Episode 203 – What Big Data Revealed About BNPL Consumers with Kristin Carlson and Zach Tondre, LexisNexis Risk Solutions

Yvette Bohanan

May 31, 2023

POF Podcast

A 2022 report from the CFPB titled Buy Now, Pay Later: Market trends and consumer impacts noted that from 2019 to 2021, the dollar volume of BNPL originations through five lenders surveyed grew by 1,092 percent, from $2 billion to $24.2 billion. BNPL adoption continues, not just in the US but globally, making it one of the most watched payment methods in the world.

BNPL’s rapid ascent has raised questions about its use: Who is choosing to use BNPL? Is it helpful or harmful to consumers’ credit and overall financial well-being? Does it improve financial inclusion? Is it displacing other forms of payment at checkout? A quick online search leads to mixed data points and inconsistent findings, often based on modest sample sizes and extrapolations.

To answer these questions, we sat down with Kristin Carlson, Global Products and Analytics Data Scientist, and Zach Tondre, Director of Credit Risk Market Planning at LexisNexis Risk Solutions, who conducted a longitudinal study to obtain insight into these and other questions.

Links: https://risk.lexisnexis.com/insights-resources/article/buy-now-pay-later-credit-risk

Yvette Bohanan:

Welcome to Payments On Fire, a podcast from Glenbrook Partners about the payments industry, how it works, and trends in its evolution.

Hello, I’m Yvette Bohanan, a partner at Glenbrook and your host for Payments on Fire. We have been discussing buy now, pay later, or BNPL for quite some time. Seemingly coming out of nowhere in the past few years, buy now, pay later, has captured the attention of consumers, merchants, investors, and regulators around the globe. As a reminder, BNPL comes in two primary flavors. The first is a specialized long-term financing, a multi-month line of credit for a specific, usually expensive purpose like home repair or a medical bill. Examples of companies offering this flavor would be CareCredit, GreenSky, and Wisetack. The second flavor is shorter-term interest-free financing or installments focusing on more modest purchases. These are often described as pay-in-3 or pay-in-4 offerings. Examples include PayPal’s Pay-in-3, aptly named, Afterpay and Klarna. This flavor is the version we’re focusing on today with our guests.

Glenbrook has covered BNPL on this podcast, in our education workshops and in conversations with the multinational merchants who attend our round table. And no matter the forum, there are always basic questions that everyone has been struggling to answer. Namely, who is using buy now, pay later? Is it good for consumers? Are BNPL users ditching other forms of payment in favor of buy now, pay later? All great questions. And today I’m talking with some people who have been looking at the data, quite a bit of data, on buy now, pay later, and it turns out they’ve uncovered some answers and a few surprises. So joining me from LexisNexis Risk Solutions are Kristin Carlson, Data Scientist on the Global Products and Analytics team. And Zach Tondre, Director of Credit Risk Market Planning. Kristen, Zach, welcome to Payments on Fire.

Zach Tondre:

Thanks so much.

Kristin Carlson:

Thank you so much for having us.

Yvette Bohanan:

I am looking forward to this conversation. You’ve done a lot of work, so I’m super excited to have you on the show. Let’s take a step back though for a second before we dive into that. For anyone listening who’s unfamiliar with LexisNexis Risk Solutions, can you just start with a brief description of how the company fits into the broader payments landscape and how your organization supports that mission?

Zach Tondre:

Sure. Yeah, so I think to summarize it briefly, LexisNexis Risk Solutions really mitigates kind of multiple different risk points around payments and lending and other financial services. And we do that through a variety of different means. And so everything from mitigating risk that revolves around device to identity resolution, document authentication, fraud and identity, some of the risk around compliance as well as credit risk. And so that kind of summarizes the areas that we focus and how our products, which are big data and analytics driven products, support the market.

Yvette Bohanan:

Great, thank you. So you’re all about risk, understanding risk, helping people minimize risk and building the right controls around it. So now let’s get into what we’re here to talk about. The insights that you’re sharing today have been quite an undertaking. You’ve analyzed, I would say significantly more buy now, pay later applications than any other study out there that I’ve been able to get my hands on. And you’ve also done a lot of work with those applications. What started you down this path? How did you get into this whole project?

