordable loans have higher expected default probabilities than the highest 5% of non-goal-qualifying loans. (Slide 8)

The use of the affordable housing goals to force a reduction in the GSEs’ underwriting standards was a major policy error committed by HUD in two successive administrations, and must be recognized as such if we are ever to understand what caused the financial crisis. Ultimately, the AH goals extended the housing bubble, infused it with weak and high risk NTMs, caused the insolvency of Fannie and Freddie, and—together with other elements of U.S. housing policy—was the principal cause of the financial crisis itself.

When Congress enacted the Housing and Economic Recovery Act of 2008 (HERA), it transferred the responsibility for administering the affordable housing goals from HUD to FHFA. In 2010, FHFA modified and simplified the AH goals, and eliminated one of their most troubling elements. As Fannie had noted, if the AH goals exceed the number of goals-eligible borrowers in the market, they were being forced to allocate credit, taking it from the middle class and providing it to low-income borrowers. In effect, there was a conflict between their mission to advance affordable housing and their mission to maintain a liquid secondary mortgage market for most mortgages in the U.S. Th e new FHFA rule does not require the GSEs to purchase more qualifying loans than the percentage of the total market that these loans constitute.¹³⁷

This does not solve all the major problems with the AH goals. In the sense that the goals enable the government to direct where a private company extends credit, they are inherently a form of government credit allocation. More significantly, the competition among the GSEs, FHA and the banks that are required under the CRA to find and acquire the same kind of loans will continue to cause the same underpricing of risk on these loans that eventually brought about the mortgage meltdown and the financial crisis. This is discussed in the next section and the section on the CRA.

4. Competition Between the GSEs and FHA
for Subprime and Alt-A Mortgages

One of the important facts about HUD’s management of the AH goals was that it placed Fannie and Freddie in direct competition with FHA, an agency within HUD. This was already noted in some of the Fannie documents cited above. Fannie treated this as a conflict of interest at HUD, but there is a strong case that this competition is exactly what HUD and Congress wanted. It is important to recall the context in which the GSE Act was enacted in 1992. In 1990, Congress had enacted the Federal Credit Reform Act.¹³⁸ One of its purposes was to capture in the government’s budget the risks to the government associated with loan guarantees, and in effect it placed a loose budgetary limit on FHA guarantees. For those in Congress and at HUD who favored increased mortgage lending to low income borrowers and underserved communities, this consequence of the FCRA may have been troubling. What had previously been a free way to extend support to groups who were not otherwise eligible for conventional mortgages—which generally required a 20 percent downpayment and the indicia of willingness and ability to pay—now appeared to be potentially restricted. Requiring the GSEs to take up the mantle of affordable housing would have looked at the time like a solution, since Fannie and Freddie had unlimited access to funds in the private markets and were off-budget entities.

Looked at from this perspective, it would make sense for Congress and HUD to place the GSEs and FHA in competition, just as it made sense to put Fannie and Freddie in competition with one another for affordable loans. With all three entities competing for the same kinds of loans, and with HUD’s control of both FHA’s lending standards and the GSEs’ affordable housing requirements, underwriting requirements would inevitably be reduced. HUD’s explicit and frequently expressed interest in reducing mortgage underwriting standards, as a means of making mortgage credit available to low income borrowers, provides ample evidence of HUD’s motives for creating this competition.

Established in 1934, now a part of HUD and administered by the Federal Housing Administrator (who is also the Assistant Secretary of Housing), FHA insures 100 percent of an eligible mortgage. It was established to provide financing to people who could not meet the standards for a bank-originated conventional loan. The loans it insured had a maximum LTV of 80 percent in 1934. This went to 95 percent in 1950, and 97 percent in 1961.¹³⁹ With its maximum LTV remaining at 97 percent, FHA maintained average FICO scores for its borrowers just below 660 from 1996 to 2006. During this period, the average FICO score for a conventional subprime borrower was somewhat lower.¹⁴⁰ Beginning in 1993, shortly after Fannie and Freddie were introduced as competitors, FHA began to increase its percentage of loans with low downpayments. This had the predictable effect on its delinquency rates, as shown in the figure below prepared by Edward Pinto with data from FHA, the FDIC, and the MBA:

Figure 6.

Despite its reductions in required downpayments, FHA’s market share vis-a vis the GSEs began to decline. According to GAO data, in 1996, FHA’s market share among lower-income borrowers was 26 percent while the GSEs’ share was 23.8 percent. By 2005, FHA’s share was 9.8 percent, while the GSEs’ share was 31.9 percent. It appears that early on Fannie Mae deliberately targeted FHA borrowers with its Community Homebuyer Program (CHBP). In a memorandum prepared in 1993, Fannie’s Credit Policy group compared Fannie’s then-proposed CHBP program to FHA’s requirements under its 1-to-4 family loan program (Section 203(b)) and showed that most of Fannie’s requirements were competitive or better.

FHA also appears to have tried to lead the GSEs. In 1999—just before the AH goals for Fannie and Freddie were to be raised—FHA almost doubled its originations of loans with LTVs equal to or greater than 97 percent, going from 22.9 percent in 1998 to 43.84 percent in 1999.¹⁴¹ It also offered additional concessions on underwriting standards in order to attract subprime business. The following is from a Quicken ad in January 2000 (emphasis in the original),¹⁴² which is likely to have been based on an FHA program as it existed in 1999:

Borrowers can purchase with a minimum down payment. Without FHA insurance, many families can’t afford the homes they want because down payments are a major roadblock. FHA down payments range from 1.25% to 3% of the sale price and are significantly lower than the minimum that many lenders require for conventional or sub-prime loans.

With FHA loans, borrowers need as little as 3% of the “total funds” required. In addition to the funds needed for the down payment, borrowers also have to pay closing costs, prepaid fees for insurance and interest, as well as escrow fees which include mortgage insurance, hazard insurance, and months worth of property taxes. A FHA-insured home loan can be structured so borrowers don’t pay more than 3% of the total out-of-pocket funds, including the down payment.

The combined total of out-of-pocket funds can be a gift or loan from family members. FHA allows homebuyers to use gifts from family members and non-profit groups to cover their down payment and additional closing costs and fees. In fact, even a 100% gift or a personal loan from a relative is acceptable.

