■ ■ ■ Collection of Data on Ethnicity, Race, Sex, Age and Income
(§1003.4(b)): This data collection is optional for purchased
The Importance of Scrubbing and Elements
of a Scrub Screen
As indicated above, scrubbing loans and loan applications subject
to reporting is essential, as existing reporters know. Scrubbing
is usually a function where Compliance (and sometimes LOB’s)
execute one or more of the following techniques:
■ ■ ■ Select a sample of loan files—the percentage of sampling should
be contingent on the controls in place. Many banks start with a
15 percent sample of each LOB or decision center’s records, and
increase or decrease future sampling depending on the error rate.
■ ■ ■ Utilize a “HMDA Checklist”—some banks have used simple
Excel models (pivot tables) or more customized web-based
applications to track error rates. Note that the FFIEC and outside vendor HMDA software allows for scrubbing, but the
checklist can be useful.
■ ■ ■ Tailor a custom web-based application for integrating HMDA
data from multiple source loan systems.
■ ■ ■ If developing a scrub screen program within an internally-developed data application or platform, the data system should
• A scrub screen program to include:
• One column to show the data that will be reported unless
• A second column having boxes to populate if the user deems
a change should be made to any data point;
• Ability to save the data;
• A prompt to block a user from proceeding to another file
if they haven’t saved the file just scrubbed;
• A place for the reviewer to enter their name, initials, or
• Time stamping of the date(s) the screen was used and
• A reporting capability that will allow Compliance or Audit
to assess tallies on files scrubbed and the number and corresponding percentage of files with changes made, changes
by required HMDA data field, and any other elements Compliance and/or Audit would like added. This will facilitate
the tracing of errors to their source, with the goal being
rapid extinguishing of the error source, be it one or more
individuals or one or more processes.
• Each user (LOB staff, Compliance, engaged outside consultants, etc.) should:
• Enter their name or user identifier code (for QC/reviews);
• Be trained in its use by Compliance;
• Compare the entered data to the actual loan/application
file reviewed; and,
• Enter any changes in the scrub screen.
The Importance of Training
There are immense benefits of proper and thorough training. With
the many HMDA changes in play, Compliance should have already
developed, or soon develop, HMDA-related training unique to
each LOB’s reporting responsibilities. Once loan scrubbing begins,
and should any error rates begin to exceed 1% to 2% of records
for any one data point category, Compliance should provide corrective training to the pertinent staff. Once the error rates drop
to acceptable levels, the success of the training will be evident,
and everyone’s HMDA work will become lighter.
The changes and strategies outlined in this article can be overwhelming. The Bureau’s website provides a comprehensive review
of the new rules in total, and it can be referenced for better understanding of the scope of the changes. With that understanding
as a foundation, the approaches outlined in this article can help
organizations prepare for the new requirements and ensure the
necessary adjustments have been addressed. ■
ABOUT THE AUTHOR
RICHARD KOHRUMEL, CRCM, CAMS, is currently Principal of
Richard A Kohrumel LLC. Rick has over 24 years of bank regulatory
supervision experience, from being one of the country’s first
Federal Bank examiners at the Federal Reserve Bank of NY, to
Manager of the Compliance Examinations and Consumer Affairs
Department for the Southeast Region of the Office of Thrift
Supervision. In addition, Rick has been Compliance Officer/
Manager at three regional and two community banks. Lastly,
Rick has worked as a regulatory FI consultant for over five years,
including owning his own consulting practice.
At RBC Centura Bank, Raleigh, NC, Rick successfully managed
a massive HMDA data correction effort for RBC Mortgage Company
and its newly-acquired Sterling Mortgage Company. There, he
worked with 13 outside mortgage lending consultants, and local
Compliance, legal and IT staff to obtain retroactive compliance
and build a HMDA data platform combining data from three LOSs,
all to the satisfaction of FRB of Richmond examiners. Rick also
successfully remediated all HMDA issues identified by the FDIC
in its sanctioning of the former Darby Bank and Trust Company,
Vidalia, GA (now part of Ameris Bank). Rick can be contacted at
email@example.com or (770) 639-0558.
1 Consumer Financial Protection Bureau “Summary of Reportable Data,”
version 1.0 10/15/2015; http://files.consumerfinance.gov/f/201510_cfpb_
2 CFPB Final Rule and Official Interpretations, dated October 15, 2015, p.
3 Ibid., p. 116.
4 For references to grossing up non-taxed income, for example, see Appendix
Q of Regulation Z; http://www.consumerfinance.gov/eregulations/1026-