This statement is also available here as an Adobe PDF.

Statement No. 255
February 11, 2008

For Information Contact:
Charles Calomiris
212-854-8748
Richard Herring
215-898-5613
Kenneth Scott
650-723-3070

Policies designed to resolve the current home mortgage foreclosure problem should begin from an understanding of the underlying causes of this problem and particularly the role played by public policies designed to expand homeownership. An unintended consequence of these policies has been to encourage a perilously high degree of leverage among both borrowers and lenders that is a basic cause of the current problem.

Tax policy reduces the effective interest rate to borrowers and favors increased loan size and leverage by permitting the deduction of mortgage interest in computing taxable income. Middle-income home owners can obtain an additional subsidy to the extent that they negotiate lower rate mortgages that are guaranteed by Fannie Mae or Freddie Mac (Government Sponsored Enterprises (GSEs)), which, in turn, are implicitly guaranteed by the U.S. government. With regard to both tax policy and GSE guarantees, the extent of the subsidy depends on the size of the mortgage: the higher the amount of borrowing, the greater the subsidy.

For lower-income homeowners, the Federal Housing Agency (FHA) encourages leverage by providing a subsidized rate of interest and permitting down payments that are lower than would be required by private sector lenders, which could be as low as 3% on new mortgages. Moreover, the FHA has permitted cashouts on existing mortgages to reduce the borrower's accumulated equity to 5%. Again, the extent of the subsidy depends on the amount borrowed: the higher the degree of leverage, the higher the gain in subsidy.

GSEs themselves have achieved a remarkably high degree of leverage on the basis of their implicit government guarantees. Moreover, capital regulation of banks has had the inadvertent consequence of creating a highly leveraged off-balance sheet banking system that has financed a high proportion of residential mortgage debt. Since bank capital charges are applied only to mortgages retained on the balance sheet, many banks have chosen to sell mortgage loans they originate to a thinly-capitalized special purpose entity. This technique substitutes fee income for interest income and allows banks to profit from their mortgage business without increasing required capital.

While the benefits of these policies were immediately apparent, the potential costs of this highly leveraged system of housing finance were not visible so long as house prices rose steadily. Problems in servicing a mortgage could usually be addressed through a refinancing based on the increase in the value of the home since the mortgage was originated. Over the past year, as housing prices have fallen, the weaknesses of this system have become apparent. The high degree of leverage has meant that even a 5 to 10% decline in housing prices has had a devastating impact on financial markets and the real economy, a much greater impact than a comparable decline in stock market prices has had in the past. 

One consequence of the fall in housing prices, given previous public policies, has been that marginal borrowers are unable to meet their obligations. They have begun to default on their mortgages and lenders have begun to foreclose on an increasing number of properties. The increase in delinquencies and foreclosures has generated Congressional concerns and policy proposals to help borrowers who might otherwise lose their homes. It is important that any policy to mitigate the current problems avoid making the housing finance system still more vulnerable to a decline in house prices. In particular and crucially, since leverage has been a primary source of vulnerability, any remedy should not increase leverage. 

In addition, the Committee recommends that policies to deal with the current rise in foreclosures should be guided by six general principles. A proposed policy should:

1. Address the problem in a sufficiently broad way to be perceived as both effective and fair. For example, the Treasury proposal focused only those borrowers who were current but could not pay their mortgages after rates reset above the introductory "teaser" rate. This policy dealt with such a small subset of troubled borrowers that it would not have been very effective in dealing with the overall problem and it was perceived as unfair both to those who were unable to pay their mortgages before reset and to those who managed to pay.
   
2. Not undermine the incentives of the parties to negotiate an efficient, mutually beneficial outcome. For example, the prospect of government assistance may encourage both borrowers and lenders to wait for assistance rather than renegotiating their loans. 
   
3. Not undermine, by overriding contract terms, the incentives of future parties to negotiate more robust mortgages that will avoid a recurrence of these problems.
   
4. Not undercut the willingness of investors to finance residential mortgages in the future. 
   
5. Offer sufficient flexibility to facilitate a broad range of resolutions that best reflect the interests of individual borrowers and lenders and avoid the imposition of a one-size-fits-all solution. A government program that imposes a solution, rather than encouraging borrowers and lenders to renegotiate, may be harmful.
   
6. Protect the interests of taxpayers, if taxpayer funds are used explicitly or implicitly to facilitate a solution. Any subsidies should be carefully designed to avoid waste, prevent abuse and discourage gaming to increase government subsidies.

In December 2007 the Shadow Committee reviewed the Treasury proposal to mitigate foreclosures. (See Statement No. 250 on Treasury Department's Mortgage Foreclosure Program, December 10, 2007.) From the perspective of the principles enumerated above, we concluded that it was deficient in several respects. We raised concerns regarding effectiveness and fairness of the policy, its impact on incentives for borrowers and lenders to negotiate efficient solutions, and its potential distortion of incentives for the design of future mortgage contracts. Since that time several additional ideas have surfaced.

