Friday, March 28, 2008

Worth rereading: Deng, Quigley and Van Order from 13 years ago

Mark Thoma tries to unravel the mystery of why house prices matter more than rate resets. The paper by D, Q and VO is Mortgage Default and Low Downpayment Loans: The Costs of Public Subsidy, and I think it helps answer the question. The abstract:

This paper presents a unified model of the default and prepayment behavior of homeowners in a proportional hazard framework. The model uses the option-based approach to analyze default and prepayment and considers these two interdependent hazards as competing risks. The results indicate the sensitivity of default to the initial loan-to-value ratio of the loan and the course of housing equity. The latter is a measure of the extent to which the default option is in the money. The results also indicate the importance of trigger events, namely unemployment and divorce, in affecting prepayment and default behavior. The empirical results are used to analyze the costs of a current policy proposal -- stimulating homeownership by offering low downpayment loans. We simulate default probabilities and costs on zero-downpayment loans and compare them to conventional loans with conventional underwriting standards. The results indicate that if zero-downpayment loans were priced as if they were mortgages with ten percent downpayments, then the additional program costs would be two to four percent of funds made available -- when housing prices increase steadily. If housing prices remained constant, the costs of the program would be much larger indeed. Our estimates suggest that additional program costs could be between $74,000 and $87,000 per million dollars of lending. If the expected losses from such a program were not priced at all, the losses from default alone could exceed ten percent of the funds made available for loans.


When households have equity in their house, they don't default. This is why the unprecedented drop in nominal house prices nationally is so calamitous. At the same time, households don't usually default unless they face a trigger event, such as unemployment, divorce and illness. A large rate reset may also constitute such a trigger event, but as three Boston Fed economists show, these are not all that common. Trigger events, moroever, are not sufficient conditions for default, and negative equity is a necessary condition for default.

2 comments:

Anonymous said...

I think your point is that economists have known about the importance of equity vs. trigger events for at least 13 years. But I'd like to point out that economists have known this for at least 15 years. See http://archive.gao.gov/d49t13/149941.pdf

GAO's report on the VA mortgage program. Two years before the Deng, Quigley, Van Order paper that you cite, GAO was running competing risk models of an actual zero down mortgage program (as opposed to their extrapolation of data with non-zero LTVs to an LTV world). We found a 60 bp increase in credit cost for each 1% change in annual house price appreciation, found that initial LTV and equity were important determinants of default, and found that a trigger event, unemployment rate, had some explanatory power, but a lot less than equity measures did.

If you just want to push back how long housing economists have known that equity matters, you could go all the way back to von Furstenburg in, what, 1974? Actually, I think Herzog, FHA's actuary, was doing tables of default by LTV back in the 1960's.


BTW - thanks for the shotout on my zero down paper. If I'm enough of a pain ...

Unknown said...

David lindahl says that, if you want to get rich in no matter of time, then real estate investing would be the best way; inexorably when we think of real estate investing; we foresee a residential rental, either a unit or a house in fact, the real estate industry is much bigger than that.
Nevertheless, there are several things that will help you to make money in this particular field. David lindahl scam is one such great deal would be concentrate on different types of properties that are less conventional.