Data show surprisingly little impact of new mortgage rules
By Jim Parrott and Ellen Seidman and Laurie Goodman Bing Bai :: August 21st, 2014
The Qualified Mortgage, or QM rule, which was designed to protect borrowers from acquiring loans they cannot repay, was predicted by many to reduce the availability of mortgages. The law has been in effect since January of this year, so any change it’s made to lending patterns should be evident by now. After combing the data for the past several months, and despite recent headlines, the Housing Finance Policy Center team has seen surprisingly little impact on the origination numbers.
The QM Rule in brief. Since January 2014, lenders have been responsible for determining, before they make a loan, if the borrower has the ability to repay the mortgage. The government will presume the borrower has the ability to repay and protect the lender from any future borrower lawsuits if the mortgage meets the guidelines of the Qualified Mortgage or QM rule.
Headlines don’t always get it right. After seven months of the rule’s operation, some have claimed that a recent survey proved the predictions right: Banks Making Fewer Mortgages Because of CFPB Rules, Fed Says. But the headlines didn’t get it quite right. The truth is, the Federal Reserve’s Senior Loan Officer Survey actually tells the same story we’re seeing:
- There has been almost no impact in the government sponsored enterprise (GSE—Fannie Mae and Freddie Mac) or government agency (Ginnie Mae) market; and
- There has been only minimal impact on the loans banks are holding in portfolio.
And, since the GSEs and Ginnie Mae together account for approximately 80 percent of all originations, the muted impact on their loans far outweighs the slightly stronger impact we found in bank portfolios.
How would we see an impact: Four elements of the QM rule could be expected to have a significant impact on mortgage availability:
- The disqualification of loans that are interest-only (IO) loans and or have a prepayment penalty (PP) might reduce the number of loans made with those features;
- The limitation of the back-end debt to income ratio (DTI) to 43 percent might reduce the percentage of loans to borrowers with DTIs in excess of 43 percent;
- The three percent limit on points and fees might limit lender interest in making small loans;
- The requirement that an adjustable rate mortgage (ARM) be underwritten to the maximum interest rate that could be charged during the loan’s first five years might reduce the ARM share.
Mitigating the impact of these factors is the part of the QM rule known as “the patch.” This allows the GSEs and government agencies such as the Federal Housing Administration (FHA) to operate under their own standards for seven years or, in the case of the GSEs, when they exit conservatorship, whichever is sooner. Under their own standards, both the GSEs and the FHA eliminated the DTI restriction but retained the other requirements listed above.
We don’t really see these impacts on access to credit: Here’s what we see in the data:
- IO and PP loans: Were fewer of these loans made? No. Very few IO and essentially no PP loans were made before QM went into effect, and that is still the case. In the GSE and agency markets, the number of IOs was essentially zero before and after; in the non-agency market it has held steady in the 2 to 3.5 percent range.
- 43 percent DTI limit: Are there fewer loans with DTIs over 43 percent? No. Theshare of loans with DTIs over 43 percent, while different in each channel, has remained relatively steady at approximately 10 percent in bank portfolios, 17 percent for the GSEs and 35 percent for Ginnie Mae. However, in recent months, the GSE share of higher DTI loans has declined slightly. This is worth keeping an eye on, as the expectation has been that the patch will encourage continued lending to lower-income borrowers through the GSEs as well as through FHA.
- Points and fees cap: Is there less interest in small loans?As shown in Figure 1, we do see a drop in the percentage of loans under $100,000 being bought by the GSEs, or securitized by Ginnie Mae. However, the decline has been relatively small, and variable, with the March 2014 percentage of smaller loans higher than any month since January 2013. While the percentage of bank portfolio loans under $100,000 appears to have declined substantially in November 2013, and continued to fall through March 2014, our data is sparse –there was only a 400-loan decline in the number of new small loans in the bank portfolios between November 2013 and March 2014.
4. ARM underwriting rule: Has there been a drop in the ARM share?Traditionally, as interest rates increase, so does the ARM share, and indeed, as shown in Figure 2, last summer’s spike in interest rates was reflected in a very large increase in the ARM share of bank portfolio loans starting in July 2013. The ARM share in bank portfolios has declined marginally since the January 2014 peak, although it is still significantly higher than during early 2013, when interest rates were lower than they are today. The ARM share at both the GSEs and Ginnie Mae is far smaller than in bank portfolios and it has increased slightly notwithstanding the QM rules.
Why so little impact? We think there are four likely reasons why the QM impact has been muted.
- There may be little or no effect because credit standards were already tight before QM went into effect. A corollary is that the demand for non-QM loans may have declined following the financial crisis, in part because of greater house-price affordability.
- About 80 percent of lending is being done through the GSEs or Ginnie Mae, organizations covered by the patch. Some loans that might have ended up in bank portfolios, but are close to the 43 percent DTI limit, are being sent to the agencies.
- Some institutions were slow in fully implementing QM or systems to track QM loans, delaying its ultimate impact. In addition, some entities may have had preexisting commitments that had not yet closed.
- The primary effect of QM is on institutions not covered by our data. We have complete agency data, and non-agency data from the largest servicers and a smattering of others. It is possible that small banks and credit unions are being much more stringent when approving portfolio loans, which would not show up in the data we have access to.
While there remains a great deal of uncertainty over the ultimate impact of the QM rule, we do not yet see significant changes in the market. And while we will continue to track the issue carefully, better answers will likely have to await release of the 2014 Home Mortgage Disclosure Act data in the fall of 2015, as that information will include mortgage application, origination and denial activity from virtually all lenders.