How to save the housing market

As the Senate turns to crafting its own version of the stimulus package, many are calling for amendments that directly target the plummeting housing market. To be sure, house prices have been inflated in recent years, so a measured decline in house prices is warranted. The problem is that house prices have already fallen to their pre-bubble levels, and economic forecasters and financial futures markets suggest that, without action, house prices will continue falling into 2010.

These declines have a real impact on the economic climate. About two thirds of Americans own houses, and for most of them their houses constitute a substantial portion of their net wealth. As house prices decline, homeowners reduce consumption and save more to compensate. Stabilizing house prices is therefore a critical component of an economic recovery plan. Several Columbia University professors have recently released a policy proposal to address the collapsing housing market. The plan has two components, one or both of which are likely to be included in some form in proposed amendments to the Senate fiscal stimulus.

Lowering mortgage rates

The first piece, published by Glenn Hubbard (dean of Columbia Business School) and Christopher Mayer (Senior Vice Dean) calls for an immediate reduction in mortgage interest rates. Mortgage rates have historically hovered about 1.6 percent above the ten-year Treasury bond rate, but that spread has recently increased. So although the government and banks can borrow at historically low rates, those rates haven't been passed along to borrowers seeking to buy homes. Reducing rate spreads to their normal levels would bring mortgage rates down as low as 4 percent. Not only would these low rates allow many Americans to refinance their mortgages, translating to an average monthly payment reduction of over $400 per month, it would also bring many new homebuyers into the market, thereby mitigating the current excess housing supply.

Dean Baker says this implicitly means that new buyers will lose money on their homes in the future, since these low rates would inflate house prices above natural levels. But this criticism misses the point of the empirical analysis that precedes the 4 percent proposal, which looks at where house prices should be. If they've already fallen below their efficient level, and they're still falling simply because of ill-functioning credit markets, Baker's fear that we're going to create a new bubble is unfounded.

Some would argue that, in this market of high default rates and economic uncertainty, mortgage spreads should be more than their historical levels. But this ignores a crucial component of the cost of mortgage lending. In mortgage lending, investors face two risks. The first is that borrowers will default, leaving the lender with a foreclosed property. In the present market, this risk is indeed higher than it has been historically, both because default rates are up, and because the property the investor is left with may have fallen in value. But mortgage lenders also face an additional risk -- that borrowers will refinance and pay off their existing mortgages if rates fall in the future, leaving the investor with a bunch of cash (instead of a profitable investment) just when he least desires it - when interest rates are low. This prepayment risk would be far lower for newly refinanced mortgages than it has been historically, since there is virtually no risk that future mortgage rates will fall below 4% and lead borrowers to refinance these mortgages.

To reduce the transaction costs of this plan, the government could automatically refinance the 30 million Fannie Mae and Freddie Mac mortgages already on its books -- it already owns the risk on these mortgages, and such automatic refinancing would leave banks available to deal with an influx of non-GSE applications.

Encouraging loan modifications

Even a substantial reduction in mortgage rates would leave many homeowners headed for foreclosure, which is where the second prong of the proposal comes in. The second piece of the Columbia proposal, published by Mayer, Tomasz Piskorski (of Columbia Business School) and Edward Morrison (of Columbia Law School), addresses the thorny issue of loan modifications. A growing number of policy makers see such modifications, or "workouts," as a promising way to stem foreclosures. House values have fallen, and foreclosed houses typically sell for less than their market value. So although many borrowers cannot afford their original mortgage, they often can afford a mortgage worth more than the house would net at a foreclosure auction. In such cases, loan modifications benefit everyone involved. The homeowners get to stay in their homes with an affordable mortgage, and lenders receive more than they would under foreclosure.

So why aren't lenders and borrowers arranging loan modifications on their own? To some extent they are. When a homeowner takes out a loan, the resulting mortgage is an asset for the lender - the promise of a future cash flow. Banks keep some mortgages on their own books - they are known as "portfolio loans." In such cases, it is relatively easy for banks to contact delinquent borrowers and arrange a mutually favorable workout. In other cases, however, banks securitize their mortgages, selling the cash flow (and the risk that the borrower will default) to independent investors. Although the bank may still service the mortgage (billing the borrower, routing cash flows, etc.) they do not receive the mortgage payments.

Although securitized loans represent only about 15 percent of the total mortgage market, they account for over half of all foreclosure starts in 2008. Yet securitized mortgages are also less likely to be modified, even when modification is in the interests of both the borrower and the lender, for two reasons. First, the pooling and servicing agreements (PSAs) governing securitized loans are often vague about when loans can be modified - and they sometimes specifically prohibit workouts - so servicers are hesitant to modify loans when doing so may leave them open to lawsuits from investors. Second, even aside from issues of legality, servicers frequently do not want to arrange loan modifications. When a house goes into foreclosure, servicers collect fees that often exceed what they would receive if the loan were successfully modified. Since servicers must initiate loan modifications, many workouts that could benefit both the borrower and the investor are left on the table.

The Mayer, Morrison, and Piskorski plan addresses both of the impediments to securitized loan modifications. The legislation they propose would indemnify servicers who make good faith efforts to successfully modify loans from investor lawsuits. Secondly, they propose using TARP funds to pay mortgage servicers a monthly fee mirroring the cash flows from borrowers to investors. This would effectively align servicers' incentives with those of the lenders they represent, thereby motivating them to carry out loan modifications that benefit both investors and borrowers. They estimate this plan would prevent one million foreclosures over the next three years, at a cost of not more than $11 billion.

Both of these plans are likely to be hashed up and repackaged before making it into a final stimulus bill, but there are good reasons to consider each of them. They might also serve as a welcome opportunity for bipartisan cooperation. Republicans have been the strongest advocates for the Columbia plan so far, but Democrats have recently signaled their willingness to compromise, particularly on issues targeting the housing market. Let's hope they do.