Fannie Mae and Freddie Mac, once the twin kingpins of the home loan market, have been languishing in Federal conservatorship since September 6, 2008. While their continued presence is an embarrassment to many on both sides of the political aisle, it is generally understood that the agencies can't be axed without seriously disrupting the market, and that a phase-out should occur over a period of years -- to coincide with a phase-in of the institutions that will replace them. The problem has been the absence of a coherent game plan for developing those institutions.
But now there is at least a preliminary game plan, which is contained in the draft on housing finance reform recently released by leaders of the Senate Banking Committee. The draft has received a positive response in some quarters as evidence that bipartisan cooperation is still possible. The bipartisan consensus, however, is pretty much limited to the view that Fannie Mae and Freddie Mac have to go. The structure proposed to replace the agencies is replete with provisions from the left designed to promote affordability for the disadvantaged, and from the right designed to protect taxpayers.
A lot of ink is being spilled on which side of the aisle is getting the better of the bipartisan deal, but that is a sideshow. The critical question is whether the basic structure proposed as a replacement for Fannie/Freddie would meet the objective of creating effective secondary mortgage markets?
The Proposed Structure
The proposal would eliminate the major structural weakness of the Fannie/Freddie model, which was the blending of private shareholders and political meddlers. Under pressure during the bubble years to meet both the demands of investors for rising earnings and the demands of politicians for rising allocations to disadvantaged groups, the agencies assumed massive risks that did them in when the bubble burst.
The Federal Mortgage Insurance Corporation (FMIC) that would replace Fannie/Freddie would be wholly owned by the federal government, with insurance functions similar to those of FHA, and regulatory functions similar to those of FDIC. Its weaknesses would be those of government corporations, which are much better understood than those of private/public hybrids.
FMIC would insure the mortgage-backed securities that meet its standards, and would regulate the various participants in the security creation process. These include originators who make the loans for sale in the secondary market, aggregators who pool mortgages and sell insured securities, and guarantors who place their guarantee in front of that of the government and will be subject to a capital requirement of 10 percent. The major objective is to provide the same degree of security to investors in mortgage-backed securities as Fannie/Freddie do now, but with the private sector assuming a major part of the risk exposure.
Could the Proposed Structure Work?
The model proposed is a new one that has not been tested, and to my knowledge has no antecedents anywhere. Whether or not it will work as desired, therefore, is not clear.
There will be no shortage of loan originators, since we have a large industry of mortgage banks that do this; and there will be no shortage of aggregators, because we have a large industry of investment banks; but whether there will be private guarantors willing to do what the scheme requires is uncertain.
Guarantors play a critical role in the draft proposal because their capital is the buffer against loss by the government. Only when the guarantor 's capital is depleted does the government insurance kick in. And while there are now many firms that guaranty securities, including mortgage insurance companies, these guarantees are always limited. None of them will bet the firm on a single security. To avoid such risk concentration under the draft proposal, the guarantor will need an enormous amount of capital to place at risk.
For example, if a guarantor backs 10 securities of $1 billion each, which provides only limited risk diversification, it would require $1 billion of capital to meet a 10 percent requirement. It is not at all clear that the premiums it could charge would justify an investment of this magnitude.
Of course, the capital requirement could be scaled down, as could the requirement that the guarantor assume 100% exposure on every security. But such adjustments would be resisted by those determined to "protect the taxpayer".
No Provision For Reviving a Needed Private Secondary Market
A major omission from the draft proposal is a game plan to create a private secondary market that would be more robust than the one that collapsed during the financial crisis. Indeed, the need for a private secondary market is not even recognized in the draft.
The need arises from the excessively restrictive underwriting rules adopted by Fannie/Freddie after the crisis, which are bound to be adopted by FMIC. Discretion in the loan underwriting process has been largely eliminated and large numbers of good loans, including loans to the self-employed and investors in particular, are not being made. A recent study by Laurie Goodman, Jun Zhu and Taz George estimates an annual shortfall of over a million loans because of reduced availability.
A newly-constituted private secondary market would make such loans possible, but that market should be more robust than the one that collapsed during the financial crisis.
Weakness of the Now-Defunct Private Market
The critical weakness of the market that imploded during the crisis was the lack of risk-sharing among securities. Every security had "credit enhancements" that were designed to allow each one to stand on its own feet, and there was no provision for redistributing the enhancements to where they were most needed. This meant that if nine securities had more credit enhancement than they needed and one had less, that one would fail.
In this structure, the issuer of a security had no liability except for whatever commitments the issuer had contributed to the credit enhancement. The Dodd/Frank legislation attempted to deal with this by requiring issuers to have 5 percent exposure, but there have been no takers.
Could the Private Secondary Market Reemerge Under the Draft Proposal?
On the optimistic assumption that the guarantors required by the draft proposal emerge to do what is required for FMIC to replace Fannie/Freddie, at some point they might well expand their reach into the private market. They would do this by reducing their premiums, liberalizing their underwriting requirements, and limiting their exposure to each security they guarantee, perhaps to 5 percent. Essentially, this would recreate the same type of market structure that existed prior to the crisis, with every security standing on its own bottom.
In sum, the new model designed by the Senate Banking Committee may or may not provide an adequate replacement for Fannie/Freddie, and even if it did, it would not provide the basis for a robust private market.
A Thoroughly Tested Alternative Model
The Committee has another way it can go. It can adopt the Danish model which combines originators, aggregators and guarantors into one entity, called a mortgage bank. The mortgage-backed securities issued by the bank would be guaranteed by the government, but as liabilities of the bank, they would also be protected by the total capital of the bank.
A major advantage of the mortgage bank approach is that it should evolve into a robust private secondary market. As the banks establish their operating record, they will begin offering securities that carry only their own guarantee, and eventually the government will be out of the picture altogether -- as is the case in Denmark.
There has never been a default on a mortgage security issued by a Danish mortgage bank. During the worst phases of the recent financial crisis, it was business as usual in the Danish market.
The Danish model should appeal to the right side of the political aisle because the risk exposure of the government is buffered by 100 percent of mortgage bank capital, and over time the government's exposure will disappear altogether. The Danish model should appeal to the left side of the political aisle because it drastically simplifies the mortgage lending process for borrowers.
In contrast to the US system, where months may pass between the date when a loan is closed and the date when the loan is converted into a security, in the Danish system, each borrower is funded directly by the secondary market. The mortgage bank places the mortgage directly with investors simply by adding it to an open bond issue covering the same type of mortgage. This means that borrowers can shop secondary market prices on-line to find the best price for their loans, leaving only the mortgage bank's markup to be negotiated with the bank.
The system could also be used as an efficient way to channel government support to disadvantaged groups. This could be done by creating one or more special mortgage securities on which the government would bear the risk.
In short, the Danish mortgage bank model could provide the basis for a true bipartisan approach to mortgage reform.