<img src="https://ws.zoominfo.com/pixel/pIUYSip8PKsGpxhxzC1V" width="1" height="1" style="display: none;">

How to Avoid Accidental Lending Bias During the Mortgage Boom

2 min read
Sep 2, 2020

With demand for home mortgage refinancing at a record high, mortgage lenders are working at a frenzied pace to keep up. In some cases, they are discovering there just are not enough resources or hours in the day to refinance every eligible applicant.

The big question, then, is who gets served?

Is it consumers with a long-standing relationship or good customers or members who use many of the financial institution’s products and services? Is it loans of a certain size? Customers who already have a mortgage with the FI?

Before making this decision, make sure you take a step back and evaluate the demographics of those segments. While it may seem fair to pick a group and stick with it—consistency is a good policy when it comes to Fair Lending—it matters who is in the group. If it turns out your group inadvertently excludes a protected class of borrowers, you may end up in violation of the disparate impact definition of discrimination.

The implications can go beyond Fair Lending. Just reflect on what happened when large national banks prioritized Paycheck Protection Program (PPP) loans for large companies instead of accommodating the small businesses—the real customers that the program was intended to serve.

Looking for Hidden Fair Lending Bias

Fair lending bias is an avoidable issue if your FI takes the right steps to detect and correct it. It is all about being proactive and consistent to ensure Fair Lending bias does not sneak into your loan process. 

The solution is so simple: Conducting a Fair Lending gap analysis of groups before drafting any policy or product. Analyzing the distribution of protected classes in your proposed policies, products, and services allow you to see if there could be any unintended consequences to decisions that could result in Fair Lending violations.

Multi-variate modeling (where variables like credit score, LTV, and DTI are included) is useful for this task, especially if you have a large data set. Multi-variate modeling allows you to see the relationship between multiple variables, providing more insights than the “simple comparison means.” This allows you to better investigate disparities in model prediction, understand the impact, and make proactive changes.

Fair Lending analysis will also give you a chance to correct disparities and find a better way to allocate limited resources for refinancing. Not only can this help prevent your FI from accidentally injuring a consumer, but it also shows examiners that your institution takes a proactive stance when it comes to Fair Lending.

Is your FI taking steps to manage the potential risk of limiting mortgage refinancing to a select group of consumers? Make sure you are not doing anything to inadvertently discriminate against a protected class.


Related: How to Build a Strong Fair Lending & Redlining Compliance Management System


Subscribe to the Nsight Blog