International Monetary Fund

Please send us your feedback

IMFSurvey Magazine: Policy

What Next for Fannie and Freddie?

Fannie Mae, Freddie Mac were initially established to provide liquidity to the residential mortgage market (photo: Karen Bleier/AFP)


What Next for Fannie and Freddie?

By Paul Mills
IMF Monetary and Capital Markets Department

October 3, 2008

  • United States at crossroads for its mortgage finance system
  • Several options exist after takeover of two mortgage giants
  • Longer-term decisions left to the next U.S. administration

The United States is at a crossroads with regard to its $10.6 trillion mortgage finance system following the effective government takeover of mortgage giants Fannie Mae and Freddie Mac.

Exercising powers recently bestowed upon him by Congress, the U.S. Treasury Secretary on September 7 acquired securities issued by the two largest housing government-sponsored enterprises (GSEs)—the Federal National Mortgage Association (Fannie Mae) and the Federal Home Loan Mortgage Corporation (Freddie Mac).

At the same time, the newly established regulator—the Federal Housing Finance Agency (FHFA)—placed the GSEs under regulatory management (conservatorship), and replaced their existing managers. But this dramatic intervention leaves open a number of questions. Will the government choose an explicitly government-insured—and likely subsidized—mortgage market, or will it choose a clear private sector alternative?

In the process, the authorities need to resolve the more fundamental questions of whether the GSEs should continue to add to the large subsidies directed to homeownership in the United States and, if so, whether mortgage borrowing should be the preferred method to finance house purchase .

This article looks at what prompted the dramatic government intervention, its impact, and, most pertinently, the future of the GSEs.

What are the GSEs?

The two GSEs were initially established to provide liquidity to the residential mortgage market, thereby reducing mortgage rates and promoting homeownership. They were formed as competing private sector companies (with federal charters) in the late 1960s (although Fannie Mae was originally established in the 1930s as a public sector corporation).

They purchase mortgages that conform to standard criteria from lenders, package them into mortgage-backed securities (MBS) enhanced with credit guarantees, and then sell them in the market.

In addition, the GSEs purchase mortgage-related securities, loans, and other types of assets for their investment portfolios. These purchases are financed by debt securities issued at low yields due to investors' assumption that the Treasury would not let the GSEs default.

What prompted the GSE crisis?

Essentially, the GSEs were off-balance-sheet "special purpose vehicles" of the U.S. government. They were highly leveraged combinations of specialized mortgage bond insurer and investment fund. The GSEs had an incentive to expand their own portfolios of mortgage-backed securities using the low funding costs coming from the implicit taxpayer guarantee to benefit shareholders and managers. They also enjoyed lower capital requirements than comparable private sector entities.

The degree to which U.S. mortgage borrowers benefited from this arrangement is debatable. A Federal Reserve study estimated that 30-year mortgage rates were around 7 basis points (0.07 percent) lower on average due to their existence, with the large majority of the implicit taxpayer subsidy going to share- and bondholders and managers.

The IMF estimates that currently U.S. GSEs Fannie Mae, Freddie Mac will suffer $95-$125 billion in credit losses (photo: Paul Richards/AFP)

The GSEs had already suffered interest-rate hedging difficulties in the 1990s and accounting scandals in the early 2000s. A number of commentators, including the IMF, called for them either to be wound up or more tightly regulated, or for a fundamental change in their business objectives.

Increased capital requirements

The GSEs were able to avoid such outcomes by claiming their activities promoted homeownership and by intensive political lobbying. However, following their accounting problems, the former regulator increased the GSEs' capital requirements somewhat and froze the growth of their portfolio holdings.

The GSEs' immediate problems arose from the weakness of the U.S. housing market that resulted in growing losses on their own holdings of risky MBS (including subprime and Alt-A mortgage exposures), and increasing provisions on their mortgage portfolios and guarantees as delinquency and foreclosure rates on prime conforming mortgages (the bulk of their business) began to rise.

