W. Scott Frame
Financial Economist and Policy Adviser
Federal Reserve Bank of Atlanta
Bankers Club of Chicago
May 27, 2010
Over the past several months, we have all been closely following the legislative process surrounding financial system reform. That process appears to be coming to a close as Capitol Hill looks ahead to a summer recess followed by midterm elections.
Notably, an important set of issues relating to housing finance was largely not addressed in the recent legislative process. Chief among these is a reconsideration of the future scope of government involvement in our mortgage finance system—both in primary and secondary markets. I expect housing finance reform to be a legislative priority in 2011, driven primarily by a desire to resolve the uncertainty surrounding the future of Fannie Mae and Freddie Mac. Together, these government-sponsored enterprises have already cost taxpayers over $111 billion.
Tonight, I would like to share my thoughts on redefining the scope of government involvement in secondary mortgage markets. In particular, I would like to outline some options for the conventional-conforming market historically served by Fannie Mae and Freddie Mac. Please bear in mind that these are my own views and not necessarily those of the Atlanta Fed or the Federal Reserve System.
Current government participation in mortgage markets
I would like to begin by providing some background on the federal government's current participation in mortgage markets.
The federal government directly and explicitly projects itself into the mortgage finance space through the provision of FHA (Federal Housing Administration) and VA (Veterans Affairs) mortgage insurance and the guarantee of the timely payment of principal and interest on securities backed by such loans through Ginnie Mae (Government National Mortgage Association). These programs have historically been self-sustaining and are typically used by first-time and low- and moderate-income homebuyers. In 2009, FHA and VA mortgage insurance was obtained for about 25 percent of all mortgage originations, and Ginnie Mae accounted for 25 percent of all mortgage-related security issuance. These market shares are high by historical standards.
Some analysts have identified potential issues with the FHA program that may ultimately result in significant cost to taxpayers. As a result, it is likely that we will see additional programmatic changes—especially in terms of underwriting standards and the financing of insurance premiums. Nevertheless, I expect that the FHA and Ginnie Mae will continue to play an important, but more limited, role in our mortgage finance system.
The government also participates in secondary mortgage markets—albeit in an indirect and more opaque way—through Fannie Mae and Freddie Mac. These government-sponsored enterprises, or GSEs, are quasi-public/quasi-private financial institutions. On one hand, each was created by Congress and maintains an exclusive federal charter. On the other hand, shares of Fannie Mae and Freddie Mac are traded on the New York Stock Exchange. This unusual governance arrangement results in two sometimes opposing corporate objectives: one, fulfilling certain social policy goals and supporting related political constituencies, and two, maximizing shareholder value.
By law, Fannie Mae and Freddie Mac are limited to operating in the secondary mortgage market. Their participation in this market takes two forms. The first is the issuance of credit guarantees against mortgage pools, called securitization. And the second is leveraged investment in mortgages and mortgage-backed securities. Today, Fannie Mae and Freddie Mac together hold about $1.7 trillion in assets and have another $4 trillion in net credit guarantees outstanding. This $5.7 trillion in obligations reflects about one-half of all U.S. residential mortgage debt outstanding.
Specific provisions in the GSEs' charter acts have long created a market perception that each institution's debt and mortgage-backed obligations are implicitly backed by the U.S. government. Such perceptions significantly reduce the required yield on securities issued by Fannie Mae and Freddie Mac, and some of these savings are then passed on to homeowners with conventional-conforming mortgages.
As we are all aware, once residential real estate prices began falling in 2007, the incidence of mortgage default and foreclosure increased dramatically. As a result, both Fannie Mae and Freddie Mac became financially distressed in 2008—owing to their singular exposure to U.S. residential real estate and thin capital cushions. And following the deteriorating housing and mortgage market conditions, the availability of new mortgage credit receded both in terms of prices and quantities.
The federal government responded to this situation by placing both Fannie Mae and Freddie Mac into conservatorship and entering into senior preferred stock agreements with each institution. This intervention calmed bond market investors by effectively guaranteeing the GSEs' obligations—and thereby improving the flow of capital to the residential mortgage sector. Indeed, in 2009, conventional-conforming mortgages represented 65 percent of all originations; and together, Fannie Mae and Freddie Mac accounted for 72 percent of all mortgage-related security issuance.
So, what next?
The failure of Fannie Mae and Freddie Mac provides us with an opportunity to redefine the federal government's role in secondary mortgage markets.
One option, of course, would be to simply recapitalize Fannie Mae and Freddie Mac and "return them to the wild" through an initial public offering. If policymakers are inclined to go down this road, I believe it is important to consider certain changes. I will suggest four.
First, I would like to see significantly increased capital requirements for Fannie Mae and Freddie Mac that are consistent with those faced by other regulated financial institutions. This measure will reduce perverse incentives for large volumes of U.S. mortgage-related risks to reside in only two institutions. My second recommendation is that GSE investment portfolios be strictly limited. Fannie Mae and Freddie Mac's massive balance sheets provide little social benefit in terms of liquidity and pose material risks to the institutions themselves and the financial system more broadly. A third suggestion is that any federal guarantees for GSE obligations be made explicit and priced by the government. This clarification will provide transparency and improve financial stability. Finally, I would recommend stripping out the GSEs' affordable-housing mission. Over the years, both institutions were market laggards in this area, and it is unclear whether these particular activities had any discernible effect on affordable-housing outcomes. In fact, some analysts have suggested that the GSEs' affordable-housing goals contributed to the institutional failures.
