Guest post by Lawrence J. White, Robert Kavesh Professor of Economics, NYU Stern School of Business.
In the summer of 2008, Fannie Mae and Freddie Mac’s financial positions deteriorated sharply: the result of inadequate capital (equity financing) for the risks in the residential mortgages that they held and had securitized. On September 6, 2008, their regulator, the Federal Housing Finance Agency (FHFA), removed senior management and placed these government-sponsored enterprises (GSEs) into conservatorships. Since then, the FHFA and the U.S. Treasury (which extended almost $188 billion to keep them solvent through 2011) have run them.
As a result of what was in effect their bankruptcy, many observers would have been willing to offer long odds on a bet that Fannie and Freddie would still be in conservatorships nine years later. They would have lost that bet. The Dodd-Frank Act of 2010 was silent on the subject, as is the House-passed Financial CHOICE Act. While both firms have stabilized and run cash-flow surpluses since 2012, there is no resolution in sight.
Although it is not an occasion for a celebration, this ninth anniversary provides a good opportunity to review the U.S. housing finance system and Fannie and Freddie’s role. This review addresses U.S. housing policy more generally, offering thoughts about sensible ways forward.
Let’s start with what Fannie and Freddie do: They operate in the secondary residential mortgage market (primarily mortgages for single-family homes, but also for multi-family). That is, they buy already-originated mortgages. Then, either they package the mortgages into pass-through mortgage-backed securities (MBS) and sell the MBS (primarily to institutional investors); or they hold the purchased mortgages on their own balance sheets, funding the assets primarily by issuing bonds. In the first case, in return for a fee, Fannie and Freddie guarantee that the MBS investors will receive their principal back even if the underlying mortgages borrowers default. Thus, although the MBS investors bear the interest-rate risk (including pre-payment risk) on these securities, Fannie and Freddie bear the credit risk.
There is a good reason that we call Fannie and Freddie GSEs. Although each has traditionally had the “look” of an ordinary corporation, complete with a board of directors and shares traded on the New York Stock Exchange, both are creatures of the federal government. Their charters come from acts of Congress (rather than from, say, the state of Delaware); the President of the United States could appoint 5 (of their 18) board members; they had a special line of credit with the Treasury; and they enjoyed various other privileges that were not available to ordinary corporations.
Fannie and Freddie have always been and continue to be special. Prior to 2008, their stand-alone financial conditions would have generated a AA- bond rating from the credit rating agencies, but they were generally able to borrow at better-than-AAA interest rates (barely above Treasury rates). Despite the disclaimers of various Treasury Secretaries—as well as the statement on their securities that they were not full-faith-and-credit obligations of the federal government—the debt markets acted as if the Treasury would rescue the GSEs (in reality, rescue their creditors) if they got into financial difficulties. The events of September 2008 proved market participants correct. And since then, investors have been even more inclined to treat the GSEs’ securities as safe.
The GSEs do face important restrictions. They cannot originate mortgages, nor can they become involved in financial (or other) activities outside of residential mortgages. Their regulator caps the value of the conforming residential mortgages that they can buy (and guarantee): as of 2017, the ceiling for a single-family mortgage is $424,100 in most states, but rises to $636,150 in “high-cost” areas, such as coastal California and expensive metropolitan areas. (For comparison, in June 2017, the median sales price for an existing home was $263,800, so that an 80-percent mortgage would be about $210,000.) Conforming mortgages also are supposed to have a 20-percent down payment (or other another source of credit support, such as private mortgage insurance), and the borrower is expected to meet appropriate creditworthiness standards. Finally, the FHFA also obliges the GSEs to act in a manner that supports affordable housing, especially in underserved communities.
The net impact of all of this specialness is that Fannie and Freddie subsidize home ownership. Prior to 2008, the common academic estimate was that conforming mortgages carried interest rates that were about 20-25 basis points lower than they would have been in the absence of the GSEs. The continued interest-rate spread between conforming loans and jumbo loans that exceed the regulatory caps suggests the subsidy persists (despite substantial increases in the guarantee fees charged for MBS).
Another consequence of the GSEs’ specialness is their size. As of 2008, Fannie and Freddie together guaranteed $3.5 trillion in outstanding MBS and had another $1.5 trillion in residential mortgages directly on their balance sheets. Combined, this accounted for nearly one half of the $10.5 trillion of residential mortgages then outstanding. Although they have shrunk the mortgages on their balance sheets, they have expanded their MBS, and as of 2017 they still account for over 40% of residential mortgages.
The continued subsidy and scale of the GSEs are not accidental. They are simply one component of a larger set of public policies and institutions that encourage the construction and consumption of housing—mostly to promote home ownership, but for rental housing. The list is long, including an alphabet soup of federal agencies (FHA, VA, USDA, and HUD), favorable tax treatment (including the deductibility of mortgage interest payments and capital gains exclusion), and a myriad of state and local government provisions designed to encourage more housing. (For an extended discussion, see here.)
A reasonable characterization of U.S. public policies with respect to housing is: “Too Much Is Never Enough!”
The results are not surprising. Subsidies bias the U.S. economy’s investment in favor of residential housing, leaving the country with insufficient plant and equipment for goods and services production; poorly maintained roads, bridges, tunnels, schools, hospitals, and other publicly provided infrastructure capital; and too little education and training (human capital).
Further, although advocates often claim that many of these the policies encourage home ownership―helping to achieve “the American Dream”―the data don’t show much responsiveness to these policies. As the following chart shows, the U.S. home ownership rate is currently at 63.7 percent: down from a peak of 69.2 percent in 2004, and below the levels of the late 1960s! Compared with other advanced economies, the United States usually appears only in the middle of the pack, even though those countries engage in far less subsidy of home ownership (see the table below).