Kristin Carlson:

That’s a great question. So just like you mentioned, with the growth of buy now, pay later, over the last few years, it really started with the pandemic that kicked off that incredible growth and it’s sustained since then. So when that growth occurred, we of course started getting many questions from our clients about how buy now, pay later might be eating into existing traditional credit trade lines and credit applications. They wanted to understand what a consumer looked like that was using buy now, pay later. And just a lot of questions around, if this is such a new and disruptive payment and lending platform, who is using it because we want to understand if that’s going to hurt our bottom line or that sort of thing.

So that was initially what kicked off this idea that, boy, we should be looking into this. And my team of data scientists actually grew out of the pandemic. We recognized this need to have a dedicated team of data scientists that could quickly pivot and look at macroeconomic trends, current events, how things like that could be impacting our data or our client’s data. And so that was the impetus of once we saw that buy now, pay later, growing out of the pandemic was such a hot topic, it really made sense for our team to take that, because we’re constantly exploring. We’ve got so much data at our fingertips and to have a team that can say, “Boy, we’ve never asked this question of our data, let’s try it out. Let’s see where it takes us. And if it turns up nothing, that’s fine. And if it turns into something really interesting, then we can run with it. We have the resources.” So that’s what we did with this study. It just started with a few client questions and we got the hint, we should be looking at it.

Yvette Bohanan:

Okay. And you had this tremendous data set to plow through. How did you approach the work? How did you think about structuring your research around those initial questions?

Kristin Carlson:

Yes. So the initial questions were honestly pretty easy. It was things like, how old are the consumers using buy now, pay later? Do they have experience with credit in the past and if so, are they subprime? I’ll just say that those are some of the assumptions that were floating out there in the industry. When we read up on buy now, pay later, there was a lot of positing about who could be using buy now, pay later. It looked to be like a more risky product because it doesn’t have hard credit checks, 0% interest, there’s low fees, and we would assume that it would attract this kind of risky profile of consumers. So those were pretty easy questions to start with. And so we started there by saying, “All right, let’s get some demographic features. Let’s look at their basic credit health.”

But the second part of this research was that we said, “All right, well we can’t just look at buy now, pay later users in a vacuum. We have to compare them to something. And so those groups should be banking applicants or retail card applicants.” So we’ve pulled comparison populations of a similar size, we’re talking about millions of applicants, during the same period so that we could get a really healthy group to compare and actually make insights out of. So that was the second part of the research where we then looked at longitudinally, over time, do we see changes in a person’s credit health or in their public record derogatory history? We have a lot of information on that. So we could get this broad picture of how a consumer looks when they start using buy now, pay later and a year later. And compare it to other consumers that were out there applying for more traditional credit products. So that was the second part of our research.

And then finally we said, “Well, wow, we have so many insights from this. Let’s see if we can model it. Let’s see if we can predict whether a consumer would choose buy now, pay later when given the opportunity of, ‘Boy, I need more credit. What am I going to choose?’ Can we predict if they would choose buy now, pay later, and subsequently, can we predict how often they end up using buy now, pay later?” So we had those three sections of our research, but we didn’t intend for that to be the outcome. It sort of grew out of what we… Each stage drove the next set of questions and how we could answer it.

Yvette Bohanan:

So there wasn’t some virtual whiteboard somewhere where you drew this whole arc out. It was just like starting with who are these people really?

Kristin Carlson:

Exactly. It started with just some basic questions, like I said, and we try to publish internal research monthly. So we sort of have these monthly goals to find something interesting in the population, but we really quickly had so much internal interest, it was very clear that we needed to continue the research. And that’s how we said, let’s ask more and more challenging questions and try to be one of the first groups out there to get a better understanding of buy now, pay later consumers.

Yvette Bohanan:

And you’re talking millions. I think one of the largest studies that I was able to find was in the thousands of records. So we’re talking exponentially higher numbers of records and longer time period, which gives it a lot more depth into what you’re able to answer.

Kristin Carlson:

Definitely.

Yvette Bohanan:

The time period, just to make sure everyone listening understands, you referenced this sort of surge in BNPL usage. It was over a thousand percent growth or something crazy. It’s still not the biggest retail product out there when it comes to payments acceptance, but it was growing so fast is what really captured everyone’s attention. What period are we talking about that you were pulling data for?