FHA’s credit requirements are flexible. Compared to credit requirements established by many lenders for other types of home loans, FHA focuses only on a borrower’s last 12-24 month credit history. In addition, there is no minimum FICO score -mortgage bankers look at each application on a case-by-case basis. It is also perfectly acceptable for people with NO established credit to receive a loan with this program.

FHA permits borrowers to have a higher debt-to-income ratio than most insurers typically allow. Conventional home loans allow borrowers to have 36% of their gross income attributed to their new monthly mortgage payment combined with existing debt. FHA program allows borrowers to carry 41%, and in some circumstances, even more.

It is important to remember that 1999 is the year that HUD was planning a big step-up in the AH goals for the GSEs—from 42 percent LMI to 50 percent, with even larger percentage increases in the special affordable category that would be most competitive with FHA. The last major increase in the percent of FHA’s loans with LTVs equal to or greater than 97 percent had occurred in 1991, the year before the GSE Act imposed the AH goals on Fannie and Freddie, and in effect directed them to consider downpayments of 5 percent or less. In 1991, FHA’s percentage of loans equal to or greater than 97 percent rose suddenly from 4.4 percent to 17.1 percent.¹⁴³

Again, FHA, under the control of HUD, appears to be offering competition to the GSEs that would lead them to reduce their underwriting standards. Since FHA is a government agency, its actions cannot be explained by a profit motive. Instead, it seems clear that FHA reduced its lending standards as part of a HUD policy to lead Fannie and Freddie in the same direction.

The result of Fannie’s competition with FHA in high LTV lending is shown in the following figure, which compares the respective shares of FHA and Fannie in the category of loans with LTVs equal to or greater than 97 percent, including Fannie loans with a combined LTV equal to or greater than 97 percent.

Figure 7.

Whether a conscious policy of HUD or not, competition between the GSEs and FHA ensued immediately after the GSEs were given their affordable housing mission in 1992. The fact that FHA, an agency controlled by HUD, substantially increased the LTVs it would accept in 1991 (just before the GSEs were given their affordable housing mission) and again in 1999 (just before the GSEs were required to increase their affordable housing efforts) is further evidence that HUD was coordinating these policies in the interest of creating competition between FHA and the GSEs. The effect was to drive down underwriting standards, which HUD had repeatedly described as its goal.

5. Enlisting Mortgage Bankers and Subprime
Lenders in Affordable Housing

In 1994, HUD began a program to enlist other members of the mortgage financing community in the effort to reduce underwriting standards. In that year, the Mortgage Bankers Association (MBA)—a group of mortgage financing firms not otherwise regulated by the federal government and not subject to HUD’s legal authority—agreed to join a HUD program called the “Best Practices Initiative.”¹⁴⁴ The circumstances surrounding this agreement are somewhat obscure, but at least one contemporary account suggests that the MBA signed up to avoid an effort by HUD to cover mortgage bankers under the Community Reinvestment Act (CRA), which up to that point had only applied only to government-insured banks.

In mid-September [1994], the Mortgage Bankers Association of America-whose membership includes many bank-owned mortgage companies, signed a three-year master best-practices agreement with HUD. The agreement consisted of two parts: MBA’s agreement to work on fair-lending issues in consultation with HUD and a model best-practices agreement that individual mortgage banks could use to devise their own agreements with HUD. The first such agreement, signed by Countrywide Funding Corp., the nation’s largest mortgage bank, is summarized [below]. Many have seen the MBA agreement as a preemptive strike against congressional murmurings that mortgage banks should be pulled under the umbrella of the CRA.¹⁴⁵

As the first member of the MBA to sign, Countrywide probably realized that there were political advantages in being seen as assisting low-income mortgage lending, and it became one of a relatively small group of subprime lenders who were to prosper enormously as Fannie and Freddie began to look for sources of the subprime loans that would enable them to meet the AH goals. By 1998, there were 117 MBA signatories to HUD’s Best Practices Initiative, which was described as follows:

The companies and associations that sign “Best Practices” Agreements not only commit to meeting the responsibilities under the Fair Housing Act, but also make a concerted effort to exceed those requirements. In general, the signatories agree to administer a review process for loan applications to ensure that all applicants have every opportunity to qualify for a mortgage. They also assent to making loans of any size so that all borrowers may be served and to provide information on all loan programs for which an applicant qualifies…. The results of the initiative are promising. As lenders discover new, untapped markets, their minority and low-income loans applications and originations have risen. Consequently, the homeownership rate for low-income and minority groups has increased throughout the nation.¹⁴⁶

Countrywide was by far the most important participant in the HUD program. Under that program, it made a series of multi-billion dollar commitments, culminating in a “trillion dollar commitment” to lend to minority and low income families, which in part it fulfilled by selling subprime and other NTMs to Fannie and Freddie. In a 2000 report, the Fannie Mae Foundation noted: “FHA loans constituted the largest share of Countrywide’s activity, until Fannie Mae and Freddie Mac began accepting loans with higher LTVs and greater underwriting flexibilities.”¹⁴⁷ In late 2007, a few months before its rescue by Bank of America, Countrywide reported that it had made $789 billion in mortgage loans toward its trillion dollar commitment.¹⁴⁸

6. The Community Reinvestment Act

The most controversial element of the vast increase in NTMs between 1993 and 2008 was the role of the CRA.¹⁴⁹ The act, which is applicable only to federally insured depository institutions, was originally adopted in 1977. Its purpose in part was to “require each appropriate Federal financial supervisory agency to use its authority when examining financial institutions to encourage such institutions to help meet the credit needs of the local communities in which they are chartered consistent with the safe and sound operations of such institutions.” The enforcement provisions of the Act authorized the bank regulators to withhold approvals for such transactions as mergers and acquisitions and branch network expansion if the applying bank did not have a satisfactory CRA rating.

CRA did not have a substantial effect on subprime lending in the years after its enactment until the regulations under the act were tightened in 1995. The 1995 regulations required insured banks to acquire or make “flexible and innovative” mortgages that they would not otherwise have made. In this sense, the CRA and Fannie and Freddie’s AH goals are cut from the same cloth.

