top of page

HMDA is Coming ... Review #2


While Ben's away, Gregg will play ... a special [re]issue of the Wednesday compliance newsletter while Ben is on vacation.

 

Note from Ben:

I appreciate Gregg Oberg taking over newsletter duties for the second week in a row. See you all next Wednesday!

 

Alright, Crocodile Dundee will be back tomorrow, so I've got this last newsletter to try and steal his audience. Last week we looked at scenarios in which demographic information is not required to be collected.


Many of you have already started, or will in the very near future, (no later than Jan 1; if this is news to you, call me yesterday!) to collect 2018 demographic information. Let's take a look at Ben's guide to demographic data training topics, with some updates from the questions we've seen as clients begin the "practice" collection period. The focus of my commentary will be the interaction between applicants and MLOs for demographic collection, and how you might answer some common questions the MLOs have as they begin collection.


 

We, in the regulation, I mean mortgage, business are all learning about the new HMDA requirements. One particular issue that arises is collecting the more complicated set of Government Monitoring Information (GMI) data, especially when it's based only on visual observation.


[Gregg's Commentary: Is GMI still the right phrase? We use it for simplicity, but how does it sound from an applicant standpoint? Just say the words "now its time to collect your government monitoring information" and let it sink in ... if you considered pulling the tin foil hat out of your desk, then you're not alone. There is a very real potential to create more anxiety than necessary on the part of applicants simply by using a phrase with such a negative connotation.

Many applicants, whether in old HMDA or new, are uneasy providing information to the government. This often can be due to mistrust of the government or other organizations. Combine that mistrust with the poor phrasing of GMI, and you can quickly see how this can lead to longer conversations with applicants on demographic collection, and potentially lower frequency of disclosure by applicants.


How are you handling those conversations, and are you seeing the same concern? It very likely could end up not making a difference, but is there a better term we can use?]



"Disaggregated"

Soon HMDA will require us to collect "disaggregated" demographic information. According to the dictionary, that just means we're taking this information (race and ethnicity) and breaking them down into more specific subcategories. So we report the category for "Native Hawaiian" but then also report a subcategory such as "Guamanian" or "Samoan."


[Gregg's Commentary: Ben felt the need to go to Webster here, so maybe not the simplest way to describe for training purposes? When training MLOs on demographic collection, you can refer to the two forms of data collection as "old HMDA" and "new HMDA" for simplicity, or something similar. Whatever you do call it, I'd pick one and stick to it, no need to add to stress with confusing terminology.]


A picture of the new 1003 addendum (available here) might explain better:

Guessing Game When Applicant Declines to Self-Report?


Gregg's Commentary: [Lets be honest on this one. How often do you actually have to make an identification of the applicant? The requirement to determine demographics on visual or surname basis applies only to in-person applications or those online with a video component. It then only applies when the applicant declines to provide the information. Finally, it only allows an originator to select from the old HMDA (aggregate) categories. I think sometimes MLOs are thinking of this rule in terms of "when don't I make visual ID?" Instead, consider framing the visual observation component as an exception to the norm, not the norm from which everything is excepted.]


So the question that arises is what happens when the applicant declines to provide this new demographic information. Are originators expected to become experts in demographic differences? Could you (assuming you're a well-traveled professional with high cultural sensitivity) accurately fill in the particular tribe that a native american applicant belongs to?


Well the good news here is that when the lender has to fill in the demographic information (i.e. when the applicant declines to self-report), we do not have to try and guess at the disaggregated subcategories--we only need to report the main categories.


These main categories are ...


For Ethnicity:

  • Hispanic or Latino

  • Not Hispanic or Latino

For Race:

  • American Indian or Alaska Native

  • Asian

  • Black or African American

  • Native Hawaiian or Other Pacific Islander

  • White

So the good news here is that you won't need to devote time to training originators on the difference between these ethnic and racial subcategories.


Gregg's Commentary: [the better news is they already know these categories, its just "old HMDA." New HMDA, only applicant can touch. Those fields common to old and new, fair game.]


Other Scenarios to Cover in Training

Developing a training on this demographic data topic? Well prepare for all the different scenarios that can arise. Here are several that come to mind and enough information to get you started:


Gregg's Commentary: [From the collecting perspective, demographic data is simple. THE APPLICANT IS ALWAYS RIGHT. Originators are not here to make a determination of validity for applicant demographics. And although we are required to report demographics under HMDA, we are not required to report them CORRECTLY. This is a "report as disclosed" field. 99% of the questions an originator may have as to the propriety of applicant selections can be solved by this rule; we take whatever they give us, and pass it up the chain toward the LAR. The rule of five is not important from a collection standpoint, so don't bog down individuals who don't need to know it.]


1) Borrower reports more than one ethnicity or race:


Don't correct the borrower, just report what they have provided. There are some nuances here, such as when the borrower selects more than 5 races. See Appendix B #9 for more info.


2) Borrower applies by mail, online, or by phone, and fails to provide demographic information but also fails to check box "I do not wish to provide this information."


If you meet the borrower in person to complete the application, you must ask for the demographic information. If they decline, then you report that they declined and leave the disaggregated subcategories blank. But if you do not hold an in person meeting until after the application process, then you are actually not required to collect their demographic data at all. Let's see how this plays out in practice! Appendix B #12.


[Gregg's Commentary: In these scenarios, one key distinction between the scenarios described in Appendix B #12 is affirmative declination to provide information. Take an example: if an online application comes in with nothing selected on the demographic addendum (is anybody hard coding their demographic fields for online apps?), this is failing to provide information. In the same example, had the applicant checked "do not wish to provide," this would be an affirmative declination to provide. In the former case, you must request info again and make visual ID when they decline. In the latter, your obligation under HMDA is met.


As a practical matter, could or would you want to call the applicant to obtain an affirmative declination prior to meeting them in person to complete the application?]


3) Borrower provides incomplete demographic information.


Generally, you report whatever they provided.


[Gregg's Commentary: There is a distinction between what we 'collect' and what we "report." From the MLO collection standpoint, we collect whatever the borrower provided. From a reporting standpoint, we report what they provide, subject to the rule of five.]


4) Borrower provides demographic information but also selects "I do not wish to provide this information".


Except for in-person applications, you will still report the data provided. Appendix B #13.

 

In Other News

 

On My Mind ...

  • Thanks to everyone who commented on the newsletter last week in emails, I really appreciate the feedback and look forward to engaging with you in the future as I start my own newsletter. If you haven't heard back from me yet, you will shortly!

  • Ben will be back at some point this week, once he digs out of two weeks of emails I'm sure many of you will be hearing from him again soon. He's back with regularly scheduled newsletters next week

 

Thanks so much for reading our weekly newsletters. We're not always going to be perfect, but because we always do our best and try not to overpromise, we hope that we're always going to be trustworthy. Your calls and e-mails are very helpful - please keep contributing.

**These are our opinions. We're not authorized, or willing, to express those of others.**

bottom of page