Today's mortgage lending environment continues to be an enigma to the industry, policymakers and researchers alike struggling to understand why credit standards remain relatively tight compared to the "norm." Recent data from the GSEs showing that their 97% LTV programs have yet to catch fire with lenders brings the broader issue of credit standards to focus. The mortgage business is a mosaic of intersecting issues, making this issue particularly complicated for those in search of answers. In seeking answers to this question we first have to ask what does a normal mortgage market look like, what indicators do we have of credit lending policy and processes today relative to the past and what factors could explain what we are seeing?
One of the more difficult aspects of any assessment of mortgage market norms is the fact that the available historical data quickly leads one to the conclusion that since 2000, the lending market can be defined into four distinct episodes:
- the period pre-2004 (pre-boom)
- 2004-2007 (boom)
- 2007-2008 (crisis)
- 2009 and beyond (post-crisis)
The closest reference point we have to a normal credit environment is the period preceding the boom. To gain a better sense of this, consider the early payment default risk and underwriting score for each of these time periods.
The underwriting risk score is a statistically based metric designed to determine the relative level of credit risk reflected by mortgage underwriting standards over time.
This metric considers all important risk attributes of the borrower, property and loan while removing specific effects associated with the economy.
What we see by comparing those two metrics for the years between 2000 and 2014 is a fairly stable and low level of risk until about 2005, at which point scores begin to accelerate so that by 2007 they are more than double that of 2000.
This is a clear indication of the excessive risk layering and riskier credit policies in place during the boom years. The numbers also show the dramatic tightening that took place during the crisis, and its aftermath.
In fact, based on this measure, credit policies as reflected by the underwriting risk score continue to be much more restrictive than the pre-boom years.
In an economic environment that while not robust is nonetheless growing what could be holding lending back?
Two issues might hold the key to answering this question.
The first is the recurring issue of the impact of compliance and regulatory costs; the second is relative mortgage profitability. It is well known that the processes by which loans were originated were deficient in many respects to handle the massive volumes and new riskier products originated during the boom.
One way to estimate the effects of this over time is to look at early payment default rates (EPDs); i.e., loans that have gone 90 days past due or worse within the first year of origination.
Loans that go delinquent that soon after origination tend to have had some underwriting flaw at inception upon inspection of the loan files.
While by no means a perfect measure for looking at underwriting process deficiencies over time, the figure can give us a sense of the relative trend.
It is not surprising that EPD rates began rising in 2007 and peaked in 2008 at a rate about three times that of 2000. Since 2009, EPD rates have fallen to levels well below pre-crisis levels.
Without more analysis, we can only offer some hypotheses about lender process issues. It appears that by the EPD figures, lenders have made progress in improving processes; we also know this by the intense scrutiny on lenders by various regulatory authorities in the years following the crisis.
In its Mortgage Lending Sentiment Survey from 2014, Fannie Mae found that nearly three-quarters of lenders reported increases in mortgage compliance costs with a median increase of 30%.
Lenders also cited concerns over repurchase risk. While FHFA and the GSEs have made good progress in addressing uncertainties over loan repurchase policy, at the very least perceptions of continued uncertainty remain for lenders on this issue.
That lenders may be holding back a bit is not surprising as lessons from behavioral finance studies suggest that individuals tend to weight recent experience more than the past. Applied to the mortgage business and compounded by rising compliance costs, recent experience with heavy credit losses of the crisis could be dampening management enthusiasm for credit expansion in an otherwise expanding, albeit tepid economy.
Compliance costs play into the larger issue of mortgage profitability. Looking over time at what the mortgage business has earned on average and relative to other earning assets presents some picture of how profitability might be factoring into mortgage expansion decisions.
Since 2003, the general trend in mortgage yields has been declining. And relative to other earning assets, mortgage yields are running a little behind those of other bank assets.
With yields coming down from their 2003-2015 levels, further pressure on profits from higher regulatory costs and potential contingent liability from Qualified Mortgage rules and/or repurchase rules would influence the direction of the mortgage business for many firms.
The data suggests mortgage credit is tighter than it was before the boom. Exactly why that may be remains a matter of debate and interpretation. Banks are profit-maximizing enterprises. As such, if prospective mortgage assets are projected to generate attractive profits, banks enter or expand this business.
At least along the cursory measures examined here, a combination of higher compliance costs and a perception of, if not real uncertainty over repurchase risk might be leading causes for holding back a more reasonable expansion of credit consistent with the pre-boom environment.