There were two very distinct applications of the CRA. The first, and the one with the broadest applicability, is a requirement that all insured banks make CRA loans in their respective assessment areas. When the Act is defended, it is almost always discussed in terms of this category—loans in bank assessment areas. Banks (usually privately) complain that they are required by the regulators to make imprudent loans to comply with CRA. One example is the following statement by a local community bank in a report to its shareholders:

Under the umbrella of the Community Reinvestment Act (CRA), a tremendous amount of pressure was put on banks by the regulatory authorities to make loans, especially mortgage loans, to low income borrowers and neighborhoods. The regulators were very heavy handed regarding this issue. I will not dwell on it here but they required [redacted name] to change its mortgage lending practices to meet certain CRA goals, even though we argued the changes were risky and imprudent.¹⁵⁰

On the other hand, the regulators defend the act and their actions under it, and particularly any claim that the CRA had a role in the financial crisis. The most frequently cited defense is a speech by former Fed Governor Randall Kroszner on December 3, 2008,¹⁵¹ in which he said in pertinent part:

Only 6 percent of all the higher-priced loans [those that were considered CRA loans because they bore high interest rates associated with their riskier character] were extended by CRA-covered lenders to lower-income borrowers or neighborhoods in their assessment areas, the local geographies that are the primary focus for CRA evaluation purposes. This result undermines the assertion by critics of the potential for a substantial role for the CRA in the subprime crisis. [emphasis supplied]

There are two points in this statement that require elaboration. First, it assumes that all CRA loans are high-priced loans. This is incorrect. Many banks, in order to be sure of obtaining the necessary number of loans to attain a satisfactory CRA rating, subsidized the loans by making them at lower interest rates than their risk characteristics would warrant. This is true, in part, because CRA loans are generally loans to low income individuals; as such, they are more likely than loans to middle income borrowers to be subprime and Alt-A loans and thus sought after by FHA, Fannie and Freddie and subprime lenders such as Countrywide; this competition is another reason why their rates are likely to be lower than their risk characteristics. Second, while bank lending under CRA in their assessment areas has probably not had a major effect on the overall presence of subprime loans in the U.S. financial system, it is not the element about CRA that raises the concerns about how CRA operated to increase the presence of NTMs in the housing bubble and in the U.S. financial system generally. There is another route through which CRA’s role in the financial crisis likely to be considerably more significant.

In 1994, the Riegle-Neal Interstate Banking and Branching Efficiency Act for the first time allowed banks to merge across state lines under federal law (as distinct from interstate compacts). Under these circumstances, the enforcement provisions of the CRA, which required regulators to withhold approvals of applications for banks that did not have satisfactory CRA ratings, became particularly relevant for large banks that applied to federal bank regulators for merger approvals. In a 2007 speech, Fed Chairman Ben Bernanke stated that after the enactment of the Riegle-Neal legislation, “As public scrutiny of bank merger and acquisition activity escalated, advocacy groups increasingly used the public comment process to protest bank applications on CRA grounds. In instances of highly contested applications, the Federal Reserve Board and other agencies held public meetings to allow the public and the applicants to comment on the lending records of the banks in question. In response to these new pressures, banks began to devote more resources to their CRA programs.”¹⁵² This modest description, although accurate as far as it goes, does not fully describe the effect of the law and the application process on bank lending practices.

In 2007, the umbrella organization for many low-income or community “advocacy groups,” the National Community Reinvestment Coalition, published a report entitled “CRA Commitments” which recounted the substantial success of its members in using the leverage provided by the bank application process to obtain trillions of dollars in CRA lending commitments from banks that had applied to federal regulators for merger approvals. The opening section of the report states (bolded language in the original):¹⁵³

Since the passage of CRA in 1977, lenders and community organizations have signed over 446 CRA agreements totaling more than $4.5 trillion in reinvestment dollars flowing to minority and lower income neighborhoods.

Lenders and community groups will often sign these agreements when a lender has submitted an application to merge with another institution or expand its services. Lenders must seek the approval of federal regulators for their plans to merge or change their services. The four federal financial institution regulatory agencies will scrutinize the CRA records of lenders and will assess the likely future community reinvestment performance of lenders. The application process, therefore, provides an incentive for lenders to sign CRA agreements with community groups that will improve their CRA performance. Recognizing the important role of collaboration between lenders and community groups, the federal agencies have established mechanisms in their application procedures that encourage dialogue and cooperation among the parties in preserving and strengthening community reinvestment. [emphasis supplied]

A footnote to this statement reports:

The Federal Reserve Board will grant an extension of the public comment period during its merger application process upon a joint request by a bank and community group. In its commentary to Regulation Y, the Board indicates that this procedure was added to facilitate discussions between banks and community groups regarding programs that help serve the convenience and needs of the community. In its Corporate Manual, the Office of the Comptroller of the Currency states that it will not offer the expedited application process to a lender that does not intend to honor a CRA agreement made by the institution that it is acquiring.

In its report, the NCRC listed all 446 commitments and includes the following summary list of year-by-year commitments:

Table 13.

The size of these commitments, which far outstrip the CRA loans made in assessment areas, suggests the potential significance of the CRA as a cause of the financial crisis. It is noteworthy that the Commission majority was not willing even to consider the significance of the NCRC’s numbers. In connection with its only hearing on the housing issue, and before any research had been done on the NCRC statements, the Commission published a report absolving CRA of any responsibility for the financial crisis.¹⁵⁴

To understand CRA’s role in the financial crisis, the relevant statistic is the $4.5 trillion in bank CRA lending commitments that the NCRC cited in its 2007 report. (This document and others that are relevant to this discussion were removed from the NCRC website, www.ncrc.org, after they received publicity but can still be found on the web¹⁵⁵). One important question is whether the bank regulators cooperated with community groups by withholding approvals of applications for mergers and acquisitions until an agreement or commitment for CRA lending satisfactory to the community groups had been arranged. It is not difficult to imagine that the regulators did not want the severe criticism from Congress that would have followed their failure to assist community groups in reaching agreements with and getting commitments from banks that had applied for these approvals. In statements in connection with mergers it has approved the Fed has said that commitments by the bank participants about future CRA lending have no influence on the approval process. A Fed official also told the Commission’s staff that the Fed did not consider these commitments in connection with merger applications. The Commission did not attempt to verify this statement, but accepted it at face value from a Fed staff official. Nevertheless, there remains no explanation for why banks have been making these enormous commitments in connection with mergers, but not otherwise.

