Navy Federal Credit Union, the largest credit union in the United States, is currently facing a class action lawsuit regarding its lending practices. Plaintiffs claim that the credit union is discriminating against people of color and other minorities.

A CNN Investigative Report released in December found substantial evidence to back the plaintiffs’ claims, citing that in 2022, 77% of white applicants saw their loans approved, whereas only 48% of black applicants were approved. Additionally, many of these black applicants were in good financial standing, with plenty of income (some over 200K), great credit scores, and little to no debt. Yet, regardless of their good standing, they were denied loans at Navy Federal, and forced to turn to other financial institutions where they were approved.

However, many argue that the problem is not limited to Navy Federal, but that the Navy Federal lawsuit is simply a result of an outdated system rife with inaccurate information and inherent biases that cripple the entire lending system worldwide.

To gain some insight into the discriminatory nature of current lending metrics and discuss solutions for more equitable loan decisioning, CUSO Magazine sat down with Tim Ray, CEO of VeriFast, a platform that operates in the screening and underwriting industry—primarily focused on tenant screening as well as mortgage underwriting.

CUSO Magazine: Thanks so much for joining us today, Tim. We’re excited to hear more about your thoughts on the situation and how you think things can hopefully be resolved to create more equitable lending in the future. 

Tim Ray: Happy to be here! 

CUSO Magzine: As a lending professional, what are your thoughts on the effectiveness of current loan decisioning tools and how they relate to the Navy Federal lawsuit?

As lenders, all we’re trying to do—credit unions, banks, REITs—is assess an applicant’s ability to pay. Are they going to be a good client? Do they have the ability to pay for the mortgage, lease, or loan that they’re applying for? Until recently, the only metric with which to measure that ability was credit scores.

However, credit checks are always a historical out-of-date view of someone’s financials by necessity. As showcased by the lawsuit against Navy Federal, several historical biases are baked into the way an arbitrary credit score is calculated. Therefore, our current system, using just the credit score to make the decision, is negatively impacting immigrants, people of color, and other marginalized groups.

While it’s true Navy Federal had a higher discrepancy in loan approvals than other financial institutions, part of this lawsuit is to start the conversation about the disparities that do exist within credit scores and credit checks and the unfair lending practices that they perpetuate. So on a deeper level, this has nothing to do with Navy Federal per se, they just happen to be the largest credit union in the nation, so they’re the ones that are being called out in this lawsuit. But overall, it’s not them specifically, this is an issue across the board for everyone. It’s a systematic problem. 

Navy Federal, the credit unions at large, and the banks are playing within the rules that the system and the credit bureaus have created because there hasn’t been another tool until today. Unfortunately, that tool is just not accurate. So ultimately, this is about us having a conversation to change our lending practices and then make a more financially inclusive underwriting system or screening process.

CUSO Magazine: You mentioned that with the way credit scores are traditionally calculated, there are a lot of inherent biases in that system. Can you touch on that?

Tim Ray: Immigrants, people of color, or those who come from financially illiterate backgrounds are historically going to have a much harder time getting a loan—regardless of what their true financial situation is like. 

Credit scores often incorporate qualifications such as job titles, age, ethnicity, etc. into the final score. We don’t know what the scores are made of, but what we are seeing—and what Navy Federal is in hot water for—is that applicants of color, with hundreds of thousands of dollars of income, no debt, and credit scores over 750 are being rejected for a loan in large numbers, disproportionate to the number of white applicants being approved. 

Realistically, if an applicant is making $200,000 a year, their ratio is good, and they have no debt, there is no logical reason why they’re getting rejected. Right? But because of those inherent biases, they are.

Additionally, those with a lack of credit will have issues as well. Let’s say someone has been taught by their parents to utilize cash and debit cards when they were young, as credit cards come with a risk of building debt. Now they want to buy a house or a car but have no credit. Even though they have plenty of income, a good amount of money in the bank, and responsible spending habits, when they apply for a product, the system tells them they must have a credit score and rejects them. 

The system should be saying you have to have money, not that you have to have a certain score. Then we should be using our credit reports to look at bankruptcies or utilization of credit as a follow-up to whether or not you have money, not as the one-and-only factor.

CUSO Magazine: You mentioned earlier that immigrants have difficulties securing loans. When they come into the country, they have little to no credit. But that’s not an accurate reflection of how much money they have or what their lending capability should be.

