What I learned from Ashlyn Nelson about Liar Loans and Foreclosures

I’ve been writing about the new public ethics of home ownership and foreclosures, including some questions about the defaults among what seem to be higher income mortgage holders.

USC’s Lusk Center sponsored a talk by the extremely gifted young scholar Ashlyn Nelson at Indiana University. She was kind enough to do a recent presentation on two new papers:

Liar’s Loan? Effects of Origination Channel and Information Falsification on Mortgage Delinquency Wei Jiang , Ashlyn Aiko Nelson and Edward Vytlacil


Securitization and Loan Performance: A Contrast of Ex Ante and Ex Post Relations in the Mortgage Market
Wei Jiang , Ashlyn Aiko Nelson and Edward Vytlacil

And I highly recommend you take a look at both manuscripts. First, they are dealing with really significant public policy issues, and second, both studies use a very clever set of research designs and methodologies to get at the complex issue of moral hazard in mortgage lending.

My takeaways are summarized as:

1. Bank-originated, bank-held loans have a much higher survival rate and, thus, lower risk of foreclosures and delinquency.

2. With low-document bank-originated loans, the researchers could not find any observable “red-flags” for poor loan performance at the time of origination. However, among broker-originated low-doc loans, nearly 75 percent of the performance differences came down to issues that were observable at the time the loans were originated. IOW, there was moral hazard for brokers who turned a blind eye to red flags and passed the loans.

3. When the authors first conducted their analysis, they found an association between higher incomes and foreclosures–the relationship that has had me tied into knots for months.

4. Delving into these “liar loans”–the low documentation loans–they use a neat set of reasonings to gather an estimate of potential income exaggeration. Now, this is an estimate of real income for individuals gathered from neighborhood income weighted upwards for the comparative affluence of homeowners. However, it’s a nice idea, and using this approach, they are able to rough out that, on average–on average!!–liars exaggerated their incomes by close to $1,830 a month. That’s a lot. I would love to see the spread on the exaggerations.

5. This means essentially that individual liars were gaming the the debt-to-income ratio in order to get more house than they could really afford. When they got rid of the fudging, the relationship between income and delinquency returned to expectations.

I had to leave before I got to their answers for whether secondary buyers were able to cherrypick better performing loans from banks. So I shall have to read the manuscripts to get at those answers, but from what I could tell from the introduction, if banks were engaging in moral hazard, they weren’t particularly good at it as they got the short end of the stick on a lot of loans.

Ah! This is the sort of scholarship that makes me want to make my own work better!

Add to: Facebook | Digg | Del.icio.us | Stumbleupon | Reddit | Blinklist | Twitter | Technorati | Yahoo Buzz | Newsvine