by Calculated Risk on 6/12/2008 11:18:00 PM

Here is a new research paper with some important conclusions about the percentage of foreclosures among homeowners with negative equity. From Christopher L. Foote, Kristopher Gerardi, and Paul S. Willen at the Boston Fed: Negative Equity and Foreclosure: Theory and Evidence

As a consequence of the recent nationwide fall in house prices, many American families owe more on their home mortgages than their houses are worth—a situation known as “negative equity.” The effect of negative equity on the national foreclosure rate is of obvious interest to policymakers, but this effect is difficult to study with datasets that are commonly used in housing research. In this paper, we exploit unique data from the Massachusetts housing market to make three points. First, during a specific historical episode involving a downturn in housing prices—Massachusetts during the early 1990s—less than 10 percent of a group of homeowners likely to have had negative equity eventually defaulted on their mortgages. Thus, current fears that a large majority of today’s homeowners in negative equity positions will soon “walk away” from their mortgages are probably exaggerated. Second, we show that this failure to default en masse is entirely consistent with economic theory.
...
A foreclosure requires both negative equity and a household-level cash-flow problem that makes the monthly mortgage payment unaffordable to the borrower. Cash-flow problems without widespread negative equity do not cause foreclosure waves. Even if borrowers are having trouble making payments, they will always prefer to sell their homes rather than default, as long as equity in their homes is positive so they can pay off their outstanding mortgage balances with the proceeds of the sales. Similarly, widespread negative equity will not result in a foreclosure boom in the absence of cash-flow problems. Borrowers with negative equity and a stable stream of income will, in most cases, prefer to continue making mortgage payments. Thus, we argue that negative equity does play a key role in the prevalence of foreclosures, but not because (as is commonly assumed) it is optimal for borrowers with negative equity to walk away from affordable mortgages.
emphasis added

The authors present a model to explain why homeowners with negative equity, but sufficient cash flow, will not walk away. See section 3: The basic economics of default from the borrower’s perspective

I think this model is helpful for understanding the behavior of homeowners with minimal negative equity, but may be flawed for a simple reason: the probabilities in the two state model are what the homeowner believes will happen, and homeowners deep in negative equity will assign probabilities of zero to the good outcome and one to the bad outcome.

Here are the equations as presented by the authors (see paper for description):
V₁^H = rent₁ + 1 / (1+r) * [3/4 P₂^G + 1/4 P₂^B]
V₁^M = mpay₁ + 1 / (1+r) * [3/4 M₂ + 1/4 P₂^B]
But notice what happens when we make the good outcome zero for deep underwater homeowners (instead of 3/4) and the bad outcome 1 instead of 1/4 (and adding stigma term). The choice simplifies to the obvious:

Value = (rent₁ - mpay₁) + Stigma_i
Where Stigma includes "moving costs, default penalties that take the form of limited future access to credit markets, sentimental attachment to the home, or even the presence of moral qualms associated with defaulting on one’s debts".

This is really the problem: deep underwater homeowners who perceive the probabilities of a negative outcome as 1 (and are probably mostly correct), will walk away from their homes unless Stigma is greater than (mpay - rent). And Stigma for many of these homeowners really depends on if it becomes socially acceptable for middle class Americans to walk away (ruthless default).

Finally, there may be problems when comparing to the Boston housing bust of the early '90s - although prices did decline about 30% from the peak in real terms (according to Case-Shiller), lending standards were tighter in the late '80s compared to the recent bubble, and few homeowners bought at the peak with no or negative equity (like during the current boom). Also, the current bubble was much larger than the late '80s bubble in Boston, and some areas in the U.S. will probably see real price declines in excess of 40% (maybe even 50% or more), and these homeowners will be deeply underwater.

This is an interesting paper. I believe it is like that a majority of homeowners with negative equity will not walk away from their homes. But I believe we need to know the number of homeowners deeply underwater, and try to understand their probable behavior.