QM Impact on the Mortgage Market

The American Banker has a story (paywalled) about the impact of the QM rule on subprime lending. Subprime loans are unlikely to qualify for the full QM safe harbor because they are typically priced at more than 150 bps above prime. This means that subprime borrowers now have a possible foreclosure defense if they can show that the lender failed to properly account for their ability to repay. The result is to increase the risk lenders incur with subprime loans. But how much does this matter?

The American Banker story contains an estimate that only 5% of today's mortgage volume would fall outside of QM. Today, however, is an ultra conservative mortgage market. We can get a rough sense of what the impact of QM would have been in the past by looking at the market share of subprime and Alt-A loans.  

The CFPB's rulemaking has an estimate for the 1997-2003 period roughly 70% of originations would have been QM, with 22% non-QM and 8% consisting of loans that can no longer legally be made.  (The CFPB's rulemaking also estimates a very minimal impact on mortgage pricing, p.582 and following in the rulemaking.)

My own back-of-the-envelope estimate roughly in line with that.  Let's say that all of subprime originations and Alt-A originations would not have qualified for QM.  Some prime loans wouldn't either, but they aren't easily broken out of the data. Below is a chart showing the market share of subprime and of subprime and Alt-A originations by dollar amount. If my estimate is correct, then this is approximately the share of mortgages that would have been covered by QM in the past (assuming that nothing else had changed). Of course, had QM existed, it's possible a lot would have been different, and if it really will chill non-QM originations, it suggests we might well have avoided the housing bubble.

A few more comments on the QM rulemaking:

1. Some commentators have taken the mortgage industry's sigh of relief over the QM rulemaking as a sign that it is a weak or bad rulemaking. I think that's poorly considered. I haven't seen any consumer groups expressing disappointment. It might not be exactly what every consumer group hoped for, but a win is a win. The mortgage industry's relief is explained by two factors.  First, QM isn't going to have much effect on current business because the mortgage industry isn't doing much in the way of subprime anymore. And that's fine. QM is meant to preclude the type of poorly underwritten loans made during the housing bubble, not create a right to litigate underwriting of every mortgage loan. And second, industry reactions are in part determined by expectations. The financial services industry beat itself into a frenzy between 2009 and present worrying about the CFPB as national credit commissar. I think many in the industry started to believe their own rhetoric. Perhaps the QM rulemaking will encourage some more realistic appraisals of the agency by the financial services industry.
2. Some commentators have been baffled by the lack of down payment or FICO requirements in QM. I'm not sure why. Down payments don't show ability to make monthly payments. They're relevant from a creditor perspective in terms of ensuring borrower skin-in-the-game, but they don't seem to have much to do with ability to repay. That's why downpayments are likely going to be part of the interagency Qualified Residential Mortgage (QRM) investor protection rulemaking, but aren't in QM. And mandating FICO scores are just kind of silly for ability to repay. Would QM really require a 620 or 660 FICO? What does FICO add that income and asset verification and back-end DTI doesn't? 
3. My original QM post (now corrected) missed the sunset provision regarding the GSEs. This is important. At the sooner of 2021 or the end of conservatorship, GSE purchase eligibility ceases to satisfy parts of the QM definition. In other words, GSE underwriting is only being used as a partial proxy for ability to repay while the GSEs are under tight federal control. 
4. Finally, I missed two more behavioral economics tidbits in the rulemaking, both on page 557. This is all I can find in 804 pages. 

"Consumers may underestimate the true costs of homeownership or be overly optimistic about their own future (or even current) financial condition."

"Further, consumers may not understand or may underestimate the costs they will incur in the event of default, such as the loss of the borrower’s own home, costs of relocation, and the borrower’s loss of future credit, employment and other opportunities for which credit reports or credit scores weigh in the decision."

I'm not sure how much behavioral economics is necessary to get to these insights, but there it is.  So much for the behavioral bogeyman.