2018 HMDA Prep: Sample Sizes & Error Thresholds for Examiner Testing

11/16/2017 - By Jennifer Paradise, CRCM

In an earlier post entitled 2018 HMDA Prep: Examiner Guidelines for Testing HMDA Data, we provided a brief overview of the HMDA Examiner Transaction Testing Guidelines issued by the Federal Financial Institutions Examination Council (FFIEC) in August 2017. The guidelines detailed how HMDA Loan Application Register (LAR) data, collected beginning January 1, 2018 will be validated by examiners. In this article, we elaborate upon two topics that should be of importance to compliance officers -- how the testing sample is selected and the specific thresholds for resubmission due to errors.

Testing Samples

First, recognize that all data fields could potentially be selected for review. Likewise examiners may select certain only fields for transaction testing.

The samples used for testing are based upon the number of entries on an institution’s LAR (the “LAR Count”). The "Total Sample Size," as depicted in column A of the HMDA Transaction Testing Sample Sizes and Thresholds Table (“HMDA Table”) within the guidelines as pictured below will be randomly chosen by examiners.

*For institutions with fewer than 30 LAR lines, the full sample size is the institution’s total number of LAR lines. The Resubmission Threshold number remains at 3. Accordingly, the Resubmission Threshold percentage will be higher for institutions with fewer than 30 LAR lines (Federal Financial Institutions Examination Council, 2017)

Transaction testing consists of two defined stages. In Stage 1, examiners begin by reviewing a subset of the Total Sample, referred to as the Initial Sample, in the HMDA Table above. If there are less than two errors identified during the review of the Initial Sample, no further review is required. That’s right. The Initial Sample Threshold, as depicted the HMDA Table is two, regardless of the size of your LAR. If you passed Stage 1, do your happy dance because your institution’s LAR passed the Initial Sample review!

If there are two or more errors in the Initial Sample, examiners will proceed to Stage 2. Now the remaining Total Sample is subject to mandatory review of all data fields that had one or more errors in the Initial Sample and the review may or may not include any or all data fields found to have no errors in Stage 1. This could mean a few more files to pull, or a mountain of files, depending on the examiner.

Thresholds for Resubmission

If the total number of errors in any data field equals or exceeds the “Resubmission Threshold” as depicted in the HMDA Table, examiners will require the institution to correct and resubmit its LAR. It’s important to note that for most community banks and credit unions, where the LAR Count is less than or equal to 500, the Resubmission Threshold is just three. In addition, an institution could be required to correct and resubmit its LAR if the examiner has reason to believe that errors in any one or more individual data fields make analysis of the HMDA data unreliable.

The Resubmission Threshold is higher for institutions with a relatively low number of LAR entries. Our hope is that all our clients will pass Stage 1 and be able to do the HMDA Happy Dance. In the meantime, you can count on us to assist in any form that may be. Do not hesitate to reach out to your financial institution consultant at Saltmarsh Cleaveland & Gund. You are not alone.

About the Author | Jennifer Paradise, CRCM

Jennifer is a consultant in the Financial Institutions Advisory Group at Saltmarsh, Cleaveland & Gund and has been serving the financial institutions industry since 1990. She is primarily involved in performing fair lending, loan internal audit, loan compliance, and other consulting services for the firm's financial institution clients throughout the region.

View Full Bio