Kristin Carlson:

Yes, that’s a great question. So we pulled all of these consumers that had engaged somehow with buy now, pay later from October of 2020 through December of 2021. So we had 15 months of consumers that had been engaging in buy now, pay later, we could see their activity with it. But then, and we had to keep that timeframe because when we’re doing this analysis, we then aged that whole group 12 months. So we needed to have enough information through December of 2021, we needed to get it through December of 2022. And so that’s where we have essentially tracked 15 months of initial activity and then looked over time at those same consumers for 12 months. So I might mention that 12 and 15 month period again, but hopefully that clears it up from the get go.

Yvette Bohanan:

That’s great. That’s helpful, thanks. Okay, and the other thing I wanted to make sure everyone sort of understood before we stepped into some of your insights and your findings on all of this, as you worked through the data, sounds like certain cohorts emerged and groups that sort of had the same attributes of consumer and you gave them names. One was one time, light, occasional, regular, and super users. What are the general attributes of each of the cohorts? For example, how often do they use buy now, pay later? I’m guessing that’s maybe one attribute. Whether or not they’re credit scorable? How do you characterize each of those cohorts? Can you just sort of step us through that?

Kristin Carlson:

Yes, that’s a perfect question because I’ll be coming back to these profiles time and time again. We quickly realized with this research that a person using buy now, pay later just once or a couple times in that 15-month period is probably using it for pretty different purpose than someone that comes back 10 or 20 times in a 15-month period. So these five profiles one time, it’s pretty simple, just once in 15 months do we see them. Light was two to five times. Occasional, six to nine times. Regular, 10 to 19 times, and that kind of breaks that monthly or more than monthly. And then super users, which were 20 or more times, it was a very long tail where we did see 30 or 40 times in that 15-month period.

And so those five cohorts really were helpful because all of the data that we pulled and then said, “All right, let’s get the average credit score of each profile. Let’s look at how scorable they even are.” We actually found that on a whole variety of dimensions that these five cohorts really rank ordered with the measures that we were looking at. So what I mean by that is let’s say, let’s take score ability. We expected, our assumption about this group was that a person that ends up using buy now, pay later very regularly, that they’re relying on it maybe because they aren’t credit scorable. They haven’t used credit before or very infrequently, so they are a thin file. We don’t know that much about them. We thought that’s who would generally make up that regular and super user bunch. But we actually found that to be the opposite. The opposite case was true, that they were the most likely bunch to be credit scorable. So regular users of buy now, pay later are actually the most likely to have a history with credit in the past. And that really surprised us.

And not just that, but when we also looked at then credit score aspect, that also rank ordered. So the consumers using buy now, pay later, just once or a couple times, they had the highest credit scores on average starting at around 640. And finally when we get down to the occasional, regular and super users, they subsequently had lower and lower credit scores, all the way down to 606, was the average credit score of a super user. So on the whole, when we look at buy now, pay later consumers or users, they do have lower credit scores. It’s about 620 is the average credit score for a buy now, pay later user while it’s about 670 for our banking applicant population that we compared them to, so clearly a difference.

But the level of risk within buy now, pay later really breaks out when you look at how often they’re coming back to buy now, pay later. So these five cohorts were pretty important for us to have figured out right away, so we could constantly be comparing new dimensions against them and understanding where that risk is really derived from within buy now, pay later users as a whole.

Yvette Bohanan:

How old is the buy now, pay later user, and that sort of thing. We always talk about in our workshops that how we choose to buy is psychographic and demographic to some degree. Did you observe anything with generationally? Do you find super users who are 24 and super users who are 64 or do you see any kind of distinction or delineation there in the cohorts for any kind of dimension?

Kristin Carlson:

Yeah. Well, okay, for many dimensions that I can think of right off the top of my head, yes. We see when it came to percent highly leveraged on existing trade lines, rank orders, percent delinquent on an existing trade line, rank orders, other things like bankruptcy or evictions, all of that rank ordering with how often someone ends up using buy now, pay later. Just meaning they have some inherent risk and things that have gone wrong in their past. And that’s likely one of the drivers of where they’re continuing to come back to buy now, pay later. When it comes to age, I don’t have that in front of me in terms of how the average age, let’s say, shook out for each of the cohorts, but in general, just as a whole, buy now, pay later users are younger on average than banking and retail card applicants.

However, I’ll hopefully get to this in a little bit, but some of the surprise definitely came when we looked at prime consumers that use buy, now pay later frequently. A prime consumer is more likely to be slightly older, and there’s certainly a large enough percentage of buy now, pay later users that are over the age of 35. 47% to be exact, are over the age of 35. And so we can see that we can’t just classify buy now, pay later users as, oh, they’re all young. That’s definitely not the case. They’re on average younger, but they’re not all young by any means.

