The Holes In the Administration's Housing Reform

Forgot Password?

Getty Images

A foreclosed home stands boarded up on February 9, 2012 in Islip, New York. A New York State Department of Financial Services Foreclosure Relief Unit van visited Islip to provide individuals who are facing foreclosure with counselors who can assess where homeowners are in the pre-foreclosure or foreclosure process. Islip, which is located in Suffolk County, has the highest foreclosure rate in New York State.

Tweet This

Tweet This

Tweet This

Subscribe to AEI's Financial Services e-mails.

Empty promise: the holes in the administration's housing finance reform plan Download PDF In the year since US Treasury Secretary Timothy Geithner announced the Obama administration’s options for reforming the housing market, the administration has said and done nothing to indicate which option it prefers or how its plan will be implemented. In early February, however, Geithner reported to the Financial Stability Oversight Council (FSOC) that the administration wants to make progress on housing finance reform this year, winding down government-sponsored enterprises (GSEs) Fannie Mae and Freddie Mac and bringing private capital back into the market. For any such plan to be credible, it must do much more than wind down the GSEs. The Federal Housing Administration (FHA) creates competitive obstacles to the revival of the private securitization market that are at least as serious as the GSEs, and because of the Dodd-Frank Act a number of formidable legal obstacles now exist that must be cleared away before a private securitization market will come back. If the administration is serious, its plan must address all these issues.

Key points in this Outlook:

In a report to the Financial Stability Oversight Council in early February, Geithner declared, “We also want to make progress this year in building the foundation for reforms to the mortgage market in the United States, including a path for winding down the GSEs. . . . Our plan will wind down the GSEs and bring private capital back into the market, reducing the government’s direct role in the housing market and better targeting our support toward first-time homebuyers and low-and-moderate income Americans.”[1] Unfortunately, for his statement to be credible, Geithner has to recognize that winding down Fannie and Freddie will not in itself materially assist the revival of the private securitization market. Other steps are necessary, including reducing the role of the FHA and substantially changing the housing provisions incorporated in the Dodd-Frank Act.

Geithner’s statement came almost one year to the day after he had announced a series of options for doing exactly what he just outlined—winding down the GSEs and reducing the government’s role in the housing market. That original plan, announced on February 11, 2011, contained three options, (1) eliminating the government’s role entirely, except for low- and moderate-income housing; (2) keeping the government as a backstop in case of another financial collapse; and (3) replacing the GSEs with a program in which a government agency would guarantee mortgage-backed securities issued by privately capitalized firms. It also proposed to limit the role of the FHA to providing loans for low-income housing. In the year that has passed, the administration has steadfastly refused to indicate which of  these options it actually favored, and Geithner’s most recent statement—except for the reference to reducing the government’s “direct” role—does not provide any further guidance.

Shortly after Geithner’s statement last year, Alex Pollock, Ed Pinto, and I released a white paper (the WPP white paper) that followed up on the administration’s apparent willingness to consider a wholly private system of housing finance.[2] The plan detailed how a private market would work and how to eliminate Fannie Mae and Freddie Mac and reduce the role of the FHA while stimulating the revival of a private housing finance system. Political agreement between the House Republicans and the administration on a private housing finance system is a real possibility, but the administration must first express an interest in such a conversation, which it has not done.

The lack of any progress on housing finance reform in the past year was emphasized in late February when the Federal Housing Finance Agency (FHFA), the GSEs’ regulator, issued what it called a “strategic plan” for their future.[3] Given its subordinate position within the administration, FHFA’s plan could not go beyond the terms the administration laid out a year ago, but it noted out that its limited recommendations were consistent with both the administration’s year-old plan and the leading congressional proposals introduced to date. The FHFA cannot do much to move the ball forward until the administration and congressional Republicans come to an agreement about the major features of the housing finance system they want in the future.

Certainly, winding down the GSEs is an essential precursor to the revival of a private securitization market. However, it is not sufficient. The FHA is also a major player, especially for home purchase loans (as distinguished from refinances) and poses the same competitive problem as the GSEs for the revival of a private system. Between 2004 and 2010, the FHA has increased its market share in home purchase loans from 9 percent to 42 percent. Other government agencies—Veterans Administration and the Department of Agriculture (Rural Housing Administration)—had another 12 percent.[4]

"Winding down the GSEs is an essential precursor to the revival of a private securitization market. However, it is not sufficient."In November 2011, under pressure from the National Association of Realtors, Congress increased to $729,750 the size of the mortgages FHA is permitted to insure. This gives the FHA an even wider scope of operations in the housing market than the GSEs, which are limited to acquiring mortgages no larger than $625,500. Accordingly, without FHA reform, including limits on its growth, the agency could take over the market now still dominated by Fannie and Freddie. Under those circumstances, winding down the GSEs will not substantially reduce either the amount of moral hazard in the US housing finance system or the government’s liability for housing losses. That would simply move it from the GSEs to the FHA.

