Sign up to receive our monthly newsletter!

Analysis of the Final 2018 Public HMDA Data

Sep 11, 2019 by Leonard Ryan

The Federal Financial Institutions Examination Council (FFIEC) on August 30th announced the availability of 2018 Home Mortgage Disclosure Act (HMDA) aggregate data.

The public disclosure data was made available in April of this year at the institution level. This was the first year for both the expanded HMDA data requirements announced in 2015 plus the S.2155 Partial Exemption passed by Congress and incorporated into HMDA regulations last summer.  

New data points and mid-year rule changes created significant challenges for lenders, their vendors, and regulators. Implementation and testing of systems and processes were continuous throughout 2018 and well into 2019. Many questions went unanswered, resulting in confusion that became more apparent with the data submissions. It is anticipated that most issues have been addressed and that 2019 data submitted in March 2020 will not repeat these issues and the aggregate data will be released much earlier next year.

What is happening in the industry?

Steady industry consolidation is continuing to be reported with about three percent fewer institutions reporting in 2018 (5,683) vs. 2017 (5,852) despite having the same reporting criteria for eligibility. Originated loans additionally decreased by 12.6% largely due to a 23.1% decrease in refinance lending.

Analysis of the individual submissions indicates that 349 lenders submitted HMDA data for the first time. Of this, 205 of them appear to be new institutions. This means that there were about 360 lenders that exited HMDA reporting (6.3%) through attrition, dropping below the reporting threshold or as a result of merger/acquisition activity.

1,141 lenders increased their reporting by 25% or more while 1,046 reported applications dropped by 25% or more from 2017 to 2018. Of the top 25 reporting institutions, three more than doubled their submitted data in 2018 (Bank of America, Huntington National Bank and CBNA).  

The Effects of the S.2155 Partial Exemption.

The passage of S.2155 created a partial exemption in the number of data fields that depository institutions (banks and credit unions) were required to report. Year 2018 data was the first year covered under the partial exemption. In all, 39.6% of HMDA reporters took advantage of the reduced reporting, which represented just 2.8% of the total records submitted. This falls in line with a report issued last year by QuestSoft which indicated that there would be a small number of applications affected, but that many loans affected would fall in rural areas. Additional research will need to be completed to determine the resulting outcome.

However, it appears that institutions claiming the partial exemption are being asked to provide the full and complete HMDA data set to their examiners. It appears that the only benefit of the partial exemption is to reduce exposure through the public file.

The Story for Minority Borrowers

There was good news for low/moderate-income (LMI) borrowers in the 2018 HMDA data. First lien home purchase loans for 1-4 family, site-built, owner-occupied properties to LMI borrowers increased from 26.3% to 28.1% with refinance loans increasing from 22.9% to 30.0%.

In contrast, first lien home purchase loans for 1-4 family, site-built, owner-occupied properties made to Black and Hispanic-white borrowers rose only slightly, but denial rates continued to be higher for these two classes of applicant.

HMDA By the Median and Average Numbers

Finally, lending by various measurements changed:

  • FHA and VA Lending share of first lien home purchase loans for 1-4 family, owner-occupied properties decreased slightly from 22.0% in 2017 to 19.3% in 2018, with refinance loans decreasing from 13.0% in 2017 to 12.8% in 2018.  
  • Non-depository lending for 1-4 family, owner-occupied home purchase loans accounted for 57.2% of volume, up 1%.
  • High-Cost HOEPA lending accounted for just 6,681 originated loans or 0.04%, showing that these loans continue to be almost non-existent in the market.
  • Only 57,000 applications were for reverse mortgages (0.4% of lending) with 424,000 applications for commercial purpose.
  • Only 10% of applications reporting ethnicity and race included a disaggregated category.  
  • The most commonly reported disaggregate groups selected were Mexican (3.0%) and Other Hispanic (1.5%).
  • Applicant demographics collected based on Visual Observation account for 6.3% of applications for race and ethnicity and 6.5% for gender.

For more information about the 2018 HMDA data, visit