How to avoid another housing crisis (maybe)
Matthew C Klein
No matter how you count it — consumer spending, employment, fertility, divorce — the US housing bust was extremely costly. While cycles of greed and fear are inevitable features of human nature, things did not have to happen this way.
Smarter government responses in the initial phase of the downturn would have been very helpful. Perhaps in a repeat performance, policymakers will do better, although we wouldn’t want to rely solely on their learning capabilities when there are reasonable preventive measures available as well. Continuous workout mortgages and shared responsibility mortgages would be useful financial innovations, for example.
A new paper unveiled at today’s semiannual Brookings Papers on Economic Activity by Janice Eberly and Arvind Krishnamurthy provides some additional theoretical insight into these questions. The main innovation is its attempt to distinguish between periods when households face temporary hits to their incomes and long-term declines in house prices that leave borrowers with good reason to engage in “strategic default.”
Most of the paper focuses on what can be done to mitigate a crisis that has already started, although there is a short bit at the end discussing the best ways to alter future mortgages to make them less dangerous for borrowers and their neighbors to prevent a repetition of the 2002-2010 cycle.
Staying current on mortgage payments can be tricky when hit with an unexpectedly large expense or a temporary loss of income (perhaps due to unemployment in a bad economy).
Lenders certainly have the option to foreclose and may even profit from doing so if the borrower’s home is worth more than the principal owed on the mortgage. But foreclosures create nasty spillovers for everyone else by blighting the neighborhood and depressing home values. That in turn can make the downturn in spending and employment worse, increasing the probability that other borrowers may be forced to default.
This suggests it is in the interests of lenders with exposure to more than one mortgage in the community to come up with a solution that allows borrowers to stay in their homes as long as possible. But how?
Eberly’s and Krishnamurthy’s insight is that borrowers who have no incentive to default (perhaps because they have high credit scores, or because they have positive equity in their homes) are agnostic about how they reduce their monthly payments. That gives lenders leeway to pursue several different options that affect the net present value of their mortgages in different ways.
The best option for lenders, at least in this theoretical model, is payment deferral. Borrowers would cut their monthly expenses by a large amount upfront in exchange for a promise to make up the difference later. There are different ways of structuring this but the simplest is to refinance existing mortgages into new ones with much lower interest rates and much longer maturities.
Principal reductions in this situation are unattractive as a first resort, at least in the theoretical model, because they remove the option of writing down the outstanding balance in the future should house prices fall significantly:
Principal reduction can be helpful, but is a less efficient use of government resources, since it back-loads payments to households who cannot borrow against these future resources to support consumption today, and also because it is most helpful in reducing strategic default, rather than payment-distress-induced default.
Defaults resulting from payment distress have a greater negative impact on home prices, since distressed borrowers carry their distress into the rental market and reduce housing demand more than default resulting from strategic considerations.
Of course, lenders may be reluctant to make even this type of modification during a downturn, and for good reason. They may doubt the borrower will be able to afford the lower payments if he remains unemployed, and they should also be concerned that the mortgaged home’s value will decline enough that the borrower concludes it is in his interests to default.
The authors recommend that the government get around this problem by paying lenders to modify their loans or by offering to refinance the mortgages directly, presumably incurring some credit risk (and upside) in the process. They do not model it, but they suspect that having the government offer to pay households directly would be even more effective, although we suspect that this would effectively subsidise lenders and discourage them from modifying any loans at all.
While this is theoretically interesting, most people who lost their home in the recent downturn did not do so because of temporary setbacks that they could have waited out. They were not illiquid but insolvent. Even now, millions remain underwater on their mortgages and only keep paying because they would rather not deal with the costs of moving and reduced access to credit.
Some borrowed so much that they could only hope to repay their debts if house prices continued to rise at double-digit rates, lacking income barely sufficient to cover their teaser payments. They would not be helped by a maturity extension scheme, no matter how generous, unless we are to strain all credulity and believe that the downturn in house prices was just a temporary blip.
But many others were given in a once-in-a-lifetime chance to buy a house because of the absurd relaxation of credit standards. These borrowers had little reason to continue making payments on their one asset now that it was worth far less than their mortgage. Only a reduction in principal sufficient to restore their net worth above zero would keep them from walking away.
(One reason the Home Affordable Modification Programme was such a failure was that it allowed banks to take advantage of people who could afford to make payments but would have been better off defaulting immediately.)
Eberly and Krishnamurthy briefly discuss this situation. They think the challenge in designing the optimal principal reduction scheme is separating the Ponzi borrowers, who could never service their debts out of their income, from those who rationally want to avoid pouring their money down the drain when they have negative equity. Their solution is to get the government to subsidise a standardised modification contract for all borrowers.
Their cursory treatment of this issue is regrettable since it leads us to suspect that they got the causality of the crisis backwards. Research by Atif Mian and Amir Sufi shows that income growth in the bubbly period, weak as it was, would have been far lower had it not been for the consumption funded out of home equity withdrawals.
(Former Fed Chairman Alan Greenspan also recognised the importance home equity played in the 2000s, for all the good that did us.)
Borrowing against rising home values was only possible, however, because reckless lenders were finding all sorts of people who should never have been taking out mortgages and throwing money at them to buy housing.
One result was a construction boom, but the credit expansion far exceeded the increase in housing supply. Sometime in 2006, although there was regional variation, the mortgage originators ran out of warm bodies. At first house prices flattened, then declined as the marginal borrowers began defaulting and dumping supply onto the market.
That was what drove the subsequent changes in employment and income, not some random shock from the heavens, much less one that prevented otherwise solvent borrowers from extracting home equity to sustain current consumption.
So while we appreciate the insight that the best thing to do for a solvent borrower who loses his job when home prices are stable may be to refinance him into a new mortgage with lower payments and a longer maturity, the overall analysis may not be that relevant for the boom and bust cycle we actually experienced.
Eberly’s and Krishnamurthy’s proposed financial innovation is to suggest that fixed-rate mortgages ought to contain the option to refinance into floating-rate loans even when the borrower is underwater. Since short-term interest rates tend to move up and down with the health of the economy, this would allow borrowers to sharply curtail their payments during a downturn and then make up some of the difference later as rates rise.
Presumably lenders would only agree to such a deal by charging much higher interest rates upfront given the downside risks. Giving lenders upside ought to lower this risk premium while also moderate the bubbly phase of the housing cycle. We thus wonder whether it might be better to automatically index the principal (and by extension, the monthly payments) to the value of comparable homes in the neighborhood.
Indexing reduces the need to trust that lenders will honor their generous refinancing options if the payments are always floating — a legitimate concern given their track record with HAMP. Plus, local prices would not be able to get too far out of line with local incomes since the higher payments would push people to sell, creating an automatic macroeconomic stabiliser.