The Bush administration, now preparing for its second term, has confirmed that it supports the creation of a new regulator to oversee all the government-sponsored enterprises concerned with housing--Fannie Mae, Freddie Mac, and the twelve Federal Home Loan Banks--and that it favors increased competition among these entities. This paper makes specific recommendations for structural and regulatory requirements to ensure that competition is enhanced along with regulation, with resulting improvement in the performance of the American mortgage-finance system. Although they would not require it, these changes would also prepare for the possibility of privatization of all the housing GSEs.
With the reelection of President Bush and with a stronger Republican majority in the new Congress, legislation addressing regulation of the government-sponsored enterprises concerned with housing now appears highly likely. From the point of view of economic policy, it is essential that regulatory reforms be pro-competitive. This indeed is the administration's position.
Treasury under secretary Brian Roseboro recently said, as reported by Reuters News, that "the administration welcomes the idea of greater competition from Federal Home Loan Banks with Fannie Mae and Freddie Mac in buying mortgages and holding them or selling them in capital markets," and further that "we believe increased competition is good, that [it] increases the breadth and depth of the capital market." On a closely related subject, the under secretary reconfirmed the administration's position that "the importance and the evolution of our housing finance markets require that all of the housing enterprises be included in a single program of world-class supervision. We see the need for this for the Federal Home Loan Banks just as we see it for Fannie Mae and Freddie Mac." Such a change would reflect the fact that "over time the activities of the Federal Home Loan Banks and the other housing GSEs have converged."
Another Treasury official, Gregory Zerzan, deputy assistant secretary for financial institutions policy, reiterated the message that the administration wants "legislation that combines the oversight of housing GSEs into a single regulator." "Having a single regulator," Mr. Zerzan convincingly explained, "should create the institutional expertise necessary to monitor and evaluate those risks both with regard to the enterprises individually, as well as with an eye towards the health of the housing finance system as a whole."
Our August 2004 Financial Services Outlook, "The Housing GSEs: Through Competition to Privatization," argued that the Federal Home Loan Banks (FHLBs), twelve GSEs with aggregate assets of about $900 billion, offer the opportunity to develop a much more competitive secondary-mortgage sector. Such a more economically efficient sector would significantly reduce the duopoly profits now enjoyed by Fannie and Freddie, as these twelve additional competitors, each also having GSE status, create supplier, product, and price alternatives. The economic benefits of GSE status would thus be pushed by competitive pressure through the GSEs into the hands of the customers, even before or without privatization.
Essential to this competitive outcome is the creation of a single regulator for the housing GSEs and a regulatory regime that fosters level-playing-field competitive conditions for all fourteen GSE competitors.
Our Financial Services Outlook argued further that a competitive secondary-market sector, enabled by a common GSE regulator, could set the stage for the possibility of complete privatization of all the housing GSEs together--with consequent private-market resource efficiency and the elimination of the government's implied guaranty of virtually unlimited debt issuance. However, that would be an independent third step, not necessarily entailed by the first two.
At least the first two of these three steps are consistent with the administration's policy position. The current troubles of Fannie Mae, meanwhile, seem to make legislation concerning the housing GSEs probable early in the new year. Senator Richard Shelby's bill, passed by the Senate Banking Committee (which Shelby chairs) last spring, focused on creating a new regulator for all housing GSEs. It is the purpose of this paper to assume the formation of this regulator is likely and to suggest a number of specific characteristics that the regulatory regime must have to promote a competitive outcome.
A focus on insuring a competitive outcome is essential if the new, stronger regulator is to avoid the perverse effect of strengthening the perceived government ties and implied guaranty, thus extending the duopoly status and duopoly profits of Fannie and Freddie.
The following topics need to be addressed:
- The structure of the new regulator
- Capital requirements
- Treatment of gains and losses on derivatives
- Government-appointed directors
- New activities
- Interest-rate risk
- Affordable housing programs
The Structure of the New Regulator
To replace today's GSE regulatory structure, the common regulator should be a new agency, as provided in Senator Shelby's bill and proposed by the administration. It should not be a continuation of either the current Fannie or Freddie regulator, the Office of Federal Housing Enterprise Oversight (OFHEO), or the FHLB regulator, the Federal Housing Finance Board (FHFB). OFHEO was designed as a very constrained regulator by the politics of its charter act of 1992, and the FHFB is the lineal descendant of the late and unlamented Federal Home Loan Bank Board, with continuing baggage of that connection (as one example, its appointment of directors to the boards of the regulated FHLBs). Both OFHEO and the FHFB were designed to oversee too narrow a domain. The new organization, with the ability to view the housing GSE sector as a whole, would have a broad perspective and more balanced bases of comparison.