The largest of the commitments, in terms of dollars, were made by four banks or their predecessors—Bank of America, JPMorgan Chase, Citibank, and Wells Fargo—in connection with mergers or acquisitions as shown in Table 14 below.

Table 14. Announced CRA Commitments in Connection
with a Merger or Acquisition by Four Largest Banks and Their Predecessors

Compiled by Edward Pinto from the NCRC 2007 report CRA Commitments, found at http://www.community-wealth.org/_pdfs/articles-publications/cdfis/report-silver-brown.pdf, NCRC testimony regarding Bank of America’s $1.5 trillion in CRA agreements and commitments in conjunction with its 2008 acquisition of Countrywide found at http://www.house.gov/apps/list/hearing/financialsvcs_dem/taylor_testimony_-_4.15.10.pdf.

Given the enormous size of the commitments reported by NCRC, the key questions are: (i) how many of these commitments were actually fulfilled by the banks that made them, (ii) where are these loans today, and (iii) how are these loans performing?

Currently, in light of the severely limited Commission investigation of this issue, there are only partial answers to these questions.

Were the loans actually made? The banks that made these commitments apparently came under pressure from community groups to fulfill them. In an interview by Brad Bondi of the Commission’s staff, Josh Silver of the NCRC noted that community groups did follow up these commitments.

Bondi: Who follows up…to make sure that these banks honor their voluntary agreements or their unilateral commitments?

Silver: Actually part of some of these CRA agreements was meeting with the bank two or three times a year and actually going through, ‘Here’s what you’ve promised. Here’s what you’ve loaned.’ That would happen on a one-on-one basis with the banks and the community organizations.¹⁵⁶

Nevertheless, when the Commission staff asked the four largest banks (Bank of America, Citibank, JPMorgan Chase and Wells Fargo) for data on whether the merger-related commitments were fulfilled and in what amount, most of the banks supplied only limited information. They contended that they did not have the information or that it was too difficult to get, and the information they supplied was sketchy at best.

In some cases, the information supplied to the Commission by the banks, in letters from their counsel, reflected fewer loans than they had claimed in press releases to have made in fulfillment of their commitments. The press release amounts were JPMorgan Chase (including WaMu, $835 billion), Citi ($274 billion), and Bank of America ($229 billion), totaled $1.3 trillion in CRA loans between 2001 and 2008, and had been presented to the Commission by Edward Pinto in the Triggers memo.¹⁵⁷ No Wells Fargo press releases could be found, but in response to questions from the Commission Wells provided a great deal of data in spread sheets that could not be interpreted or understood without further discussion with representatives of the bank. However, the Commission terminated the investigation of the merger-related CRA commitments in August 2010, before the necessary data could be gathered. For this reason, the Wells data could not be unpacked, interpreted in discussion with Wells officials, and analyzed.

After I protested the limited efforts of the Commission on this issue in October 2010, the Commission made a belated attempt to restart the investigation of the merger-related CRA commitments in November. However, only one bank had responded by the deadline for submission of this dissenting statement. As with the bank responses, additional work was required to understand the information received, and there was no time, and no Commission staff, to follow up.

As a result of the dilatory nature of the Commission’s investigation, it was impossible to determine how many loans were actually made under their merger-related CRA commitments by the four banks and their predecessors. This in turn impeded any effort to find out where these loans are today and hence their delinquency rates. It appears that in many instances the Commission management constrained the staff in their investigation into CRA by limiting the number of document requests and interviews and by preventing the staff from following up with the institutions that failed to respond adequately to requests for data.

Where are these mortgages today? Where these loans are today must necessarily be a matter of speculation. Some of the banks told the FCIC staff that they do not distinguish between CRA loans and other loans, and so could not provide this information. Under the GSE Act, Fannie and Freddie had an affirmative obligation to help banks to meet their CRA obligations, and they undoubtedly served as a buyer for the loans made by the largest banks and their predecessors pursuant to the commitments. In a press release in 2003, for example, Fannie reported that it had acquired $394 billion in CRA loans, about $201 billion of which occurred in 2002.¹⁵⁸ This amounted to approximately 50 percent of Fannie’s AH acquisitions for that year.

In the Triggers memo, based on his research, Pinto estimated that Fannie and Freddie purchased about 50 percent of all CRA loans over the period from 2001 to 2007 and that, of the balance, about 10-15 percent were insured by FHA, 10-15 percent were sold to Wall Street, and the rest remain on the books of the banks that originated the loans.¹⁵⁹ Many of these loans are likely unsaleable in the secondary market because they were made at rates that did not compensate for risk or lacked mortgage insurance—again, the competition for these loans among the GSEs, FHA and the banks operating under CRA requirements inevitably raised their prices and thus underpriced their risk. To sell these loans, the banks holding them would have to take losses, which many are unwilling to do.

What are the delinquency rates? Under the Home Mortgage Disclosure Act HMDA), banks are required to provide data to the Fed from which the delinquency rates on loans that have high interest rates can be calculated. It was assumed that these were the loans that might bear watching as potentially predatory. When Fannie and Freddie, FHA, Countrywide and other subprime lenders and banks under CRA are all seeking the same loans—roughly speaking, loans to borrowers at or below the AMI—it is likely that these loans when actually made will bear concessionary interest rates so that their rate spread is not be reportable under HMDA. It’s just supply and demand. Accordingly, the banks that made CRA loans pursuant to their commitments have no obligation to record and report their delinquency rates, and as noted above several of the large banks that made major commitments recorded by the NCRC told FCIC staff that they don’t keep records about the performance of CRA loans apart from other mortgages.