Yvette Bohanan:

Right. And I think when we have a video, we play a lot of times with Max Levchin introducing his idea of launching Affirm. And the whole premise was, “This is for all of those people who are in their twenties who are allergic to credit cards and who are going to someday have to buy a mattress or something that they can’t put on just on their debit card.” And that was the whole thesis of Affirm. And what you’re saying is, yeah, maybe that was true, but that’s not the whole picture here.

Kristin Carlson:

I think it’s got appeal for a much wider swath of the consumer population than what he originally set out to solve for.

Yvette Bohanan:

The other thing that was an interesting observation when I was looking through the material was that there are consumers in depressed geographies using buy now, pay later twice as often as consumers in more affluent geographies, and that sort of separation there. And it sort of brings to light the question of is buy now, pay later helpful or hurtful when it comes to financial inclusion? And I know Zach, you’re pretty passionate about this topic.

Zach Tondre:

When we think about financial inclusion, I think financial inclusion can mean different things depending on the context that we’re talking about. And so when we look at financial inclusion, we’re really just thinking about what it takes for a consumer to get a new account or a line of credit through a variety of different digital avenues. And so really at the end of the day, there needs to be enough data to kind of in an automated fashion, verify a consumer’s identity, score them for credit risk, make sure that they pass regulatory compliance checks. All of those things have to happen in order for a consumer to go from being an invisible into being included in the mainstream financial services.

I mean, when we think kind of back, I guess even a couple decades back, where we looked at consumers that needed loans that were in depressed geographies to just be kind of simple about it, they would possibly go down to the neighborhood loan shark at the corner, right? Because that’s where they could get credit. They had a relationship with that person. Obviously the terms and rates were pretty terrible. Anywhere where there’s kind of this big disparity where there’s kind of some room to maybe offer a product that’s more competitive, more fair, a lot of different lenders started stepping into this space.

And a lot of times that may start off with a shorter term loan with a fintech lender, but then eventually now they’re going to establish credit and they’re going to become more and more substantial kind of in the credit universe and they’re going to become more included in more prime products. And so there’s this kind of journey of the consumer emergence to be able to get into these prime products. And so I think with BNPL, it’s really going to depend. I think it offers a great avenue for increased financial inclusion, but there has to be this kind of downstream reporting. There has to be something that allows the consumer to get some positive credit history out of the transaction for that to be able to translate to establishing credit for larger installment loans or home loans or car loans, whatever that might be.

And so I think it’s really going to depend. We haven’t really gotten to that part of the story yet in the US of where those BNPL transactions are going to land as far as credit reporting goes. And so it’ll be interesting, I think time will tell, but even the consumers kind of having this activity and going out and seeking these BNPL transactions that does establish them in kind of a digital identity, it starts down the road, but it doesn’t quite get us all the way there. And I think that a lot of that story will be told in the coming months.

Yvette Bohanan:

Kristin, you’re shaking your head yes. What did the data say to you? What did you see coming out of the data that sort of amplifies what Zach’s mentioning here?

Kristin Carlson:

Yes. Well, I see this question in at two angles, which is this initial, if you don’t have access to credit and buy now, pay later presents itself, it is clearly a place that you can get credit and that is technically financially inclusive because you’re able to get credit. But when it comes to, like Zach mentioned, I mean there isn’t, the national credit reporting agencies don’t track buy now, pay later products. So they’re not able to say to another lender, “Hey, this consumer has several buy now, pay later loans out. They might not be able to pay off this additional one that they just inquired about.” And so that feedback loop of either a consumer positively making payments, it’s a more simple platform to understand for some consumers.

The fact that it’s not having that positive feedback cycle, we actually saw that in our data where we looked at credit score ability among the banking, retail and buy now pay later populations. And we’ve looked at it over time. What we would expect is that for the consumers that applied, they went out and applied for a traditional credit product that’s being tracked by the national credit reporting agencies. Well over time they become scorable at a faster rate than buy, now pay later users who are at about half. So it’s about eight to 4% is the difference, where 4% over time saw an improvement or being able to be credit scorable in that 12 month period. But it showed up very clearly that because of the lack of reporting that’s happening, consumers that are using buy now, pay later, they’re only getting that improvement because they likely took out another traditional credit product at the same time. And so that allowed them to become more credit visible, but buy now, pay later isn’t doing that. So that angle of being financially inclusive is not living out to the potential that it has.

