A preliminary look at Home Mortgage Disclosure Act data for 2013 shows applications down 8% from the prior year, but loan denial share up by 13%.
The Federal Financial Institutions Examinations Council will not release the 2013 HMDA data until September. But a mortgage analytics firm has gotten access to what it claims represents about two-thirds of 2013 market volume, simply by asking lenders to send it their data.
Nearly 400 lenders, "including quite a few top lenders," responded to Mortgage TrueView's survey request, which went out to more than 1,000 firms after the March 1 deadline to submit to the FFIEC, said executive vice president Becky Walzak.
The Media, Pa.-based compliance and business intelligence firm likes big data analyses, said Walzak, and the annual HMDA report, with millions of data points from thousands of lenders, was an ideal database to tackle.
Data to aid lender compliance is the why the company did the survey, its first for HMDA. Mortgage TrueView had the 2010, 2011 and 2012 HMDA numbers, but was looking for fresher data.
Walzak points out a new plethora of servicing rules (nine in total) came down the pike as of Jan. 10 of this year, making it a requirement for servicers "to track and monitor data." But, she said, "People just weren't there yet."
With increased vigilance by regulators like the Consumer Finance Protection Bureau, which promises to closely analyze the information, HMDA data may be able to help lenders predict if they are going to get one of those dreaded visits that tie up their conference rooms for weeks at a time.
If there are startling numbers in the HMDA data, "you need to be prepared to explain that to the CFPB," she said.
The CFPB has just announced a proposed rule that would require more mortgage data reporting from lenders. "The HMDA data do not provide adequate information about certain loan features that helped contribute to the mortgage crisis, such as adjustable-rate mortgages and non-amortizing loans," the Bureau noted.
Walzak would not disclose the names of specific lenders that responded to Mortgage TrueView's survey, but she did give an example of one potential regulatory risk, based on an aggregate of the five largest lenders that participated: Denials to African-American women averaged 35% for the five, she said.
Denials aren't the only reason a mortgage application goes unfunded. Apps can be withdrawn or be deemed incomplete, for example. This makes the "mortgage unfunded" category always considerably higher than the denial number. Other findings from the preliminary data include:
— An 8% year-over-year decline in mortgage applications.
— Origination rates dropped from 52% to 44%, a significant fall which would mean there were more unfunded applications than funded ones.
— Refinancing apps dropped from 66% to 59% of total apps, while purchase applications increased from 30% to 35%. That's broadly in line with anecdotal experience.
— Denial rates were up 13% YOY. Denials based on credit history jumped 16%, while denials based on collateral (such as refis) dropped by 19%.
— Denials for white applicants increased 24%, nearly the same as denials to non-whites, up 22%.
— Denials for Hispanics increased by 20%, while denials to non-Hispanics rose nearly as much, by 17%.
— Denials for men and women showed equality between the sexes, but nothing to be bragging about as both categories were up 24%.
One bright spot in the numbers was a big increase in percentage of loans purchased, up 59%. Walzak said that smaller lenders may have had fewer funding options, so they looked to correspondent lenders to buy their production.
An analysis of the industry's year-over-year origination volume was not a part of this report, but Walzak attributes the general downward originations trend to an increase in lender underwriting standards and a general tightening of credit.
Mortgage TrueView touts a free download of its survey results on its website.
Mark Fogarty, Editor at Large at National Mortgage News, brings more than 30 years of experience to his analyses of the mortgage market.