However, in the past few years, Bank of America has been reporting the performance of CRA loans in its annual report to the SEC on form 10-K. For example, the bank’s 10-K for 2009 contained the following statement: “At December 31, 2009, our CRA portfolio comprised six percent of the total residential mortgage balances, but comprised 17 percent of nonperforming residential mortgage loans. This portfolio also comprised 20 percent of residential net charge-offs during 2009. While approximately 32 percent of our residential mortgage portfolio carries risk mitigation protection, only a small portion of our CRA portfolio is covered by this protection.”¹⁶⁰ This could be an approximation for the delinquency rate on the merger-related CRA loans that the four banks made in fulfilling their commitments, but without definitive information on the number of loans made and the banks’ current holdings it is impossible to make this estimate with any confidence. In a letter from its counsel, another bank reported serious delinquency rates on the loans made pursuant to its merger-related commitments ranging from 5 percent to 50 percent, with the largest sample showing a 25 percent delinquency rate.

Further investigation of this issue is necessary, including on the role of the bank regulators, in order to determine what effect, if any, the merger-related commitments to make CRA loans might have had on the number of NTMs in the U.S. financial system before the financial crisis.

⁵⁹ Speech at Morehead College April 14, 2009.

⁶⁰ Raghuram G. Rajan, Fault Lines, p.44.

⁶¹ Report to Congress on the Root Causes of the Foreclosure Crisis , January 2010, p.xii, http://www.huduser.org/portal/publications/hsgfin/foreclosure_09.html.

⁶² Issue Brief: HUD’s Affordable Housing Goals for Fannie Mae and Freddie Mac, p.5.

⁶³ Final Rule, http://fdsys.gpo.gov/fdsys/pkg/FR-2004-11-02/pdf/04-24101.pdf.

⁶⁴ HUD PDR, May 2005, HUD Contract C-OPC-21895, Task Order CHI-T0007, “Recent House Price Trends and Homeownership Affordability”, p.85.

⁶⁵ Steve Cocheo, “Fair-lending pressure builds,” ABA Banking Journal, vol. 86, 1994, http://www.questia.com/googleScholar.qst?docId=5001707340.

⁶⁶ See NCRC, CRA Commitments, 2007.

⁶⁷ Federal Register,vol. 69, No. 211, November 2, 2004, Rules and Regulations, p.63585, http://fdsys.gpo.gov/fdsys/pkg/FR-2004-11-02/pdf/04-24101.pdf.

⁶⁸ Larry Kudlow, “Barney Frank Comes Home to the Facts,” GOPUSA, August 23, 2010, available at www.gopusa.com/commentary/2010/08/kudlow-barney-frank-comes-home-to-the-facts.php#ixzz0zdCrWpCY (accessed September 20, 2010).

⁶⁹ Document in author’s files.

⁷⁰ Fannie Mae Foundation, “Making New Markets: Case Study of Countrywide Home Loans,” 2000, http://content.knowledgeplex.org/kp2/programs/pdf/rep_newmortmkts_countrywide.pdf.

⁷¹ GSE Act, Section 1332.

⁷² Id., Section 1333.

⁷³ Id., Section 1335.

⁷⁴ Id., Section 1354(a).

⁷⁵ Fannie Mae, “Opening Doors with Fannie Mae’s Community Lending Products,” 1995, p.3.

⁷⁶ Fannie Mae, Memo from Credit Policy Staff to Credit Policy Committee, “CHBP Performance,” November 14, 1995, p.1.

⁷⁷ HUD, “The National Homeownership Strategy: Partners in the American Dream,” available at http://web.archive.org/web/20010106203500/www.huduser.org/publications/affhsg/homeown/chap1.html.

⁷⁸ Id., Chapter 4, Action 35.

⁷⁹ The term “flexible” has a special meaning when HUD uses it. See note 8 supra.

⁸⁰ HUD, Urban Policy Brief No.2, August 1995, available at http://www.huduser.org/publications/txt/hdbrf2.txt.

⁸¹ “Deconstructing the Subprime Debacle Using New Indices of Underwriting Quality and Economic Conditions: A First Look,” by Anderson, Capozza, and Van Order, found at http://www.ufanet.com/DeconstructingSubprimeJuly2008.pdf.

⁸² HUD’s “National Homeownership Strategy – Partners in the American Dream,” http://web.archive.org/web/20010106203500/www.huduser.org/publications/affhsg/homeown/chap1.html.

⁸³ ”Fannie Mae’s Role in Affordable Housing Finance: Connecting World Capital Markets and America’s Homebuyers,” Presentation to HUD Assistant Secretary Albert Trevino, January 10, 2003.

⁸⁴ http://www.washingtonpost.com/wp-dyn/content/article/2008/06/09/AR2008060902626.html.

⁸⁵ See Footnote 32.

⁸⁶ Fannie Mae, internal memo, Adolfo Marzol to Dan Mudd, “RE: Private Label Securities,” March 2, 2005.

⁸⁷ William J. Clinton, Remarks on the National Homeownership Strategy, June 5, 1995.

⁸⁸ HUD Press Release, HUD No. 99-131, July 29, 1999.

⁸⁹ http://frwebgate.access.gpo.gov/cgi-bin/getdoc.cgi?dbname=2000_register&docid=page+65043-65092.

⁹⁰ HUD, Office of Policy Development and Research, Issue Brief No. 5, January 2001, p.3.

⁹¹ Jamie S. Gorelick, Remarks at American Bankers Association conference, October 30, 2000. http://web.archive.org/web/20011120061407/www.fanniemae.com/news/speeches/speech_152.html.

⁹² Daniel H. Mudd’s Responses to the Questions Presented in the FCIC’s June 3, 2010, letter, Answer to Question 6: How influential were HUD’s affordable housing guidelines in Fannie Mae’s purchase of subprime and Alt-A loans? Were Alt-A loans “goals-rich”? Were Alt-A loans net positive for housing goals?

⁹³ Fannie Mae, “Discussion of HUD’s Proposed Housing Goals,” Presentation to the Department of Housing and Urban development, June 9, 2004.

⁹⁴ Josh Rosner, “Housing in the New Millennium: A Home Without Equity is Just a Rental With Debt,” June, 2001, p.7, available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1162456.

⁹⁵ Id., p.29.

⁹⁶ http://fdsys.gpo.gov/fdsys/pkg/FR-2004-11-02/pdf/04-24101.pdf, p.63601.

⁹⁷ Fannie Mae, “The HUD Housing Goals”, March 2003.

