Are you evaluating your models for fair lending compliance?

Disconnects and delays in fair lending review can prove costly

No matter how hard developers try to identify and avoid discriminatory variables in models, these variables can still lead to unintentional discrimination.

Even seemingly innocuous lifestyle questions, such as which brand of deodorant or shampoo someone prefers, can potentially signal a person’s gender, which is a prohibited basis for making lending decisions. As a result, these models – often used by marketing departments – can lead to unintentional discriminatory segmentation if they aren’t closely monitored.

A model review team needs to understand how models are developed, how model variables work, and whether the weighting of each variable could have an unintentional fair lending impact. Otherwise, models might unintentionally affect protected groups in unfair ways, including:

Excluding or pricing higher majority-minority census tracts
Creating disparate impacts in low- and moderate-income areas
Weighting female applicants less favorably than male applicants

Misunderstandings about the models in use can result in penalties, consumer reimbursements, and an increased risk of legal and reputational harm.

Collaboration to identify fair lending risk in models should happen early and often

The need for a fair lending compliance program in models at the time of internal creation or vendor selection is just as important as the need for validation. How organizations approach fair lending review alongside model validation can significantly influence levels of risk and efficiency.

Yet, in many organizations, model risk management, legal, and compliance departments collaborate less than they should, if at all. In some cases, the fair lending topic might enter the conversation so late in the process that extensive and costly rework is necessary for compliance.

Addressing disconnects and improving communication between model selection, creation, and validation with fair lending compliance teams can help manage risk and improve productivity, saving organizations substantial time and money.

Teams should evaluate specific models through the lens of fair lending practices

Regulators have not released specific guidance on which models require fair lending review. Organizations should establish definitions and guidance on what constitutes a need for review and when to bring model risk management and compliance teams together.

Models to consider include those that influence:

Third-party fintech models
Marketing and targeting applications
Credit applications
Underwriting and pricing
Any customer-facing offering that uses a model
Potentially, fair lending analysis and redlining models

Organizations should also review any models that might come from mergers, acquisitions, partnerships, and third-party vendors.

In addition, models developed and offered by fintechs continue to expand. These models should receive a close evaluation before use because fintechs are subject to different regulatory oversight of fair lending and model risk laws and regulations. The responsibility for reasonable risk management ultimately rests on the financial services organization.

Your teams should review fair lending on models early in the process

When organizations reach the point of validating a model, a review of fair lending compliance should be conducted at that time or, ideally, much sooner.

New changes can introduce unintentional and unforeseen fair lending violations, so organizations should review existing models for compliance any time they receive updates. If the model review process historically has not incorporated a fair lending review, then organizations should review all applicable models to look for discriminatory attributes that might not have been previously considered.

In general, it’s best to not change models before consulting compliance or fair and responsible banking departments. Organizations should have model review policies and procedures that outline trigger events regarding whether legal and compliance departments are required to review a model and include an up-to-date inventory of all models used and their corresponding owners.

Collaborative model validation can improve your processes

Crowe specialists can help your organization bridge the gap between validating models and steering fair lending. We eliminate unnecessary boundaries by bringing the right experience to the table when you need it.