As the Shelby bill provided, the common GSE regulator should be headed by a director, which would represent a significant and professionally attractive responsibility. The board model of the FHFB has proved an unappealing structure, which has had problems functioning well since its inception in the early 1990s.
Differential capital requirements are a source of regulation-created competitive advantage or disadvantage, as well as regulatory arbitrage. The leverage capital requirement for Fannie and Freddie, written into legislation in 1992, is lower than for the FHLBs or for private financial institutions, and it is one source of Fannie and Freddie's very rapid growth and very high return on equity.
This requirement is for minimum regulatory capital of 2.5 percent of on-balance sheet assets. The corresponding requirement for FHLBs is 4 percent--or 60 percent more capital required against the equivalent assets. The FHLB requirement is also set by legislation, the Gramm-Leach-Bliley Act of 1999. In the wake of Freddie and Fannie's respective accounting scandals, OFHEO has imposed on both of them a 30 percent required capital surcharge. Their required capital ratio then becomes 3.25 percent (1.3 x 2.5 percent), leaving the normal FHLB requirement still 23 percent higher.
Nearly everyone, domestically and internationally, agrees that portfolios of residential mortgages merit lower capital requirements than those consisting of other types of loans. As confirmed by the conclusions of the Basel II international risk-based capital project, in the case of capital for credit risk only, the economic capital required for residential mortgages is much lower. What is the right leverage-capital requirement for housing GSEs, specialized in this high quality asset or its variations, such as mortgaged-backed securities and mortgage-collateralized FHLB advances?
The new GSE regulator should be charged with answering this question and developing capital requirements for all housing GSEs in which the same risk requires the same capital--in other words, which are pro-competitive. The congressional charge to the new regulator should explicitly require that it create as level a regulatory playing field as possible within its fourteen-GSE domain.
A similar charge is needed for risk-based capital requirements. The risk-based capital approaches under current rules are completely different for Fannie and Freddie, on one hand, and the FHLBs, on the other. In my opinion, the Fannie and Freddie risk-based capital approach using stress tests, while very complex, may be the better concept. In any case, the new GSE regulator should develop a single, consistent risk-based capital system for all housing GSEs. Given their specialization in bond market-financed mortgage assets, this might well look different from the approaches for depository institutions, but it would apply consistently across the $5 trillion housing GSE sector.
Treatment of Gains and Losses on Derivatives
All practitioners and observers of financial markets are aware of the vagaries introduced into financial reporting, especially for holders of mortgages, by Financial Accounting Standard (FAS) 133. (This standard was supposed to clarify the effects of derivatives.) Indeed, the issues created by FAS 133 have been central to the accounting problems and disputes at both Fannie and Freddie. One effect of applying the complex rules of FAS 133 is that large amounts of "accumulated other comprehensive income" (AOCI) are apt to result, residing in the net worth section of the balance sheet.
Reflecting the conclusion that the economic meaning of these numbers was in considerable doubt, all financial institution regulators, and also both OFHEO and the FHFB, excluded AOCI from the calculation of regulatory capital.
The current debates about Fannie Mae's accounting make it apparent that Fannie's reported AOCI, in addition to unrealized marks to fair value, contains realized losses on terminated interest rate swaps to the tune of over $8 billion. These notable losses represent cash that is without question gone and can never return. However, in a provision that is itself quite debatable, FAS 133 not only allows but requires the deferral of such losses into AOCI, instead of their recognition as an expense in the income statement.
Whatever one may think of this accounting rule, a strong recommendation to the new GSE regulator is that realized losses (or gains) on derivatives must be included in the calculation of regulatory capital, while the rest of AOCI should continue to be left out.
President Bush has already declined to make appointments to the boards of Fannie and Freddie. In the case of the FHLBs, the FHFB has continued to appoint its allotted number of directors to their boards. The appointment of directors by the regulator to regulated entities has been criticized multiple times in government studies as an undesirable conflict. Moreover, it is obvious that a common regulator of all housing GSEs could not be appointing directors to some and not to others.
Congress, in structuring its GSE reform bill, should simply eliminate all government-appointed directors from housing GSEs--thereby clarifying their relationship to the government and the regulator and removing complex conflicts of interest.