The urgency of reviving the private-sector securitization system was emphasized in the FHFA’s strategic plan, which noted, “No private sector infrastructure exists today that is capable of securitizing the $100 billion per month in new mortgages being originated.”[5] Thus, an administration proposal to get housing finance reform moving is essential, but if the administration’s plan for reviving the private securitization market is to have any credibility it must also provide a mechanism for reducing the FHA’s role, as well as that of the GSEs. Finally, and equally important, it will be necessary to change the regulatory landscape that now—after the Dodd-Frank Act (DFA)—is hostile to private securitization and the revival of a private securitization market.

This Outlook will address each of these issues.

Competition from the GSEs and FHA 

The GSEs and FHA are serious competitors for a private housing finance system because they are subsidized by the taxpayers, who take much of the mortgage credit risk associated with their activities. Taxpayers have already invested more than $180 billion in the GSEs to keep them afloat, and the total loss on these two firms will most likely eventually reach $300 billion. As long as the taxpayers are required to take these risks, the presence of Fannie and Freddie in the housing finance markets will impede the revival of a private securitization market. No private securitizer can hope to survive when its competitors are backed by the government and, ultimately, the taxpayers.

The same is true of the FHA, which insures mortgages that are then securitized through Ginnie Mae, another government agency subsidized by the taxpayers. As described in a recent Outlook, if the FHA were subject to the accounting standards that are applied to private mortgage insurers, it would be deeply insolvent.[6] The only reason it is still able to function is that its insurance obligations are backed by the full faith and credit of the US government. In other words, the taxpayers will eventually be required to recapitalize the FHA just as they have had to recapitalize the GSEs.

One obstacle to winding down the GSEs and reducing the FHA’s role is the mistaken idea that, without a government guarantee of the credit risk, institutional investors will not be willing to buy mortgage-backed securities (MBS). The truth is very nearly the opposite; as long as the government continues to take the credit risk on mortgages, the institutional investors with the largest long-term investment portfolios—insurance companies and private pension plans—will not be interested in acquiring these securities. Institutional investors are compensated for taking credit risk, allowing them to earn the higher returns they need to meet their obligations. Thus, insurance companies and private pension funds buy corporate bonds and other fixed-income securities, which carry no government guarantee and in some cases involve substantial credit risk.

The data on this are very clear. According to the Federal Reserve’s flow of funds reports, in 2006, at the height of the mortgage boom, these private institutional investors had total assets of $12.1 trillion.[7] Of this sum, $2.4 trillion, or 19 percent, was invested in corporate and foreign bonds. These instruments carry no credit support from governments and are attractive to institutional investors precisely because they earn the yields they need by taking credit risk. Financial instruments that do not involve credit risk—because the government is taking it—are not attractive because their yields are too low. Thus, of the $12.1 trillion invested by insurers and private pension plans in 2006, only $768 billion (6.3 percent) was invested in GSEs or other government-backed MBS.[8]

"If Congress was willing to reverse itself on a limit already embodied in legislation, the risks of entering the field were even higher than anyone had anticipated."Similarly, as of the end of the third quarter of 2011, the same private institutional investors held assets totaling $12.3 trillion, of which $670 billion (5.4 percent) were invested in GSEs or other government-backed MBS, while $2.9 trillion (23 percent) were invested in corporate and foreign bonds.9 In other words, the largest institutional investors in the United States have no significant interest in government-backed MBS. Instead, they prefer to bear the credit risk of corporate securities that produce the higher yields they need to meet their obligations over time.

This suggests that private institutional investors such as insurance companies and private pension funds would have a strong interest in a private market in mortgages and MBS, where they, rather than the taxpayers, could take the credit risk associated with mortgages. Indeed, this would be a rare win-win for the United States because it would allow these private investors to diversify their portfolios away from corporate debt, helping them achieve more financial stability while making large sums available to the housing market. And needless to say, it would remove from the taxpayers the burden of mortgage credit risk.

Another obstacle to winding down the GSEs is the idea that a government-backed mortgage is less expensive for homeowners than a private mortgage. However, this is true only because the taxpayers are taking the mortgage credit risk. If the government charged for the credit risk it assumes through the GSEs and FHA/Ginnie Mae system, its mortgages would cost roughly the same as private mortgages.

