Thursday, February 18, 2016

Housing: Part 119 - Review of Household Debt and Credit Report

Yesterday, I referenced the potential influence of the bankruptcy reforms in 2005 on the apex of the housing boom, including possible evidence for the more intense late-boom price swings in Arizona, Florida, and Nevada.  Today, I will look at a few more graphs.  I have looked at these before, but I think there is some new significance related to recent additions to my overall narrative.

First is the graph of mortgage originations, by FICO score.  As my version of events is shaping up, I think we can see a couple of distinct events.  Now, I think we can probably pin down the sharp decline at the end of 2003 on the accounting scandals at the GSE's and the subsequent pressure on them to play it safe.  Note that the decline in originations at that point is weighted, oddly, toward the higher FICO scores.  These are the borrowers who would have been using GSE loans.

In the following graph, we can see that FICO scores were actually rising until 2003.  They fell until early 2007.  But, we can see here that, (1) typical FICO scores of borrowers were still within recent norms, and (2) the decline did not come from a surge of low FICO score borrowers, but from the decline in high FICO score borrowers, probably due to the decline in GSE activity.

Also, it is likely that as short term rates rose from 2004 to early 2006, high FICO score borrowers were engaging in fewer opportunistic refinancings.

In the last graph, we see delinquency levels, by state.  I have scribbled in the point in time where funding for subprime loans had completely dried up and AAA subprime securities were trading at a discount.  By this time, there had been distress in the subprime funding industry for a year or more.

I am currently reviewing Gary Gorton's work on the crisis.  He points out that subprime borrowers and lenders depended on the ability to refinance.  A lot of observers have discussed the problem of teaser rates and rate resets.  But after the run on shadow banks, there were basically no subprime loans available.  Many of those defaults after mid-2007 were related to negative equity.  But, even if those households had been willing or able to refinance, there simply was nobody to borrow from.  Should teaser rate resets have become a problem?  Maybe.  But, really, there is no way of knowing, because the liquidity crisis removed the option of refinancing, even for households who might have reasonably expected to be able to refinance.


  1. Tyler Cowen links to a review of Tim Geithner's Coursera course on the Financial Crisis:
    Would you care to comment?

    1. Intersesting. The commentary at the link isn't necessarily wrong. There is a lot of "mood affiliation". Loaded statements like "All the bad actors on Wall Street got bailed out but responsible home buyers whose only mistake was buying in 2005, 2006, 2007 or 2008 didn’t." verge on being factually incorrect, in the service of tribal signals. But, a bunch of low down payment loans do make the system less robust. It's not wrong. But, if you cut a hemophiliac, and he bleeds to death, did you kill him or did the hemophilia? There's a lot of causal density, which is probably what helps everyone defend the most convenient version of the story.

      It looks like Geithner is downplaying the moral hazard complaint and arguing that stimulus should have happened sooner. I'd say that puts him well ahead of most. He has the causation backwards on credit expansion leading to home price appreciation, but everyone does, so I don't know that I can fault him much for that - at least until he gets a copy of my book. :-)