Tim Ray: Correct. If I were to move to the sunny state of Florida tomorrow from Toronto, I would have a tough time buying a house with my lack of credit. What that says is that realistically, this universal metric that we’re all using to assess someone’s ability to pay and their worthiness to do business with, it’s like a leaky boat, it has a lot of holes in it and a lot of exceptions to it. We’re looking at the wrong metric here.

CUSO Magazine: What is the alternative?

With the advent of open banking and federal regulation around open banking and banking data in real-time, we have much better, more active tools to assess risk and income that we should be using instead. Unlike outdated credit reports, using bank data allows lenders to verify income and financial behavior with real-time financial data.

At VeriFast, we’re finding value in giving people true transparency to verify all their income sources, not just their pay stubs, and looking at their combined financial metrics like their balances, balance trends, and core expenses versus discretionary purchasing to see their ability to withstand financial shock from adverse financial events. So we’re combining AI and bank data to perform real-time cashflow underwriting as an alternative or in addition to credit, where tenant screening and mortgage underwriting are primarily driven by credit checks previously. 

At the end of the day, the industry must transition away from using an arbitrary credit score as a blunt force object to evaluate if a customer is worthy to do business with and instead look at their real-time financials such as income and assets as a first barometer if they’re a good potential client or not. 

CUSO Magazine: So what do you believe the right method or metric would be? You mentioned earlier using open banking and data to look at real-time statistics. Do you believe that is what the industry should use? That there should be an industry-wide reform?

Tim Ray: The industry needs to transition towards cashflow underwriting as the first metric that people look at concerning income from all income sources, and then bring in other things such as balance trend overall, gross income minus your expenses, etc. What’s the applicant’s actual net cash flow? 

Our default income metric today is the gross-to-rent ratio, which is simply crossing your fingers and hoping that there’s no way you can mess this up if you are a three-to-one income-to-mortgage/rent ratio. But as we know, no matter how much you make, you can always find a way to spend more. We often see former professional athletes and celebrities making tons of money and spending way more than they make and going broke. 

What financial institutions need to look at is the core fundamentals of people’s financial behavior and come up with a financial behavior score or a cashflow score that looks at a bundle of different metrics. That’s ideally the solution to create fairness in lending because the numbers are blind, right? The numbers will say what they are and they don’t look at age, race, religion, or anything like that.

CUSO Magazine: You’ve mentioned open banking as a tool for loan decisioning. Why open banking is so important when it comes to determining an applicant’s lending capabilities and why is it important for the credit union?

Tim Ray: There are multiple reasons open banking is so important. With open banking, we can get data from point to point. From the bank where it originates to the underwriter, sometimes it’s another operating unit within the same bank or another institution altogether. Thanks to this, we have much faster, more time-efficient, and more cost-efficient underwriting practices with more data that allows us to get a more accurate picture of the applicant’s financial behaviors and borrowing capabilities.

Open banking also ensures the validity of the data used for underwriting. In the current system, when we download data from the source as a PDF and then we send the PDF to another source, the first step is to look at the efficacy of the documents. We’re wasting energy and manpower on determining if the document is even real data. The fact remains that any of us can go on a PDF editor and edit a bank statement and change a one to a three, and the underwriter on the other side of the unsecured email channel is taking that data at face value, which I think is crazy. We should never be taking PDF documents at face value as proof of anything. It’s like trying to buy something with monopoly money.

If the system can transition to open banking and use just direct source data in general, whether it’s insurance data, bank data, or whatever that data is, if we go from point to point, we can ensure the chain of custody of the data stays intact. We can also operate faster and be more productive, which then drives growth. Ultimately for credit unions, this is why open banking is so important. It is going to offer better data and drive profitability because it’s going to reduce the cost of underwriting from your system.

CUSO Magazine: Do you think AI will have a place in the future of loan decisioning? Do you think that is a better, more effective, bias-free solution? 

Tim Ray: There’s the saying, “garbage in, garbage out.” So, AI inherently is not the solution on its own because if you feed AI with poor data—as in biased credit scores—you are feeding it from historically bad data, right? If you feed it data with biases, it will produce biased results. Garbage in, garbage out. 

However, AI which uses clean data or banking transaction data is an inherently better way. And yes, we are using AI and machine learning to help us with the categorization of deposits and the grouping of income and weighting people’s financial behavior scores.

In my view, ten years from now, every financial institution will use direct connect source data or open banking to determine the lending capabilities of each applicant. We’re not there yet, but it is something we should be moving toward to create better, more equitable lending for all.