Three must-have features for dashboard monitoring of fair lending

Fair lending in the US is governed by the Fair Housing Act and the Equal Credit Opportunity Act.  Both are designed to protect consumers from unfair and discriminatory practices. The penalties for breaking the law and harming consumers are extremely high.

Fair lending professionals focus on four crucial items that have the greatest impact on fair lending analysis and compliance: Underwriting, Pricing, Redlining and Steering. They perform analyses and create useful reports with data analytics. But often, their message is not communicated effectively to senior executives who have little time to dive into the details.

Analytics and monitoring dashboards are an important part of doing this job well and serve as a basis for crafting a powerful message to senior leaders about potential compliance risk.

But what are the key elements that a good dashboard needs to have in order to get that message across effectively? We see three must-have features that every dashboard should have to monitor fair lending risk.

Get your KPIs right

First, it’s critical to select the right indicators to monitor. Whether an organization chooses to monitor loan-to-value ratios (LTV) or FICO scores – or combinations of these and others – will be specific to each organization and its line of business.

We suggest choosing data fields that have impact on underwriting and pricing decisions, such as debt to income ratios. This ratio indicates how much of a borrower’s income is going toward debt repayment. The higher the ratio, the riskier the borrower. This can be combined with the borrower’s age or ethnicity, which will indicate the quality of applications that are coming through the pipeline.

HMDA data is a great source of KPIs to help monitor your redlining risk, or the risk that your organization is illegally discriminating against borrowers because of the location of an applicant’s residence.

If banks limit the source of data analyzed to only that from competitor banks, banks can learn how many loans were made by competitors who are reporting within the same census tract. The data can also provide insights about whether their own advertising, marketing and market penetration is at risk. By focusing on the information that matters most, one can get a deeper insight on the risks in a portfolio.

Get the historical view as well

A second must-have feature is historical trend analysis. Monitoring how a KPI evolves over time can provide early warning signs if there is risk ahead. It can also be used to see if a strategy or plan is working or not.

In general, we say that key indicators are “works in progress,” meaning they can always be redesigned and improved as new data becomes available which can make the indicator a better predictive factor of the future.

Be able to drill-down on your data

Once you’ve got the right KPIs and historical data sets, banks need to be able to use that information to gain a complete picture of what’s happening by drilling down into the data. Drilling down means taking a general view of your analysis and narrowing it to the most granular level, usually to the application level.

For example, if a bank has built a KPI to monitor the FICO levels of minority applicants on a monthly basis, and the current month’s report indicates a sudden decrease in FICO scores overall, a bank can drill down to the loan level and discover more. Sometimes, the score decrease is related to a data error in the FICO score field in the database.

Drilling down not only gives a more narrow view, it can also help users see data from a different angle or point of view. This enhances the user’s understanding of the analysis and the data that supports it. 

Challenges in setting up fair lending compliance

Getting fair lending monitoring right has evolved into a technical challenge. 

One reason is the lack of skilled individuals available on the job market. Compliance has become an exercise that requires the use of sophisticated data science technology. Because of the rapid advances in technology, people working in compliance are being asked to do data-heavy analyses and some requires the use of tools, such as R or SAS.

In addition, companies struggle with incomplete, missing or messy data, which complicates the ability to do more than a basic analysis of loan profiles. All these factors leave a company exposed to compliance risks.

No matter where a company is on its journey of fair lending compliance, the goal should be to get a single data repository for their analytics with a data governance policy in place.

They should use well-defined reports that are produced automatically on a periodic basis and be alerted about anomalies. And their compliance monitoring should be automated and able to perform “mock” regulatory exams for compliance.

This way, lending institutions can fulfill their societal role, providing capital on a fair basis, regardless of race, religion or origin. For many people in the US, the American dream remains home ownership.