Most critics of Fannie and Freddie's regulation point out the problems of divided regulation between HUD and OFHEO and argue that regulation of new activities is too weak. The administration has included approval of new activities in its list of responsibilities that should be vested in the new GSE regulator.
Most observers of the FHLBs, on the other hand, maintain that the FHFB, reflecting its historical legacy, has typically descended too far into micromanaging the details of the FHLB's activities.
The new GSE regulator should have a common approach to new activities approval, focused both on preserving the GSEs' housing finance mission and on fostering innovation and competition.
GSE debates, including those about risk-based capital discussed above, cannot avoid the interest-rate-risk issues that leveraged portfolios of fixed rate, freely pre-payable mortgages necessarily involve.
The regulatory regimes for interest-rate-risk management at Fannie and Freddie, on one hand, and the FHLBs, on the other, are extremely disparate. The common GSE regulator will have the opportunity, and the responsibility, to develop a sound and equitable approach to this issue so that the same interest rate risk position has the same treatment among all housing GSEs. This principle is fundamental. It applies whether the various GSEs have the same or different asset mix.
The administration has consistently maintained that receivership, with its implications for holders of GSE securities, must be possible under GSE reform legislation. As Assistant Secretary of the Treasury Wayne Abernathy said recently, "There is a need to have effective receivership language." The goal is to encourage market discipline to be exerted by GSE debt investors and to bring further into question their belief that they would benefit from an implied guaranty by the U.S. government.
These are sound objectives, and, given the extremely low-probability events under consideration, they may not be particularly onerous to the GSEs. An alternative, which would reflect the fact that GSEs are more like debt-issuing corporations than like depository institutions, would be simply to insert into the GSE charter acts the provision that they are subject to the bankruptcy code. This would entail "effective receivership language" in a very straightforward way.
Although it would make the legislation more complex and it would involve other committees of jurisdiction, in a perfect GSE reform bill, taxation would also be made equivalent among housing GSEs. This would involve applying federal corporate-income taxes to the FHLBs, as is already the case with Fannie and Freddie. The anachronistic "REFCORP" payments now made by the FHLBs to the Treasury should then be turned into a tax credit. These payments are a relic of the now long-ago thrift bailout of 1989. As an effect of applying de jure (as opposed to the current de facto) income taxation, FHLB dividends paid to other corporations would become normal inter-corporate dividends eligible for the dividends-received deduction, which would give FHLB shareholders the normal tax treatment. From the Treasury Department's point of view, it could look forward to receiving total payments from the FHLBs reflecting the 35 percent corporate tax rate, rather than the 20 percent REFCORP rate.
This equivalency of taxation would not be an aspect of creating a common regulator, but it would be a pro-competitive element of reform, removing disparities among GSEs.
Proponents of Fannie and Freddie often point out that FHLBs have no affordable housing goals as a percent of lending. FHLB representatives reply that all of Fannie and Freddie's affordable housing lending is a business done for profit, whereas under their own affordable housing programs, FHLBs actually give away hundreds of millions of dollars, the requirement being 10 percent of after-REFCORP profit, a contribution with much greater cost for the donor and arguably more benefit to the recipients. The FHLB requirement is statutory, from the Financial Institutions Reform, Recovery, and Enforcement Act of 1989. The specific levels of Fannie and Freddie's goals are regulatory.
The new GSE regulator should, as part of its responsibilities to ensure a level competitive playing field, undertake to understand the comparative benefits and costs of these affordable housing programs. In the short run, both programs could continue; in the long run, such understanding could lead to their modification or harmonization.
Finally: Privatization As an Option
The creation of a common GSE regulator and a competitive GSE secondary market sector could lead to the ultimate goal of privatization of all housing GSEs at the same time--although, as we have noted, it would not necessarily do so.
All steps proposed here for the common GSE regulator are consistent with privatization and should be supported by those who (like me) favor ultimate privatization. However, they can be equally embraced by those who want reform and greater market discipline, but do not support privatization and believe in the continuing value of having GSEs. These proposals leave the longer term open to alternate possibilities, while they would greatly improve the present situation.
1. Reuters News, "Bush Still Seeks GSE Reform--Treasury's Roseboro," September 14, 2004.
2. Remarks of Brian Roseboro at FHLB Stockholders Forum, September 14, 2004.
3. GSE Report, "Treasury Representatives Reiterate Call for Single GSE Regulator," September 20, 2004.
4. Market News International, September 16, 2004.
Alex J. Pollock is a resident fellow at AEI.