When lending inequities fuel housing disparities

Lenders must do their due diligence to reduce credit risk. But what if the tools and resources they’re using don’t paint the whole picture?

Most financial lending is based upon traditional credit scoring, an assessment of creditworthiness based on analysis of bureau and demographic data. Unfortunately, these conventional methods tend to disadvantage certain populations – particularly people of color – by inadvertently exacerbating existing inequities in our credit system.

There are ways lenders can more accurately gauge the creditworthiness of all borrowers, including those impeded by current credit-scoring practices.

Sign up for the free BAI Banking Strategies newsletter and get industry insights delivered to your inbox.

Regardless of where it’s developed, an individual’s credit score is based on the information fed into it. The credit bureau system favors those who borrow (and repay) money for things such as cars and housing – in general, an ideal measure for middle class households.

Lenders can take several actions to more fairly assess the creditworthiness of the unbanked and underbanked by leveraging their data and analytic resources in new ways, including:

Visual analytics tools: The first step in solving a problem is identifying and measuring it. By applying visual analytics to their data, banks can more readily spot areas of concern. As they say, a picture is worth a thousand words.

Non-traditional data: Non-traditional payment data related to rent, utilities and even streaming services has proven an effective risk measure for those with no credit bureau or bank account. For those with bank accounts, banking transactions have been strong indicators of creditworthiness globally.

AI and machine learning: Large volumes of data are a game changer for financial institutions in managing credit risk, detecting fraud and financial crime, and attracting loyal customers. AI/ML enables banks to efficiently sift through large datasets and capture behavior for segments not well served by traditional data sources.

For U.S. households that are “unbanked” or “underbanked,” the existing system tends to reinforce societal gaps. There is little to no credit data to capture and help establish one’s creditworthiness in the current system.

We know that access to financial services promotes economic and social mobility, so finding ways to include these millions of individuals in the credit cycle is critical to both economic growth and social justice. Owning a car, for example, can open job opportunities further from home – and home ownership is one of the most effective ways to build long term wealth.

Overcoming lending inequities

Recent data tell us that negative effects on the unbanked and underbanked are real. Consider the findings of a collaborative study by the Center for NYC Neighborhoods and SAS that examines home ownership disparities in New York City. Through AI-powered analysis of the city’s housing data, the Black Homeownership Project revealed clear neighborhood inequities.

Neighborhoods with a higher proportion of Black and Hispanic homeowners, for instance, have lower home values even when home age and square footage are equal. On top of that, the total cost of acquiring home purchase loans is higher for Black and Hispanic borrowers than for other races, even when controlling for differences in down payments and home values. And Black and Hispanic borrowers experience higher interest rates, lower lender credits and higher discount rates.

Identifying such problems is only the starting point. Technology can do much more by helping to solve them. In New York, visual analytics dashboards are helping the Center for NYC Neighborhoods design programs and advocate for data-driven policies that help address key challenges Black homeowners face, including more foreclosures and more tax liens than their non-Black counterparts.

Likewise, it is only through proactive change that the financial services industry can help reverse such inequities and close home ownership’s racial divide. As a first step, advanced analytic techniques, applied to large and diverse data sources, can help make lending practices fairer. This approach can give lenders better insights into the populations and borrowers that they serve.

At the end of the day, home ownership is a dream that should be an attainable reality for all. For lenders and borrowers alike, using analytic technology to ensure lending practices are as fair as they can be is a win-win.

Naeem Siddiqi is a senior advisor for risk research and quantitative solutions at SAS.