All this week, we’ve been looking at the data and algorithms behind credit scores. While many lenders use FICO scores, thousands also rely on FICO’s competitor, VantageScore.

It was founded by the three credit bureaus — Experian, Equifax and TransUnion — in 2006. The company, which is independently managed, says its scoring model is more inclusive and predictive of credit risk than traditional models.

“Marketplace Tech” host Kimberly Adams recently spoke with Silvio Tavares, president and CEO of VantageScore, about what he and his team consider when they’re designing their algorithms. The following is an edited transcript of their conversation.

Silvio Tavares: The reality is, establishing someone’s credit worthiness requires an enormous amount of judgment. It’s really important [to consider] the decisions around which data is factored in — and how. Let me give you an example of that. VantageScore was one of the first companies to come up with the idea of eliminating paid medical debt. Two-thirds of the collections accounts were related to medical debt. And when we looked at the data, what we found was that if you had actually paid off your medical debt, just the fact that you had a medical debt collection didn’t really indicate anything about credit worthiness. We also determined that a lot of those paid medical debts disproportionately hit people of color. So those are the key types of decisions that [as] modelers, we have to take into account. It’s about getting the best objective data, and looking empirically to see what types of data sets are best at empirically predicting whether someone is creditworthy or not.

Kimberly Adams: What are you doing at VantageScore to make sure that all of the data feeding into your algorithm and your model comes from consumers that have, in a knowledgeable way, agreed that you can have this information?

Tavares: We work with banks and lenders, and fintech lenders, to make sure that they are collecting data in a way that is transparent and where the consumer has provided their consent. But really most of that is going to be done by the bank, it’s not going to be done by VantageScore. We focus on the algorithm and making sure that it’s fair, it’s equitable and, most of all, that it’s predictive, and a really good predictor of whether the consumer is going to repay.

Adams: How can consumers be informed with what that consent actually means?

Tavares: We’re specifically not looking at, OK, what kinds of stores did you shop at? You know, did you buy big wheels for your car at a reputable dealer? We’re specifically not looking at that. We’re looking at general measures of your overall financial health and cash flow. But it’s also fair to say the reality is that the market is evolving very rapidly. The history of credit models is that they’re not based on this bank account-type data. But probably over the next 10 to 15 years, bank account data is going to be one of the primary sources for credit scoring, just because it’s timely, it’s often more accurate and it gives a much better picture of the consumers actual behavior. Many of the consumers who will benefit the most are the new type of consumers, the consumers that have been historically underserved, who are going to be able to access better credit terms, access more credit products, because of these new algorithms that we are developing.

Adams: What role does machine learning play in all of this?

Tavares: Machine learning is a general term, which is a subset of artificial intelligence. And it basically refers to looking at data — typically unstructured data — and then using that data to come up with a prediction. You know, machine learning is a super controversial topic. And it’s one that we really focus on quite a bit. But we really focus on analytics and using advanced analytics, all sorts of different analytics techniques, by the way, including machine learning. And we look at ways in which we can improve the predictive power of our scores. But also, importantly, include more people. So that’s why we were pioneers and including things like rental data, which is not a loan obligation, but it’s a really good predictor of what you’re going to do. So we believe algorithms are about making smart choices about how you use the data to include more people, but also, very importantly, understand how that data will actually predict effectively the consumers behavior.

Adams: Your speaking about rental data again reminded me of a question I meant to ask you earlier. As you’re trying to include this rental data, it can be complicated because some landlords don’t provide data on who’s renting from them or paying rent on time, consistently, especially when you’re talking about some of these more vulnerable groups like recent immigrants or people who are low income. What have you all been doing at VantageScore to try to obtain access to that type of consumer data for the most vulnerable people, who may be paying their rent in cash to a landlord who maybe doesn’t have the most incentive to report that?

Tavares: The first thing that we did is we were the first to incorporate that rental data when it’s furnished to the credit bureaus, right? The second key thing that we’re doing is looking at the other ways in which we could potentially get that data. Many consumers, particularly limited-income consumers, they get paid on payroll cards, which is a type of card or debit card. We’re looking at how can we access data from those providers and incorporate them into our algorithm. We’re also looking at bank account and debit account transactions. And if the consumer will give us permission to access those, then we can use those in our algorithms as well. The reality is that there are so many different types of data sources today, but there are more datasets than we have time to look at. And so it’s really just about prioritizing those so that we can incorporate those and get the best datasets with the consumer’s permission. That’s going to open up that really true view of creditworthiness.

Related links: More insight from Kimberly Adams

You can learn more about how VantageScore designed its latest credit score model, 4.0, around the concept of financial inclusion here.

VantageScore and FICO are considered to be the top two competitors in the credit scoring market. This explainer from Credit Karma lays out some of the key differences between the two.

But we should also note that FICO is changing, too. Its latest models, FICO Score XD and UltraFICO, “responsibly tap into alternative data sources to help lenders identify new credit-ready consumers and increase access to credit not only in the U.S. but globally.”

And we also want to point out that lenders get to select which scoring models they want to use. The ones we can see as consumers aren’t always the scores used to determine if you get a loan.

Clarification (July 13, 2022): This web story has been updated to clarify how FICO is updating its models.