This also leaves unconsidered the unmeasured costs of the financial crises that occur when the losses that accumulate in the government’s housing finance system finally come to light. This occurred in the late 1980s and early 1990s, when the savings-and-loan industry collapsed, causing a recession and many job losses, plus about $150 billion in taxpayer expenses in absorbing the losses of the savings and loans. It most recently occurred again when the 13.4 million subprime and other weak mortgages held or guaranteed by the GSEs defaulted in unprecedented numbers, driving down housing prices and causing both a worldwide financial crisis and a painful recession that continues in the United States. The cost to taxpayers thus far for bailing out the GSEs is $180 billion, with estimates of additional losses up to $300 billion. High GSE and FHA Conforming Loan Limits. Winding down the GSEs and FHA is particularly important because their current conforming loan limits were set when housing prices were considerably higher. Over the last several years, these prices have fallen almost everywhere in the United States, with declines of 30 to 40 percent in some areas. As a result, the prior conforming loan limits, which still apply to Fannie and Freddie, now cover a much larger portion of the housing market than they did before the financial crisis. High conforming loan limits reduce the number of mortgages available for private securitization and thus restrict the scope of any developing securitization business. Reducing loan limits can be done in stages, perhaps over five years, but it must be outlined in a law so that private securitizers will have confidence that if they make the necessary investments to start a securitization business, the available private market will be large enough to make that investment profitable.[10]

In November 2011, Congress dealt a serious blow to the idea that the scope of government’s housing finance activities could be reduced. When it was enacted in 2008, the Housing and Economic Recovery Act (HERA) provided that the conforming loan limits applicable to the GSEs and the FHA would be reduced from $729,750 to $625,500 on October 1, 2011. This step reflected a recognition that Congress had gone too far in bringing high-cost mortgages into the government’s orbit. At the higher level, mortgages on homes costing almost $1 million would have been eligible for government backing by both the GSEs and FHA.

HERA’s reduction of the conforming loan limit was a step in the right direction. If it had been left in place, the private market would have responded by moving into the areas from which the government had withdrawn, probably through the purchase of whole mortgages. Indeed, as the date approached, there were strong signs that the private sector was willing to increase its originations for mortgages above the new $625,500 limit. However, in November Congress returned the loan limit for the FHA to its previous level of $729,750, while leaving the GSEs’ limit at $625,500. This reversal sent a very ominous signal to the private sector; if Congress was willing to reverse itself on a limit already embodied in legislation, the risks of entering the field were even higher than anyone had anticipated. This action seriously set back hopes for a private market revival, a short-term victory for the Realtors that will have a long-term adverse effect on home sales. To its credit, in this instance the administration had opposed the loan limit increase but was unable to persuade its Democratic allies in the Senate.

Even if the GSEs and FHA are substantially wound down, several provisions of the DFA—which the Obama administration designed and steadfastly defends—make a revival of the private securitization market highly unlikely, and these must be cleared away before a private securitization market can develop. Ironically, although many have remarked that the DFA does nothing to reform the GSEs, very few have noticed that it contains provisions that also effectively prevent the return of a private-sector securitization system. Accordingly, the administration’s words about housing finance reform—insofar as they promise a revival of the private securitization market—are simply not credible unless it is willing to open the DFA for amendments in the areas I will discuss.

Dodd-Frank’s Impediments to a Robust Private Securitization System

The DFA dramatically increases the competitive advantage the GSEs and FHA have over private securitization. The act substantially reconfigures the securitization process by creating the concept of a high-quality mortgage—called a qualified residential mortgage (QRM)—and requires the sponsor of the securitization to retain 5 percent of the risk if the mortgages in a securitization pool do not qualify as QRMs. This, in itself, will significantly impede private securitization, as I will describe, but the act contains additional provisions that will also impair the securitization process or discourage firms from entering the business.

Exemptions for FHA and the GSEs from Risk Retention. The act exempts the FHA from the risk-retention requirement, further widening the gap between privately securitized mortgages and those insured by the FHA. By creating a path for the securitization of mortgages without a 5 percent risk-retention requirement, the act also encourages the growth of a market in low-quality mortgages.

Although the DFA did not exempt the GSEs from the risk-retention requirement, the initial draft of the regulations under the act extended the exemption to Fannie and Freddie. This will have the same effect as the act’s exemption of the FHA, widening the gap between GSE and private-sector prices and encouraging the securitization of lower-quality mortgages through the GSEs as well as the FHA. A private securitization market will not develop as long as these provisions remain in effect and the FHA and GSEs are able to insure or securitize any substantial portion of the mortgage market.