⁹⁸ HUD, Office of Policy Development and Research, Profiles of GSE Mortgage Purchases, 1992-2000, 2001-2004, and 2005-2007.

⁹⁹ Neil Morse, “Looking for New Customers,” Mortgage Banking, December 1, 2004. It may be significant that the chairman of Freddie Mac at the time, Leland Brendsel, did not attend the 2000 press conference or pledge support for HUD’s new goals. Raines must have forgotten his 1999 pledge to Secretary Cuomo and his speech to the mortgage bankers when he wrote in a letter to The Wall Street Journal on August 3, 2010: “The facts about the financial collapse of Fannie and Freddie are pretty clear and a matter of public record. The company managers, their regulator and the Treasury have all said that the losses which crippled the companies were caused by the purchase of loans with lower credit standards between 2005 and 2007. The companies explicitly changed their credit standards in order to regain market share after Wall Street began to define market credit standards in 2004.”

¹⁰⁰ Fannie Mae, disk produced for FCIC, April 7, 2010. Throughout this analysis, I have not discussed the GSEs’ compliance with the “Underserved Base Goal,” which is included in this table. The Underserved Base Goal applied mostly to minorities and involved a different set of lending decisions than the LMI goal and the Special Affordable Goal.

¹⁰¹ Fannie Mae, “GSE Credit Losses,” presentation to House Financial Services Committee, April 16, 2010.

¹⁰² Rob Blackwell, “Two GSEs Cut Corners to Hit Goals, Report Says,” American Banker, May 13, 2005, p.1.

¹⁰³ See, e.g., Barry Ritholtz, “Get Me ReWrite!” in Bailout Nation, Bailouts, Credit, Real Estate, Really, Really Bad Calls, May 13, 2010, http://www.ritholtz.com/blog/2010/05/rewriting-the-causes-of-the-credit-crisis/print/; Dean Baker, “NPR Tells Us that Republicans Believe that Fannie and Freddie Caused the Crash” Beat the Press Blog, Center for Economic and Policy Research http://www.cepr.net/index.php/blogs/beat-the-press/npr-tells-us-that-republicans-believe-that-fannie-and-freddie-caused-the-crash; Charles Duhigg, “Roots of the Crisis,” Frontline, Feb 17, 2009, http://www.pbs.org/wgbh/pages/frontline/meltdown/themes/howwegothere.html.

¹⁰⁴ Pinto, “Government Housing Policies in the Lead-up to the Financial Crisis: A Forensic Study,” Chart 52, p.148, http://www.aei.org/docLib/Government-Housing-Policies-Financial-Crisis-Pinto-102110.pdf.

¹⁰⁵ Id.

¹⁰⁶ Inside Mortgage Finance, The 2009 Mortgage Market Statistical Annual—Volume II, pp. 139 and 140.

¹⁰⁷ Inside Mortgage Finance, The 2009 Market Statistical Annual—Volume II, p.143.

¹⁰⁸ Tom Lund, “Single Family Guarantee Business: Facing Strategic Crossroads” June 27, 2005.

¹⁰⁹ Stephen B. Ashley Fannie Mae Chairman, remarks at senior management meeting, June 27, 2006.

¹¹⁰ OFHEO, “Mortgage Markets and the Enterprises in 2007,” pp. 33-34.

¹¹¹ Fannie Mae, 2004 10-K. These totals do not include Fannie’s purchases of subprime PMBS. http://www.fanniemae.com/ir/pdf/sec/2004/2004_form10K.pdf;jsessionid=N3RRJCZPD5SOVJ2FQSHSFGI, p.141 and Fannie’s 2007 10-K, http://www.fanniemae.com/ir/pdf/sec/2008/form10k_022708.pdf;jsessionid=N3RRJCZPD5SOVJ2FQSHSFGI, p.127.

¹¹² Fannie Mae, 2005 10-K, p.37.

¹¹³ Fannie Mae, Minutes of a Meeting of the Board of Directors, October 16, 2007, p.18.

¹¹⁴ FHFA Mortgage Market Note 10-2, http://www.fhfa.gov/webfiles/15408/Housing%20Goals%201996-2009%2002-01.pdf.pdf.

¹¹⁵ http://frwebgate.access.gpo.gov/cgi-bin/getdoc.cgi?dbname=2000_register&docid=page+65093-65142.

¹¹⁶ Bell, Kinney, Kunde and Weech, through Zigas and Marks internal memo Frank Raines, “RE: HUD Housing Goals Options,” June 15, 1999.

¹¹⁷ Barry Zigas, “Fannie Mae and Minority Lending: Assessment and Action Plan” Presentation, November 16, 2000.

¹¹⁸ Fannie Mae, “Action Plan for the Housing Goals Rewrite,” January 22, 2003.

¹¹⁹ Fannie Mae, “The HUD Housing Goals,” March 2003.

¹²⁰ Fannie Mae internal memo, Paul Weech to Brian Graham, “RE: Mission Legislation,” September 3, 2004.

¹²¹ Fannie Mae, “Costs and Benefits of Mission Activities, Project Phineas,” June 14, 2005.

¹²² Barry Zigas, “Housing Goals and Minority Lending,” September 30, 2005.

¹²³ Fannie Mae, “Update on Fannie Mae’s Housing Goals Performance,” Presentation to the U.S. Department of Housing and Development, October 31, 2005.

¹²⁴ Freddie Mac, internal email, Donna Cogswell on behalf of David Andrukonis to Dick Syron, “RE: No Income/No Asset (NINA) Mortgages,” September 7, 2004.

¹²⁵ Freddie Mac, internal email from Mike May to Dick Syron, “FW: FINAL NINA Memo,” October 6, 2004.

¹²⁶ Fannie Mae, internal memo, Single Family Business Product Management and Development to Single Family Business Credit Committee, “RE: PMD Proposal for Increasing Housing Goal Loans,” May 5, 2006, p.6.

¹²⁷ Id., p.8.

¹²⁸ Fannie Mae, 2006 10-K, p.146.

¹²⁹ Fannie Mae, “Business Update,” presentation. “Cash flow cost” equals expected revenue minus expected loss. Expected revenue is what will be received in G-fees; expected loss includes G&A and credit losses. “Opportunity cost” is the G-fee actually charged minus the model fee—the fee that Fannie’s model would impose to guarantee a mortgage of the same quality in order to earn a fair market return on capital.