Simultaneously, Congress charged the GSEs with providing greater support to the falling housing market by increasing the maximum loan size they were permitted to guarantee, while their regulator relaxed constraints on their MBS portfolios. These losses and additional business lines forced the GSEs to increase fees to lenders, tighten standards, and seek more capital from the market.

Initially, they were able to raise additional capital, but, as losses continued, the housing market deteriorated and their share prices dropped, this became increasingly difficult and costly. The Treasury had hoped that the powers granted by Congress in July 2008 to extend credit and buy GSE securities would have been sufficient to restore confidence. However, although the GSEs continued to finance themselves adequately in the money and bond markets, their equity prices continued to fall.

Why did the U.S. Treasury act?

The U.S. authorities intervened for several reasons. First, due to the collapse of "private-label" mortgage securitization since the autumn of 2007, the GSEs—along with Ginnie Mae, which securitizes loans made by the Federal Housing Administration (FHA)—are effectively the only issuers of MBS, together taking 95 percent of the market in the first half of 2008 (see Chart 1).

Banks and other lenders increased the mortgages they held by only $17.5 billion in the first half of 2008—less than 2 percent of total mortgage lending. Essentially, if the GSEs had ceased doing business, mortgage availability would have relied solely on the FHA since banks are currently adding only trivial mortgage amounts on their balance sheets.

Second, mortgage rates and costs were rising as a result of yields on GSE MBS increasing (due to credit concerns) and of the GSEs increasing fees (to compensate for earlier losses). To address this and stimulate the housing market, the Treasury agreed not just to guarantee the GSEs' capital position but also to buy new GSE MBS outright.

The GSE intervention lowered 30-year mortgage rates by approximately 50 basis points (0.5 percent—see Chart 2) and raised house price futures by 1.5-1.75 percent in the process. GSE fees are also being reduced following the rescue to cut the costs to borrowers.

Promote confidence

Third, Treasury intervened to promote confidence in the GSEs' MBS and debt, because of significant holdings by banks and foreign investors. Domestic banks hold around $1.1 trillion of GSE securities with foreign holders accounting for another $1.1 trillion, including large central bank holdings.

Additional declines in the value of these securities would have further eroded the depleted capital positions of U.S. commercial banks while an insolvency and possible default on GSE senior debt securities would have been viewed as an implicit breach of trust by foreign official investors, with potentially serious consequences for the value of the dollar and the ability of the United States to attract capital inflows.

In practice, to achieve these ends and minimize potential costs to the U.S. taxpayer, the Treasury and FHFA undertook to inject sufficient senior preferred equity capital into the GSEs to ensure GSE solvency, while placing them into conservatorship, rather than receivership. This avoided giving an explicit Treasury guarantee to existing GSE debt while giving the FHFA control over the GSEs' major financial decisions. Receivership would have entailed winding up the GSEs and accelerating various debt covenants. Hence the authorities achieved control and severely diluted preferred and common equity-holders without making the GSEs insolvent.

How much will these actions cost?

In July, the Congressional Budget Office estimated a probability-weighted cost of $25 billion, ranging from zero to over $100 billion. The most pessimistic outside observers put the possible cost in the region of $200-$300 billion. The Treasury's agreement with the GSEs envisages an injection of senior preferred equity of up to $100 billion in each, but this is presented as being well in excess of likely losses, in part because current operations are regarded as profitable.

The IMF estimates that currently the GSEs will suffer $95-$125 billion in credit losses (of which some $18 billion has already been provided for). If the GSEs were to be capitalized further to private sector standards, an additional $80-$100 billion in capital may be needed, much of it recouped on any future sale to the private sector. The cost to the public purse would be lowered to the extent that the GSEs retain earnings from their current high-margin business.