Risk Retention and Capital. The DFA provisions that mandate a 5 percent risk retention for mortgage securitizers were originally intended to penalize those who securitize low-quality mortgages. That idea has many flaws that go well beyond exempting the FHA and possibly the GSEs from the risk-retention requirement. Although minimum standards for mortgage quality make sense and some standards of this kind are included in the WPP white paper,[11] the risk-retention idea will make it difficult for any firm other than a large bank to carry on a private securitization business.

"Only large banks will have balance sheets and capital sufficient to carry the 5 percent retention amount for the maturity period of the loans, which could be thirty years."If a QRM is ultimately defined as a high-quality or bulletproof mortgage, most mortgages will not qualify, requiring securitizers to retain 5 percent of the risk. Under these circumstances, only large banks will have balance sheets and capital sufficient to carry the 5 percent retention amount for the maturity period of the loans, which could be thirty years. This will seriously impair competition and—if the largest banks are indeed too big to fail—again put the government’s credit behind the mortgage market. A minimum standard for securitized mortgages should be retained, but the 5 percent retention requirement should be repealed.

Risk Retention and True Sale. Because it favors the largest banks, the 5 percent risk-retention requirement will impair competition. Yet, because of the “true sale” accounting requirements, it will not be effective in preventing the securitization of low-quality mortgages. Under true sale accounting, securitized assets are no longer considered to be assets owned by the securitization sponsor, and the sponsor can immediately recognize a profit on the sale. To get this treatment, the securitization must comply with accounting rules requiring that if the sponsor of a securitization retains any interest in a securitization pool, that interest must be proportional to the risks sold.

The financial regulators’ initial proposal offered four options for how a sponsor might retain a required 5 percent interest; however, only a vertical slice through the pool—in which the sponsor takes a 5 percent share of each of the tranches—is clearly likely to clearly qualify for true sale treatment. Barring a huge loss in a pool, then, a 5 percent vertical slice does not represent a significant risk for a sponsor that will earn most of its profits from origination, distribution, servicing, and management of the mortgage pool.

Most of the risk in a privately issued securitization is rated triple-A, with only relatively small portions of the risk in the lower-rated tranches. Thus, if sponsors generally elect the vertical slice as a way of satisfying the DFA risk-retention requirements, this risk is unlikely to have much effect on the quality of the securitized mortgages. Accordingly, as recommended in the WPP white paper, it would be better to maintain mortgage quality in securitizations through a regulation that specifies the requirements for one or more categories of prime loans.[12]

SEC’s Regulation AB. The Securities and Exchange Commission (SEC) has proposed an amendment to its Regulation AB that is intended to implement section 621 of the DFA. Section 621, among other things, prohibits securitization sponsors from engaging in transactions that involve a “conflict of interest” with investors in the securitization. The act requires that if there is a prohibited conflict, the transaction must be rescinded. The provision was apparently intended to cover transactions in which a securitization sponsor sells interests in a pool and then bets against the pool by acquiring credit default swaps to protect against the possibility that the mortgages in the pool will default.

An inherent conflict exists in every transaction between a buyer and a seller—the seller always thinks the asset will decline in value, while the buyer thinks it will rise. The SEC has the unenviable task of defining and prohibiting some kinds of conflicts while permitting others. To do so, the SEC has proposed an amendment to its Regulation AB, which governs securitization transactions. The SEC’s proposed rule is 118 pages long and asks for comments on 109 separate issues.[13]

Given the complexity of the rule, it may not be possible to know in advance whether a particular transaction will transgress the rule. Many firms that would otherwise enter the securitization business may, given the serious consequences of violating the rule, be deterred by the risk of overstepping the rule’s complex elements, or by the cost of maintaining the expertise to advise on each transaction. Remedial legislation should make clear what specific transactions were meant to be covered, or section 621 of the DFA should be repealed.

Premium Cash Recapture. The DFA’s requirements for premium cash recapture will also discourage securitization by substantially reducing the economic returns to securitizers. Some subprime securitizations during the housing boom involved a substantial spread between what the borrower paid and what the investor received. The existence of this spread, called the yield spread premium, has provoked concern that it was used to compensate originators for steering borrowers into riskier mortgages than they might otherwise have qualified for. To prevent this, section 1403 of the DFA authorized the Fed to prevent differences in originators’ compensation that might create an incentive to steer borrowers into an unnecessarily high-risk loan. The Fed’s regulation requires the entire spread on a loan to be retained until all proceeds have been paid to the investors. In this position, it amounts to a sizable risk retention that may limit the number of firms that can take on the business of securitization and jeopardize the availability of true sale accounting treatment.