¹³⁰ Fannie Mae, “Housing Goals Briefing for HUD,” April 11, 2007.

¹³¹ Fannie Mae, “Housing Goals Forecast,” Alignment Meeting, June 22, 2007.

¹³² Fannie Mae, Forecast Meeting, July 27, 2007 slide 4.

¹³³ Fannie Mae letter, Daniel Mudd to Asst. Secretary Brian Montgomery, December 21, 2007, p.6.

¹³⁴ FHFA, Fannie Mae and Freddie Mac Single Family Guarantee Fees in 2007 and 2008, p.33.

¹³⁵ http://www.fanniemae.com/ir/pdf/sec/2009/q2credit_summary.pdf, p.5.

¹³⁶ Freddie Mac, “Cost of Freddie Mac’s Affordable Housing Mission,” Business Risk Committee, Board of Directors, June 4, 2009.

¹³⁷ Federal Housing Finance Agency, 2010-2011 Enterprise Housing Goals; Enterprise Book-Entry Procedures; Final Rule, 12 CFR Parts 1249 and 1282, Federal Register, September 14, 2010, p.55892.

¹³⁸ Title V of the Congressional Budget Act of 1990. Under the FCRA, HUD must estimate the annual cost of FHA’s credit subsidy for budget purposes. The credit subsidy is the net of its estimated receipts reduced by its estimated payments.

¹³⁹ Kerry D. Vandell, “FHA Restructuring Proposals: Alternatives and Implications,” Fannie Mae Housing Policy Debate, vol. 6, Issue 2, 1995, pp. 308-309

¹⁴⁰ GAO, “Federal Housing Administration: Decline in Agency’s Market Share Was Associated with Product and Process Developments of Other Mortgage Market Participants,” GAO-07-645, June 2007, pp. 42 and 44.

¹⁴¹ Integrated Financial Engineering, “Actuarial Review of the Federal Housing Administration Mutual Mortgage Insurance Fund (Excluding HECMs) for Fiscal Year 2009,” prepared for U.S. Department of Housing and Urban Development, November 6, 2009, p.42.

¹⁴² Quicken press release, “Quicken Loans First To Offer FHA Home Mortgages Nationally On The Internet With HUD´s approval, Intuit expands home ownership nationwide, offering consumers widest variety of home loan options”, January 20, 2000, http://web.intuit.com/about_intuit/press_releases/2000/01-20.html.

¹⁴³ GAO, “Federal Housing Administration: Decline in Agency’s Market Share Was Associated with Product and Process Developments of Other Mortgage Market Participants,” GAO-07-645, June 2007, pp. 42 and 44.

¹⁴⁴ HUD’s Best Practices Initiative was described this way by HUD: “Since 1994, HUD has signed Fair Lending Best Practices (FLBP) Agreements with lenders across the nation that are individually tailored to public-private partnerships that are considered on the leading edge. The Agreements not only offer an opportunity to increase low-income and minority lending but they incorporate fair housing and equal opportunity principles into mortgage lending standards. These banks and mortgage lenders, as represented by Countrywide Home Loans, Inc., serve as industry leaders in their communities by demonstrating a commitment to affirmatively further fair lending.” Available at: http://www.hud.gov/local/hi/working/nlwfal2001.cfm.

¹⁴⁵ Steve Cocheo, “Fair-Lending Pressure Builds”, ABA Banking Journal, vol. 86, 1994, http://www.questia.com/googleScholar.qst?docId=5001707340.

¹⁴⁶ HUD, “Building Communities and New Markets for the 21st Century,” FY 1998 Report , p.75, http://www.huduser.org/publications/polleg/98con/NewMarkets.pdf.

¹⁴⁷ Fannie Mae Foundation, “Making New Markets: Case Study of Countrywide Home Loans,” 2000, http://content.knowledgeplex.org/kp2/programs/pdf/rep_newmortmkts_countrywide.pdf.

¹⁴⁸ “Questions and Answers from Countrywide about Lending,” December 11, 2007, available at http://www.realtown.com/articles/article/print/id/768.

¹⁴⁹ 12 U.S.C. 2901.

¹⁵⁰ Original letter in author’s files.

¹⁵¹ Randall Kroszner, Speech at the Confronting Concentrated Poverty Forum, December 3, 2008.

¹⁵² Ben S. Bernanke, “The Community Reinvestment Act: Its Evolution and New Challenges,” March 30, 2007, p2.

¹⁵³ See Note 12 supra.

¹⁵⁴ FCIC, “The Community Reinvestment Act and the Mortgage Crisis.” Preliminary Staff Report, http://www.fcic.gov/reports/pdfs/2010-0407-Preliminary_Staff_Report_-_CRA_and_the_Mortgage_Crisis.pdf.

¹⁵⁵ http://www.community-wealth.org/_pdfs/articles-publications/cdfis/report-silver-brown.pdf.

¹⁵⁶ Interview of Josh Silver of the National Community Reinvestment Coalition, June 16, 2010.

¹⁵⁷ Edward Pinto, Exhibit 2 to the Triggers memo, dated April 21, 2010, http://www.aei.org/docLib/Pinto-Sizing-Total-Federal-Contributions.pdf.

¹⁵⁸ “Fannie Mae Passes Halfway Point in $2 Trillion American Dream Commitment; Leads Market in Bringing Housing Boom to Underserved Families, Communities” http://findarticles.com/p/articles/mi_m0EIN/is_2003_March_18/ai_98885990/pg_3/?tag=content;col1.

¹⁵⁹ Triggers memo, p.47.

¹⁶⁰ Bank of America, 2009 10-K, p.57.

IV. CONCLUSION

This dissenting statement argues that the U.S. government’s housing policies were the major contributor to the financial crisis of 2008. These policies fostered the development of a massive housing bubble between 1997 and 2007 and the creation of 27 million subprime and Alt-A loans, many of which were ready to default as soon as the housing bubble began to deflate. The losses associated with these weak and high risk loans caused either the real or apparent weakness of the major financial institutions around the world that held these mortgages—or PMBS backed by these mortgages—as investments or as sources of liquidity. Deregulation, lack of regulation, predatory lending or the other factors that were cited in the report of the FCIC’s majority were not determinative factors.