The accounting treatment of the GSEs' debt and budget on the federal government's balance sheet is still being debated. However, according to international government accounting standards, the GSEs are now "controlled" by the government and so become "state-owned enterprises." Gross liabilities and assets (each around $5 trillion) are now to be counted on the general government balance sheet. This has a significant impact on U.S. general government gross debt but a negligible effect on net debt.

Thus far, the credit rating agencies have not seriously questioned the U.S. government's retention of its AAA/Aaa sovereign rating, and the GSE intervention alone should not prompt a change. However, the operation coincides with a deteriorating cyclical fiscal position, multiple other public interventions in the financial sector, and the acceleration of entitlement spending due to an aging population. (To quote Senator Everett Dirksen, "A billion here, a billion there, . . . pretty soon it adds up to real money.") Unsurprisingly, the 5-year credit default swap spread of the United States has spiked up to 30 basis points.

Whither the GSEs?

Although they had the opportunity, the U.S. authorities did not place the GSEs in receivership. This left open the question of their ultimate future for the next administration and Congress. Nevertheless, Treasury placed two "poison pills" in their agreement to ensure that the GSEs could not continue indefinitely in their former state.

"The danger with making them fully publicly owned enterprises is that . . .the the U.S. housing market becomes dependent on taxpayers taking a concentrated financial risk during a housing slump"

First, the GSEs agreed to wind down their portfolio holdings by 10 per cent a year beginning in 2010, effectively running down the investment risk they pose. Second, Treasury will be able to charge a market-based fee for the capital guarantee provided by the taxpayer, potentially eliminating the implicit subsidy the GSEs receive. These provisions almost certainly mean that the next administration will need to address the fundamental future of the GSEs once the housing market has stabilized. What are the options?

(i) Fully publicly owned enterprises: The GSEs could be fully taken into public ownership, with private shareholder claims extinguished. As with the FHA and Ginnie Mae, the GSEs would then borrow in the markets with a clear government guarantee, although at higher yields than treasuries. In essence, the taxpayer would provide mortgage and bond insurance, so subsidizing homeownership. Unless the GSEs were then to face quantity or price constraints, private sector banks or securities issuers could not compete with such an operation and would be confined to the riskiest or more esoteric lending outside the GSEs' purview.

(ii) Reformed version of the old GSE model: The GSEs could gradually be recapitalized through retained profits with the Treasury's senior preferred equity repaid and extinguished. The GSEs would fully revert to their pre-existing "private sector" status, although with regulation of their capital and investment portfolios comparable to private sector equivalents.

(iii) Break-up and true privatization: If it was deemed necessary to retain institutions that guaranteed and standardized mortgages and MBS to foster liquidity, the GSEs could be reformed into pure MBS bond insurance companies that would be stripped of their federal charters, split up into multiple institutions to ensure competition, and privatized. However, fundamental reform of U.S. insurance regulation would probably be a prerequisite for such a course of action.

(iv) Run-off: Once the housing market has stabilized, the GSEs' business could gradually be wound down, with the investment portfolios allowed to mature, and MBS issuance gradually reduced over time to allow for replacement by private sector alternatives. Any residual public policy functions could be explicitly transferred to the FHA.

The need to clarify the status and role of the GSEs and assign public and private objectives argues against option (ii). The danger with making them fully publicly owned enterprises is that the GSEs become another means whereby covert subsidies are granted to political interest groups and the U.S. housing market becomes dependent on taxpayers taking a concentrated financial risk during a housing slump. The AAA-rating of the U.S. government could legitimately be questioned under such an arrangement.

These arguments would point to options (iii) or (iv) as the best alternatives, although the transition would be complex and probably protracted. Standardization of MBS structure could be achieved through private sector coordination, as has been accomplished on many types of financial instruments to date.

Comments on this article should be sent to

Free Email Notification

Receive emails when we post new items of interest to you.
Subscribe or Modify your profile

Write to us

The IMF Survey welcomes comments, suggestions, and brief readers' letters, a selection of which are posted under What readers say. Letters may be edited. Please address Internet correspondence to