It is quite possible that policymakers will not simply resurrect Fannie Mae and Freddie Mac but instead will take a broader view about the desirability and structure of government intervention in secondary mortgage markets. I think that this would be a healthy discussion.
In terms of "first principles," it is not clear that government-sponsored securitization is required for well-functioning mortgage markets. Nevertheless, there may be political pressure to insulate mortgage markets from significant external shocks and hence maintain stable access to mortgage credit over time. Recent testimony by administration officials seems consistent with this view. These same officials have also called for accurate and transparent pricing of any government guarantees, suggesting that any such guarantees would be explicit. Nevertheless, it is very likely that any government guarantees will be mispriced at least some of the time. If this were not the case, market participants would not seek them out.
One approach would be for the federal government to securitize conventional mortgages by issuing blanket credit guarantees for qualified mortgage pools. As mentioned before, Ginnie Mae already does this for mortgages insured by the FHA or VA. The benefit of this approach is that the explicit guarantee is directly controlled by the government, and recognition of risks and costs appears in the federal budget. However, the experience with government-run programs suggests that—relative to the private sector—such an operation would be less efficient and innovative and have difficulty retaining capable staff. Furthermore, the idea of having the federal government permanently take on a large fraction of U.S. mortgage exposure will strike many as just too much government.
Perhaps more politically palatable would be some sort of hybrid public-private model that is structured differently than Fannie Mae and Freddie Mac. Some analysts have suggested having privately owned securitizers that would have access to priced government tail-risk insurance and be subject to safety-and-soundness regulation. Relative to the fully government model, this approach would likely bring greater efficiency and innovation as well as involve more private capital at risk and hence a smaller role for government. I believe that the main challenge with this approach would be the regulators' willingness and ability to limit "mission creep" as the institutions would have an incentive to expand the government subsidy.
There are several institutional considerations that would need to be tackled with a new hybrid public-private model.
The first issue is how to charter such entities—should they be created in statute, or should chartering authority be delegated to a regulator? On the one hand, "first principles" tell us that competition is good and through this process a natural market structure will evolve. In this case, a regulator would be imbued with the ability to issue an unlimited number of charters subject to the applicant meeting some reasonable minimum criteria. On the other hand, it has been argued that important standardization and liquidity benefits are much more easily achieved with few issuers. Moreover, it has been suggested that such limits would also reduce misrepresentation and fraud. These issues merit further study.
A second issue is whether such entities would be required to be monoline firms (as Fannie Mae and Freddie Mac are today) or whether they could be affiliates or subsidiaries of other, diversified financial institutions. For example, should a large commercial bank be permitted to obtain one of these charters and securitize loan pools with explicit government backing? It is quite plausible that scope economies—or synergies—may exist between mortgage origination and securitization. However, allowing stock-owned financial institutions access to one of these charters may expand the safety net and further distort risk-taking incentives.
A third and related issue pertains to permissible ownership structures. If these entities are to be monolines, should they be owned cooperatively by originators or as stock firms by public shareholders? Our experience with cooperatives suggests that such institutions have weaker risk-taking incentives, although stock purchase requirements would add to the cost of participation.
Finally, no matter what the number and structure of these entities, we will need a regulator with political independence and broad authorities to limit taxpayer risk.
In terms of risk sharing, some analysts have suggested having the securitizer(s) purchase priced tail-risk insurance from the U.S. government that carries an explicit guarantee. However, there are some important additional questions to answer relating to the loss sharing between the private and public sectors on a pool of mortgages. For instance, what is the threshold at which the government tail-risk insurance is triggered? In instances where this level is below that of the mortgage underwriting standard and securitizers are thus exposed to risk, should some of this credit risk be shared by the mortgage originators? Put differently, should originators be required to have some "skin in the game" through a risk retention requirement? It strikes me that if such requirements are to be placed on private-label securitizations, one should consider similar rules for government-sponsored deals that face the same adverse selection problem.
Now that financial regulatory reform appears to be in the home stretch, housing finance reform is likely to be an important legislative item in 2011. A central part of these efforts will be the resolution of the GSE conservatorships and a redefinition of the federal government's role in secondary mortgage markets.
I believe that the most likely outcome is that the government will maintain a role by chartering privately owned securitization conduits that would be required to purchase government tail-risk insurance. Such conduits could actually include Fannie Mae and Freddie Mac—after each is restructured through a receivership process. They might also include the Federal Home Loan Banks or possibly even mortgage originators themselves.
The past few years have been exceptionally difficult for U.S. mortgage markets. Let us hope that the natural healing process, coupled with thoughtful changes in public policy, will lead to a much brighter future, a future where creditworthy borrowers have reliable access to mortgage credit and taxpayers are well protected.