The policy implications of this conclusion are significant. If the crisis could have been prevented simply by eliminating or changing the government policies and programs that were primarily responsible for the financial crisis, then there was no need for the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010, adopted by Congress in July 2010 and often cited as one of the important achievements of the Obama administration and the 111th Congress.

The stringent regulation that the Dodd-Frank Act imposes on the U.S. economy will almost certainly have a major adverse effect on economic growth and job creation in the United States during the balance of this decade. If this was the price that had to be paid for preventing another financial crisis then perhaps it’s one that will have to be borne. But if it was not necessary to prevent another crisis—and it would not have been necessary if the crisis was caused by actions of the government itself—then the Dodd-Frank Act seriously overreached.

Finally, if the principal cause of the financial crisis was ultimately the government’s involvement in the housing finance system, housing finance policy in the future should be adjusted accordingly.

APPENDIX 1

Hypothetical Losses in Two Scenarios (No feedback)

Scenario 1 is what was known to market professional during the 2nd half of 2007; Scenario 2 is the actual condition of the mortgage market. Second mortgage/home equity loan losses are excluded.

Assumptions used:

Number of mortgages= 53 million;

Total value of first mortgages=$9.155 trillion;

Losses on Prime=1.2%% (assumes 3% foreclosure rate & 40% severity);

Losses on Subprime/Alt-A=12% (assumes 30% foreclosure rate & 40% severity);

Average size of mortgage: $173,000

Losses in Scenario 1

Number of mortgages: 53 million

Prime=40 million

Subprime/Alt-A = 13 million (7.7. PMBS million + FHA/VA=5.2 million)

Aggregate Value:

Prime =$6.9 trillion ($173,000 X 40 million);

Subprime/Alt-A=$2.25 trillion ($173,000 X 13 million)

Losses on foreclosures: $353 billion ($6.9 trillion prime X 1.2%=$83 billion + $2.25 trillion subprime/Alt-A X 12%=$270 billion

Overall loss percentage: 3.5%

Losses in Scenario 2

Number of mortgages: 53 million

Prime: 27 million

Subprime/Alt-A:

Original subprime/Alt-A: 13 million

Other subprime/Alt-A: 13 million (10.5 F&F (excludes 1.25 million already counted in PMBS) + 2.5 million other loans not securitized (mostly held by the large banks))

Aggregate Value:

Prime= $4.7 trillion ($173,000 X 27 million);

Subprime/Alt-A = $4.5 trillion ($173,000 X 26 million)

Losses on foreclosures: $596 billion ($4.7 trillion X 1.2%=$56 billion + $4.5 trillion X 12%=$540 billion)

Overall loss percentage: 6.5%, for an increase of 86%

Note: No allowance for feedback effect—that is, fall in home prices as a result of larger number of foreclosures in Scenario 2. With feedback effect, losses would be even larger in Scenario 2 because a larger number of foreclosures would drive down housing prices further and faster. This feedback effect will likely cause total first mortgage losses to approach $1 trillion or 10% of outstanding first mortgages.

APPENDIX 2

Hypothetical Losses in Two Scenarios (with feedback)

Scenario 1 is what was known to market professional during the 2nd half of 2007; Scenario 2 is the actual condition of the mortgage market. Second mortgage/home equity loan losses are excluded.

Assumptions used:

Number of mortgages= 53 million;

Total value of first mortgages=$9.155 trillion;

Scenario 1:

Losses on prime=1.2%% (assumes 3% foreclosure rate & 40% severity);

Losses on self-denominated subprime & Alt-A=14% ((assumes 35% foreclosure rate & 40% severity);

Losses on FHA/VA=5.25% (assumes 15% foreclosure rate and 35% severity)

Scenario 2:

Losses on prime=1.6%% (assumes 3.5% foreclosure rate and 45% severity);

Losses on self-denominated subprime & Alt-A=25% (assumes 45% foreclosure rate & 55% severity);

Losses on FHA/VA & unknown subprime/Alt-A=15% (assumes 30% foreclosure rate & 50% severity)

Average size of mortgage:

Prime: $173,000 ($6.75 trillion/39 million)

Subprime/Alt-A/FHA/VA: $182,000 ($2.4 trillion/13 million

Losses in Scenario 1

Number of mortgages: 53 million

Prime=40 million

Subprime/Alt-A=7.7 million PMBS

FHA, and VA=5.2 million

Aggregate Value:

Prime =$6.9 trillion ($173,000 X 39 million);

Subprime/Alt-A=$1.7 trillion ($220,000 X 7.7 million)

FHA/VA= $700 billion ($130,000x5.2 million)

Total expected foreclosures: 4.7 million (3% X 39 million + 35% X 7.7 million + 15% X 5.2 million)

Losses on foreclosures: $360 billion ($6.9 trillion prime X 1.2%=$83 billion + 1.7 trillion subprime/Alt-A X 14%=$240 billion + $700 billion X 5.25%=37 billion)

Overall loss percentage: 3.9%

Losses in Scenario 2

Number of mortgages: 53 million

Prime: 27 million

Original subprime/Alt-A: 7.7 million

FHA/VA: 5.2 million

Other subprime/Alt-A: 13 million (10.5 F&F (excludes 1.25 million already counted in PMBS), 2.5 million other loans not securitized (mostly held by the large banks))

Aggregate Value:

Prime= $4.7 trillion ($173,000 X 27 million);

Original Subprime/Alt-A = $1.7 trillion ($220,000 X 7.7 million)

FHA/VA= $700 billion ($130,000x5.2 million)

Other subprime/Alt-A: $2 trillion ($154,000X13 million

Total expected foreclosures: 8.4 million (3.5% X 27 million=0.95 million, 45% X 7.7 million=3.5 million, 30% X 13 million=3.9 million)

Losses on foreclosures: $890 billion ($4.7 trillion X 1.6%=$60 billion + $1.7 trillion X 25%=$425 billion + $700 billion X 15% = $105 billion + $2 trillion X 15% = $300 billion)

Overall loss percentage: 9.8%, for an increase of 150%