As previously reported, in June 2021 the CFPB announced that it would conduct a voluntary review of the significant amendments to the Home Mortgage Disclosure Act (HMDA) rule adopted in October 2015, most of which became effective in January 2018.  The CFPB also announced at the time that it was discontinuing planned HMDA rulemakings, one addressing the data points that must be collected and reported and one addressing the public disclosure of HMDA data. 

The Dodd-Frank Act requires the CFPB to conduct an assessment of each significant rule or order it has adopted under a federal consumer financial law and publish a report of each assessment no later than five years after the effective date of the rule or order.  While the CFPB believed that the October 2015 amendments to the HMDA rule did not constitute a significant rule, a view that raised the eyebrows of some mortgage industry participants, it decided to nevertheless conduct a voluntary assessment of the rule and issue a report.  The CFPB recently released the report, which is entitled Report on the Home Mortgage Disclosure Act Rule Voluntary Review.  The report addresses not only the October 2015 amendments to the HMDA rule, but also subsequent amendments.

The CFPB states that the “report does not generally consider the potential effectiveness of alternative requirements to HMDA data reporting that might have been or might be adopted by the Bureau, nor does it include specific recommendations by the Bureau to modify any rules.  The Bureau expects that the findings made in this report and the public comments received in response to the November 2021 [Request For Information] on its plans to conduct the review will help inform the Bureau’s future policy decisions concerning HMDA reporting requirements, including whether to commence a rulemaking to make the HMDA Rule more effective in meeting its goals.”

The CFPB advises that it considers the balancing test that it applies to determine whether and how HMDA data should be modified prior to its disclosure to the public, in order to protect applicant and borrower privacy while also fulfilling HMDA’s public disclosure purposes, as well as the related policy guidance that it issued in 2018, to be beyond the scope of the review that it conducted.

The CFPB states that:

“Wherever possible, this review analyzed available data to estimate changes in measures (such as mortgage market coverage ratios) and determined whether these changes are attributable to the HMDA Rule.  However, in many cases this analysis was not possible given the data available to the Bureau during this review.

The primary challenge to this analysis is establishing a counterfactual—what would have occurred were it not for the HMDA Rule—to provide a baseline for evaluating the effects of the HMDA Rule.”

However, as noted in the report, the Economic Growth, Regulatory Relief, and Consumer Protection Act enacted in 2018 modified the HMDA data collection and reporting requirements for insured depository institutions and insured credit unions originating closed-end loans and open-end lines of credit below a certain level to focus their reporting on the pre-October 2015 rule HMDA data elements.  (For closed-end loans the level is less than 500 of such loans in the each of the previous two calendar years, and for open-end lines of credit the level is less than 500 of such credit lines in each of the previous two calendar years.)

Much of the data in the report is dated, as it focuses on 2019 HMDA data.  Findings in the report include the following:

  • HMDA data coverage of all first lien, closed-end originations ranged from 0.93 to 0.97 between Q1 of 2015 and Q4 of 2019 when measured quarterly.  This implies that between 93 percent and 97 percent of all first lien, closed-end originations made between Q1 of 2015 and Q4 of 2019 are observed in the HMDA data.
  • Consistent with the 2015 Rule’s increase in the closed-end reporting threshold for depository institutions, HMDA data coverage of first lien, closed-end mortgages decreased between Q1 of 2017 and Q1 of 2018.
  • Starting in Q1 of 2018, the annual HMDA data coverage of HELOCs was around 0.80.  This implies that 80 percent of HELOC originations were reported under HMDA between 2018 and 2020.
  • HMDA data coverage is positively related to census tract relative income, with the highest HMDA data coverage occurring in upper income tracts and the lowest HMDA data coverage occurring in low-income tracts.
  • HMDA data coverage of non-conventional loans, which make up a small share of the total mortgage market, is higher than HMDA data coverage for conventional loans throughout the 2015-2019 period.
  • HMDA data coverage throughout the 2015 to 2019 period is highest in urban census tracts and lowest in rural tracts.  The CFPB notes that based on the requirement that a lender must have a home or branch office in a Metropolitan Statistical Area to be subject to HMDA, “this pattern is not surprising.”
  • HMDA data coverage ratios for properties located in census tracts where most residents are not White non-Hispanic (“majority-minority census tracts”) and for census tracts where most residents are White non-Hispanic appear similar just before and after most of the October 2015 rule provisions took effect in January of 2018.  The declines in HMDA data coverage beginning in Q1 of 2018 are 3.1 and 1.1 percentage points for majority White non-Hispanic and majority-minority census tracts, respectively.
  • The 2019 HMDA data show that there is no clear pattern across racial groups on the shares of race and ethnicity information that financial institutions collected on the basis of visual observation or surname.  (When a lender takes an application in person and the applicant does not provide race and ethnicity information, the lender must record that information on the basis of visual observation or surname.)
  • The 2019 HMDA data show that the age distribution of mortgage borrowers varies across race and ethnicity. For instance:
    • The median age of Hispanic White borrowers was 41 and their average age was 43, making them the youngest group among all race and ethnicity groups.  The median ages of Black and non-Hispanic White borrowers were both 46 in 2019.
    • Denial rates also vary by the age of applicants.  With the exception of Rural Housing Service/Farm Services Agency loans, the denial rates of most closed-end mortgage applications generally increase with age.  In particular, the denial rates for applicants aged 62 or older were higher than those for applicants younger than 62, except for home equity lines of credit (HELOCs) and reverse mortgages.
  • For complete applications, in 2019, about 14 percent of applications for a loan on a one-unit property were denied.  In contrast, the denial rates for applications for a loan on a two-, three- and four-unit property were 18 percent, 19 percent, and 18 percent respectively.
  • In 2019, 65 percent of multifamily originations were for five to 24 units and about 9 percent of originations were for large multifamily dwellings with more than 150 units.
  • In 2019, over 1,700 originated manufactured home loans were secured by more than four units.  The CFPB observes that “[m]anufactured homes are a vital part of the housing market, providing affordable alternatives to site-built properties, especially in rural areas.”
  • In 2019, about 6 percent of multifamily originations were for properties with at least one affordable unit, and more than half of originations with at least one affordable unit were for exclusively income-restricted multifamily dwellings.  In general, multifamily loans with a greater number of total units also had a greater share of affordable units.
  • In 2019, the median interest rate for loans secured only by the manufactured home (i.e., not also the underlying real estate) was 8.490 percent, while the interest rate for loans secured by the manufactured home and underlying real estate was 4.750 percent.
  • In 2019, about 85 percent of all originations were directly submitted, and initially payable, to the reporting institution.
  • With regard to HELOCs:
    • They are more likely to have a smaller loan amount and higher interest rates than closed-end loans.
    • Applicants are more likely to be denied than closed-end loan applicants.
    • Borrowers are more likely to be non-Hispanic White, have higher income, and live in high-income tracts and metropolitan areas than closed-end loan borrowers
  • Consistent with the median interest rate trends by racial group, Black and Hispanic White borrowers had greater interest rate spreads above a market rate than non-Hispanic White and Asian borrowers.
  • Overall, the median total loan costs in 2019 ranged from about $3,400 to a little over $7,100 across different loan types. The average was $4,809.  The standard deviation of total loan costs was $9,700, about twice the size of the average, indicating a wide spread of the total loan costs.
  • Based on changes to the HMDA rule between 2015 and 2018:
    • For existing closed-end loan reporters the estimated increase in ongoing compliance costs for HMDA reporting per loan application (in 2018 dollars) was approximately $42 for a representative low-complexity financial institution with a loan/application register (LAR) size of 50 records, $11 for a representative moderate-complexity financial institution with a LAR size of 1,000 records, and $0.47 for a representative high-complexity financial institution with a LAR size of 50,000 records.
    • For open-end reporters, who are almost all newly reporting open-end lines of credit under the HMDA Rule (including financial institutions who previously reported HMDA data for their closed-end loan applications), the estimated increase in ongoing HMDA compliance costs per loan application was approximately $50 for a representative low-complexity financial institution with a LAR size of 500 records, $45 for a representative moderate-complexity financial institution with a LAR size of 1,000 records, and $10 for a representative high-complexity financial institution with a LAR size of 30,000 records.