Although intended to address predatory lending, this provision would prevent securitizers from realizing the economic benefits of the spread even when they are securitizing prime mortgages. If they try to make up this loss with a higher interest rate, they would be even less competitive with the GSEs and FHA. At the very least, this provision should be modified so that it does not apply to loans not properly categorized as predatory.

Volcker Rule. Although justified as preventing the use of insured bank deposits for risky trading, this rule, as enacted in the DFA, prohibits “bank-related entities” from engaging in proprietary trading and thus extends far beyond the insured banks it was intended to cover. The term “bank-related entities” includes bank holding companies and their subsidiaries, which do not have access to insured deposits. In addition, “proprietary trading” is so difficult to define that the most recent draft regulation covers almost two hundred pages and poses over one thousand separate questions to assist the regulators in drafting the final rule.

There are many concerns about the scope and effect of the Volcker Rule. One major adverse effect is its coverage of hedging transactions. Hedging is a regular and important element of every securitization because it is necessary to protect the issuer against a change in interest rates between the time a mortgage rate is locked in with the borrower and the time a complete pool can be assembled for a securitization.

Hedging transactions involve buying and selling securities for the issuer’s own account and could be interpreted to be proprietary trading. It is unlikely that the complexities associated with proprietary trading can be adequately defined in a regulation. However, until a clear line distinguishes hedging from proprietary trading, many banks or bank-related entities likely will not want to take the risk of sponsoring securitization. Accordingly, the Volcker Rule may stand permanently as a serious obstacle to the development of a robust private securitization market.

There are two possible ways to solve this problem: repeal the Volcker Rule language in the DFA, or apply it solely to insured banks and not the broader “bank-related entities.” This will enable bank holding companies and their affiliates to engage in securitization without fear of violating the highly technical Volcker Rule when it is ultimately finalized. Qualified Mortgage. Section 1412 of the DFA also outlines a concept called a qualified mortgage (QM), defined roughly as a mortgage that a borrower can afford. A number of severe consequences exist for providing mortgage credit to a borrower who could not afford to meet obligations on the loan, but the most significant is that a violation of the QM requirements can serve as a defense to foreclosure. This penalty extends to the investor as well as the securitizer and, unless mitigated by a bulletproof safe harbor, will be a serious obstacle to private securitizations. Although the DFA has a provision described as a safe harbor, it requires compliance with so many qualitative elements that it offers no assurance of safety to either securitizers or investors. Without a clear and unambiguous safe harbor, this provision is a strong disincentive for anyone to securitize a mortgage or invest in MBS.

The FDIC’s Safe Harbor. Although not specifically required by the DFA, a recent Federal Deposit Insurance Corporation (FDIC) rule has created another impediment to banks’ entry into the securitization business. In 2000, the FDIC adopted a safe harbor to address legal uncertainty about when it would use its authority as a failed bank receiver to reacquire assets the bank had sold in a securitization. The investors in any securitization want to be sure they own and can dispose of the assets in the securitization pool. As receiver for a failed bank, however, the FDIC has the right under certain circumstances to reclaim property the bank has transferred.

In its 2000 safe harbor regulation, the FDIC promised that it would not exercise this right when the securitization met the accounting standards for a true sale. However, changes in accounting rules in 2009—stimulated by some of the events of the financial crisis—provided that under some circumstances assets previously sold to a trust could be consolidated with the bank’s assets, vitiating the FDIC’s 2000 rule. Accordingly, in September 2011, the FDIC adopted new standards for when a securitization would qualify for safe harbor. Many of these, such as adequate disclosure to investors, require qualitative actions that could, if not properly done, be the basis for repudiating a bank’s securitization transaction prior to the bank’s takeover by the FDIC. As a result, some investors may become reluctant to purchase securities in a pool put together by a bank, posing another obstacle to the revival of a robust private securitization market.

Basel III. Another impediment to bank participation in securitizations involves the treatment of mortgage servicing rights (MSRs) under Basel III. MSRs are an important part of the compensation that securitizers receive. Current accounting rules require that MSRs be capitalized and amortized over time. The new Basel rules on bank capitalization place a limit of 10 percent on the use of MSRs in computing tier one capital. This substantially reduces the value of MSRs to any financial institution covered by the Basel requirements. It also creates a number of other difficulties, the most important of which is an incentive on the part of any bank mortgage originator to sell its mortgages to the GSEs or FHA with servicing released.