After inputting, reviewing, and re-reviewing, you finally filed your Home Mortgage Disclosure Act (HMDA) Loan Application Register (LAR) for 2018. You breathed a sigh of relief, and perhaps you even took an afternoon off—you deserved it. You worked hard getting it filed on time. But … are you really done with your HMDA data? Have you stopped to think about what that data says? What will regulators and others see when they review your recently submitted HMDA data? Have you looked at it so you can answer any questions that may arise?
Whether you’re a large filer or were lucky enough to get the partial exemption, your data will be analyzed. So why not analyze your HMDA data for underwriting and pricing disparities before your regulators do. Your underwriting analysis should consider the ratio of denied applications by ethnicity, race and gender as compared to all applications received. A ratio of denied applications by a prohibited basis group that significantly exceeds the ratio of denied applications by a control group is reason to do further analysis. Furthermore, a review of the ratios based on processing times and the rate of withdrawn and incomplete applications by prohibited basis groups should be conducted to determine if it appears there could be a difference in treatment during the application process. A pricing review should also be conducted, if data is available, to look at differences in pricing between prohibited basis groups and a control group. Any significant differences in pricing might warrant further review.
Having a significant disparity between groups doesn’t necessarily mean discrimination has occurred, but it does mean that you should take a deeper look. If you’re fortunate enough to have collected the HMDA plus data such as debt-to-income ratio, loan-to-value ratio, credit score, annual percentage rate or interest rate and possibly loan costs, you can electronically conduct a comparative file review looking for matched pairs consisting of similarly situated applicants where someone in a prohibited basis group was denied and someone from a control group was approved. You can also electronically compare similarly situated applicants by application or lock date to review for matched pairs for pricing to determine whether someone in a prohibited basis group was priced higher than someone in a control group. Only if you have any matching pairs would you need to take the next step and look at physical files for any other explanations for the differences in underwriting decisions or pricing.
It sounds like a lot of work, right? Why would you do this? Because your regulators will, and if there are any disparities, they will most likely ask questions during your next fair lending examination. If you’re a large HMDA filer, regulators will have even more data and can electronically conduct the comparative file analysis. They will know if you have similarly situated applicants where it appears they were treated differently because they were in a prohibited basis group. In addition, community and special interest groups also review this data and sometimes target certain institutions with questionable data for additional investigation and possible referrals to HUD. If you have more information than they do, you’ll be ready when questions are asked. You’ll have already reviewed the files and will have explanations for any disparities. The information is starting to be released; the Consumer Financial Protection Bureau released modified 2018 LAR data on March 29, 2019, and later this year the aggregate and disclosure reports will be released.
Worst case scenario: you go through this exercise and you think you may have a problem. What now? You can effectively assess the situation, look at your controls for fair lending risk, and make appropriate changes in your underwriting and pricing practices, monitoring and audit programs and training schedules and remediate any harm to customers, ensuring a better outcome in the future. Demonstrating the effectiveness of your compliance management system to detect and correct any problems goes a long way with your regulators.
And like they say, “Knowledge is power.”
Author(s)