Home ownership Rate (Percent), 1965-2Q 2017
Advanced Economies: Home Ownership Rates
At least as bad as the overall impact are the distributional implications of U.S. home ownership subsidies, many of which are skewed toward upper-income households. The mortgage interest deduction is a prime example: First, as a deduction, rather than a tax credit, it benefits only those who itemize their deductions, something that upper-income households are much more likely to do. Next, holding the interest rate constant, the subsidy is proportional to the size of the mortgage (this is also true for the GSEs’ subsidy). Upper-income households are far more likely to buy a more expensive house, obtain a larger mortgage, and thus receive a larger subsidy.
The point is that these subsidies do little to encourage greater home ownership. Instead, they push upper-income households who already buy homes to purchase larger, better-appointed houses on larger plots of land. It is hard to see the social benefits of such an outcome.
What would better policies look like? Let’s start with housing policies more generally.
The U.S. economy devotes too many resources to residential housing and not enough to other socially-beneficial investments. Subsidizing mortgage interest rates is largely a recipe for encouraging households to borrow, at the same time that it favors higher-income households. And, the 30 to 35 percent decline in housing prices after 2006 should debunk the idea that home ownership is a sure-fire way for individuals to build wealth. Houses are large, undiversified, and illiquid assets that place financial health at risk at the same time that they impair geographic mobility, hampering pursuit of better employment.
All of this favors reducing the subsidy for housing. Phasing-out the mortgage interest deduction would be a fine place to start. Politically based references to home ownership as part of “the American Dream” should similarly be phased out. The home ownership statistics referenced above should be de-emphasized. Renting should be considered a respectable alternative. And policy should always be grounded on the important insight that income transfers in cash are usually more helpful than equivalent transfers in kind.
Where there are positive externalities, policies should be targeted on the externality. For example, to encourage home ownership by first-time low- and moderate-income homebuyers, subsidizing down payments would be a far better approach than the broad mortgage interest tax deduction and the GSE subsidies. Policies also should be transparent and on-budget, rather than implicit and hidden. The FHA, with its transparent, on-budget subsidies focused more on lower-priced houses and on first-time buyers, meets this test. The GSEs do not.
Policy should also focus more on supply-side considerations aimed at reducing the cost of housing without the need for subsidy. These include the elimination of: a) protectionist restrictions on imports of building materials, such as lumber and cement; b) local building codes that increase the cost of housing construction without providing commensurate benefits in safety; and c) suburban exclusionary zoning and land-use restrictions that make it difficult to build multi-family housing and smaller houses on smaller lots in many suburban communities.
As for housing finance, reforms should be based on the following principles: a) create an efficient mortgage finance process; b) limit systemic risk through prudential regulation that has at its heart substantial equity financing requirements; and c) limit the use of the housing finance system for the delivery of subsidies to favored groups (while making any subsidies transparent).
Having begun nearly 50 years ago, securitization is a mature, efficient technology. It allows the funding for mortgage finance to expand beyond the deposits in the banking system, avoiding the maturity mismatch bedeviling depository institutions that fund holdings of long-term mortgages with short-term deposits. And, it allows a separation of the functions of mortgage finance (e.g., originating, funding, servicing) that encourages the efficiencies of greater specialization. But, as the debacle of subprime securitization in the 2000s revealed, securitization introduces agent-principal problems that the vertically integrated mortgage process of depository institutions had largely avoided.
With an appropriate regulatory structure in place, re-starting a private-label MBS process (involving neither Ginnie Mae not the GSEs in their current incarnation) should be possible. Life insurance companies and pension funds, which have long-term liabilities, are among the “natural” purchasers of such long-term assets. The regulatory structure needed to support this new MBS system will have at least two main functions: a) monitor the quality of the underlying mortgages that are included in the MBS; and b) regulate the MBS issuers, which are likely to be large and systemic, issuing GSE-like guarantees for investors. Again, adequate equity financing (specified by capital requirements) must be at the center of such regulation. Given its current expertise, the FHFA may well be the right regulatory agency for these tasks.
If these markets cannot function without some form of federal guarantees against tail-risk events, then such guarantees should be transparent and priced in a way that avoids public subsidies. One way to approximate this would be for the federal guarantees to exist side-by-side with private guarantees and to be priced on par with them.
Finally, what of Fannie Mae and Freddie Mac? The FHFA has been forcing them to shrink their portfolios and to transfer greater amounts of their credit risks to the private sector, through purchases of insurance and through the issuance of securities that function like catastrophe bonds, whereby the bond buyer is repaid less principal in the event that there are credit losses on the underlying mortgages. In the short run, this portfolio shrinkage and credit risk transfer (CRT) should continue.
In the longer run, as the private-label MBS system develops, the federal government should sell shares in Fannie and Freddie, allowing them to become private-label issuers, too. The 1987 privatization of Conrail could be a good precedent. Of course, the process should emphasize that the federal government is cutting all of its ties with the two organizations, and that the resulting private entities will need to maintain adequate equity and meet the other prudential requirements that the FHFA (or its successor) imposes on all MBS issuers.
Unfortunately, with almost nine years of the GSEs’ conservatorships behind us, it is hard to be optimistic that reforms such as these will be undertaken. It is quite possible that the 10th anniversary will merely see the continued pressures for “Too Much Is Never Enough!” But if there is a political will to make changes, the direction for desirable change, as well as many of the specifics, are clear.
Note: Professor White's previous guest post was Regulating the Credit Rating Agencies? Less Would be More.