How to Understand the Next Mortgage Crisis

There are some troubling new developments in mortgages. Again. The quick and dirty story is that increasingly, people who fall behind on mortgage payments aren't catching up. Here's what that means, why it could change the way we think about how foreclosures will evolve over the next year and what we should be doing about it:

From The Wall Street Journal:

Homeowners who fall behind on their mortgage payments have become much less likely to catch up again, a new study shows.

The report from Fitch Ratings Ltd., a credit-rating firm, focuses on a plunge in the "cure rate" for mortgages that were packaged into securities. The study excludes loans guaranteed by government-backed agencies as well as those that weren't bundled into securities. The cure rate is the portion of delinquent loans that return to current payment status each month.

Fitch found that the cure rate for prime loans dropped to 6.6% as of July from an average of 45% for the years 2000 through 2006. For so-called Alt-A loans -- a category between prime and subprime that typically involves borrowers who don't fully document their income or assets -- the cure rate has fallen to 4.3% from 30.2%. In the subprime category, the rate has declined to 5.3% from 19.4%.

"The cure rates have really collapsed," said Roelof Slump, a managing director at Fitch.

Because borrowers are less willing or able to catch up on payments, foreclosures are likely to remain a big problem. Barclays Capital projects the number of foreclosed homes for sale will peak at 1.15 million in mid-2010, up from an estimated 688,000 as of July 1.

Let's put that in a chart:


The cure rate is the percent chance that if a loan goes delinquent it will leave delinquency without needing to go to court or change the terms of the loan. No intervention was needed by the banks or the courts - the borrower in question starting paying his bills again in full. The higher, the more likely it is that someone will find himself back on track. If this was 100%, everyone who missed a payment would automatically make their next one. If this was 0%, everyone who missed a payment would end up foreclosing.

A few things to note with this new data when it comes to mortgage modification.

1) There used to be an obvious hierarchy in loan quality. People with prime loans would recover from whatever shocks lead them to miss a payment more often than Alt-As (see the graph above). People with subprime loans would be half as likely; they were poorer credit risks in and of themselves. Now with unemployment approaching 10% mortgage problems are going to be less about the credit risk of specific individuals and more about the general economic environment - the rates are now indistinguishable from each other.

2) This is the mechanism we watch to see upcoming foreclosure rates. Servicers aren't modifying a lot of loans. There is a lot of talk about more and more sophisticated servicers coming online in the next couple of months which will lead to a jump in the number of loans serviced. However it seems that the rate of foreclosures and troubled mortgages is going to increase faster, outpacing the servicers. More drastic action might be necessary.

3) 10% of prime loans are in delinquency; normally we could count on half those loans being able to find their way out of trouble on their own. Now just a handful, 1 in 20, will. This will greatly impact middle-class consumers, as much of their wealth is in housing, and underwater and collapsing housing stock impacts labor mobility. As John Robb says, the collapse of the middle-class consumer is just beginning.

4) There is a very influential Boston Federal Reserve Study - Why Don't Lenders Renegotiate More Home Mortgages? Redefaults, Self-Cures, and Securitization (pdf) which finds that servicers don't renegotiate home mortgages for rational reasons. What are they? Redefaults - that you'll modify the mortgage and it'll fail later on anyway, and Self-Cures, the cure ratio we are talking about here. In their study, as Morgan Housel points out, they estimate a time-appropriate 30% self-cure rate for their study of the 2005-2007 time frame.

Now, it appears that this cure rate has fallen over 80%, making loan modification much more rational. Yet we don't see it appearing yet in the aggregate data. So either the servicers are rent-seeking harder or the loans are riskier in re-default risk. Either way doesn't look promising for the future of mortgage modification. We may need to supplement the approach with aggressive bankruptcy courts.