Sign up to receive our monthly newsletter!

Is HMDA Data Enough for Risk Management?

Jun 1, 2020 by Loretta Kirkwood

With the expanded HMDA data requirements, it may seem impossible to consider that there is anything else lenders should be concerned about. However, it is important to address other potential data requirements. Regulatory examinations, HMDA data reviews and fair lending analysis generally expand data captured beyond what is reportable.

Federal and state regulatory examinations include HMDA data integrity reviews and the validation of what are most commonly referred to as “derived” fields. This simply means that the validation of certain HMDA data fields requires that a lender provide “other” non-HMDA reportable data to support the accuracy of the field value. The following are examples of data that lenders need to capture for verification of HMDA reportable fields:

Method of Application (Application Taken By)

  • Method of application fields for applicant and co-applicant are used to determine if the correct value is reported for Visual Observation. For example, all applications that are taken face-to-face must have a value of 1 – Collected on the basis of visual observation or surname or 2 – Not Collected on the basis of visual observation or surname.
  • It’s important to remember that if an application is taken face-to-face, “Information Not Provided” is not permitted as a reportable value for race, ethnicity or gender. For applications taken face-to-face, a lender is required to provide applicant demographic information based on visual observation or surname when the applicant(s) selects not to provide the information.
  • For applications not taken face-to-face, a lender may NOT use surname or any other factor to update applicant demographic information when an applicant chooses not to provide race, ethnicity, and gender.
  • Examiners are additionally asking lenders to provide analysis of the distribution of applications based on the method of application; i.e., face-to-face, internet, mail, telephone.

Rate Spread

  • Annual Percentage Rate (APR), Rate Lock Date and Amortization Type are required during internal audits and regulatory examinations to validate the accuracy of the reported rate spread.


  • Applicant demographic information includes reporting age for both the applicant and co-applicant.
  • Recent regulatory HMDA data integrity examinations have included the validation of age, using date of birth.
  • It is important to remember that the correct calculation for reporting age is based on the application date (not the action date) and the date of birth.
  • HMDA RELIEF includes a data integrity test that compares the calculated or derived age to the reported age using date of birth to flag any differences. A Group Edit is available to recalculate discrepancies automatically.

QuestSoft adds additional statistics and warnings over and above the CFPB validity and quality edits to ensure that our customers can more easily prepare for and pass their exams. Errors in HMDA data expose lenders to civil money penalties, resubmissions, and/or matters requiring attention (MRA) that result in restricted business practices.

Additional areas of concern for lenders include the S.2155 Partial Exemption and fair lending risk management. QuestSoft’s HMDA RELIEF software is designed to collect all HMDA data fields, while still allowing users to optionally report and modify selected fields. Data will automatically reflect the appropriate exempt field values in the submission file.

S.2155 Partial Exemption

  • Most loan origination system (LOS) vendors have established logic to convert actual loan data to HMDA exempt codes in data extracts for lenders subject to the partial exemption. Lenders who select this option will limit data available for regulatory exams and fair lending analysis.
  • Regulators frequently request actual data for many of the exempt fields during a Compliance and/or Fair Lending examination in the same way non-HMDA fields were requested prior to the expanded HMDA data requirements.

  • Fair Lending Risk Management

    • Fair lending risk management efforts will generally require expanded data collection even beyond the 110 HMDA reportable data fields to ensure data is appropriately grouped to identify risk and monitor performance.
    • Examples of additional data needed to effectively analyze performance include branch, channel, and loan program. Credit and pricing information that will likely be helpful in fair lending analysis include exception codes, bankruptcy and foreclosure information, as well as information about the type of loan such as the identification of construction permanent, single closing or workout loans. 

    Remember to work with internal lending personnel to identify specific business practices that may assist in effectively analyzing performance and limiting inaccurate results or conclusions. Review current data extracts and HMDA processes to ensure adequate information is available for exam management, data integrity testing, internal monitoring and performance analysis.