One idea is to allow the GSEs to purchase renegotiated mortgages. This would serve to increase liquidity and reduce the need for capital to support existing portfolios of mortgages by private parties. It could also encourage renegotiation of mortgages and the recognition of losses. But this approach may create the potential for unbounded subsidies to lenders at the expense of taxpayers. If renegotiation does not result in a sufficient write down of principal and interest, the mortgages may be sold to the GSEs at a price above their true market value. This could shift large losses to GSEs, and thus, potentially to taxpayers. One way to limit that potential risk would be to require that renegotiated mortgages have performed for a sufficiently long period of time before being purchased by the GSEs. Of course, the longer the required performance period, the longer the delay of GSE liquidity support to the market. Thus, there is an inherent trade-off between the immediacy of liquidity support and the risk of loss to taxpayers.

Another way to limit risk shifting would be to take the GSEs out of the renegotiation and writedown/sale picture and provide instead a government subsidy, say 20%, for such renegotiations. In a resale, private investors would have a strong incentive not to overstate current market values (and thus overpay), while sellers would have an incentive to accept realistic writedowns and the associated losses. In a writedown, the current investor would have an incentive not to understate current market values and incur a larger loss. To prevent collusive gaming against the government, the renegotiated mortgages could be required to grant the government a contractual entitlement to recapture the subsidy from a resale or refinancing of the property within a specified period. Moreover, to speed the renegotiation process, the availability of the subsidy offer could be limited to a brief period. It should be recognized, however, that it may not be possible to restrict the subsidy to investors who would not otherwise have renegotiated the terms of a mortgage contract and thus may be an inefficient use of government resources. Determining the feasibility of the whole approach would require considerable attention to designing the terms of the renegotiation process.

From the standpoint of our six principles, there are clear pros and cons to these proposals. On the positive side, to the extent these approaches succeed in reducing leverage via write downs, they could meet the first objective of reducing the vulnerability of the mortgage market to future shocks by reducing leverage. Second, the approaches are respectful of the contractual rights of lenders and borrowers, since they do not try to mandate particular terms for renegotiation. On the other hand, if measures to ensure adequate write downs are not sufficient, lenders may be able to game the system and profit from the disposal of risky loans at public expense. Furthermore, to the extent that losses are shifted to taxpayers, future borrowers and lenders may engage in risky behavior in anticipation of a similar bailout.

Another and less complicated idea for facilitating renegotiation would be to encourage servicers to enter into rental contracts with homeowners at risk of foreclosure. The servicer could accept, on behalf of the investors in the pool of mortgages, a deed in lieu of foreclosure. In exchange the borrower would be offered an affordable rental contract that would also include an option to buy the home at a predetermined price, possibly lower than the original amount of the loan, at some point in the future. The investors in the mortgage pool would receive rental income in lieu of mortgage interest payments. The potential advantages of this idea are several: this would permit people to remain in their homes and avoids the deadweight costs of foreclosure and resale. It would remove downward pressure on home prices coming from a massive influx of foreclosed homes into the market for sale. The option to purchase would encourage good maintenance of the premises. Many borrowers have zero or negative home equity and thus have become de facto renters rather than home owners. They now may simply walk away from their mortgage obligations and forfeit their homes because the option of homeownership is out of the money. For these borrowers, incentives would be greatly improved.

While this idea has merit, not the least being that the government does not rescue private parties from the foreseeable consequences of their decisions, the role of the government needs to be defined. The government could play a leadership role in clarifying legal standards for mortgage renegotiation to protect servicers who adopt this, or other potentially useful approaches to renegotiation, against the risk of litigation. Reduced litigation risk would encourage faster and more efficient resolution, but there are pitfalls to be avoided. Government encouragement of efficient renegotiation should not dictate terms, establish one-size-fits-all protocols, or trample on existing contractual rights.

Given the magnitude of foreclosures looming, a number of additional proposals are likely to be introduced and, in this political season, Congress may be pressed to adopt measures that provide short-term relief at the expense of increasing the vulnerability of the system in the long term. A recent example of this problem is the section of the proposed stimulus package supported by the House and the Administration that temporarily raises the conforming loan limit to permit the GSEs to guarantee mortgages as large as $729,750. This provides an implicit subsidy to a broader range of high-income borrowers, enabling them to achieve greater leverage. It is clearly inconsistent with the principles we propose that are designed to ensure that policies to deal with the current turmoil do not make the housing finance system still more fragile. Buying quick fixes at the expense of longer term vulnerability is not a good bargain.