Zach Tondre:

I think one other thing that I would add to that too is just that where it does get a little bit tricky is it may not even be as simple as just reporting the trade lines. Because a BNPL transaction just intrinsically, it’s just different than a lot of the lending products that are out there. Just with the term that happens, the dollar amounts. And so I think until lenders really get a handle, even once we’ve got the data and we’ve got the trade lines, lenders have to figure out how that translates to their credit risk, because they’re not going to treat it like any other installment loan or credit card product that’s sitting out on the consumer’s credit report. They’ve got to figure out what does this all mean? How much additional additional risk does it add? Maybe there’s some positive pieces to it, maybe there’s some negative, if someone doesn’t pay their buy now, pay later installment loan back, how big of a detriment is that to their credit scores? There’s just a lot to be figured out and.

I know that as the bureaus are talking about how they’re going to start bringing this data in, the idea is that even once they start getting trade lines, they’re going to kind of hold it off to the side to figure out what it all means before it starts getting affecting consumers kind of mainstream credit scores.

Yvette Bohanan:

Yeah, it’s a really good point. You buy some high end air fryer or something and then all of a sudden, you miss a payment and that’s affecting your score at the same level of missing your mortgage or something. It just, there’s some relativity to the whole thing. Yeah. The other thing that you observed was that for the folks who were credit scorable, their credit scores improved when using buy now, pay later. Everything isn’t always as it seems. Tell me a little bit about what you found out when you dug into that.

Kristin Carlson:

Yes, so we looked at the buy now, pay later users as a whole and like I said, we did that longitudinal study, tracking them over 12 months to see how their financial history changed after they started using buy now, pay later. And we actually saw that as a group buy now, pay later users increased their credit scores by about five to seven points on average. And in comparison to the banking and retail applicants, they also improved their credit scores during that same period by about two to six points. So as a whole there was something perhaps going on macro economically that supported credit scores to improve during that time. But what’s really interesting is that when we broke out those improvements by the cohorts, the five cohorts that we talked about earlier, you could clearly see that this is another area that things rank ordered.

So the consumers that used buy now, pay later, very rarely, well, they saw the least improvement to their credit scores over time. But the consumer that came back, they were using it maybe monthly, maybe more than monthly, their credit scores increased on average by about seven points, the max of that group. And when we said, “All right, that’s really interesting, it follows one-to-one with how often someone’s using buy now, pay later and how much their credit score improves over time.” We then said, “Okay, let’s look at other aspects of their trade line help.” So we pulled in bank card utilization. So if a consumer had an existing bank card trade line, we could say, “What happened to this during the same period?” And we actually saw that utilization on those bank card accounts declined by one to 2% for all of the buy now, pay later cohorts. But during that same period bank card and retail card applicants saw increases to those bank card trade lines and utilization.

So that divergence of outcomes in the same time period showed us that what’s happening most likely is a consumer that’s using buy now, pay later very frequently, they’re putting enough of their purchases, they’re shifting them away from an existing typical credit card and they’re putting it on a buy now, pay later product. And that product isn’t tracked by the national credit reporting agencies. So what happens is we see less of their debts and we see less debt on those bank card trade lines. It improves the viewpoint from how much they’re utilizing and that causes that increase to their credit score, because we just can’t see that they actually have more debt than what we can see. So that was one of the really interesting aspects of looking at credit scores over time and why that played such an important role to be doing a longitudinal study.

I also have a couple of other anecdotes about this related too. There were two groups that we got really interested in. We looked at the entire buy now, pay later population, and we said, “All right, well regular users are the really interesting group. And on top of that, subprime regular users and prime regular users, they probably have very different purposes for being a regular buy now, pay later user.” So that was where things got really interesting because we were pulling out this particular subset of consumers where let’s say you have subprime consumer, our expectation is that they’re turning to buy now, pay later frequently, because they don’t have many other good options for affordable credit, it makes a lot of sense to keep their lifestyle going, keep the everyday expenses going through buy now, pay later. And that potentially they’re overusing it and we might see some decline in credit help over time.

That was our expectation. But we ended up seeing was that over time, those subprime regular users, they saw a 12 point credit score increase in 12 months. And it didn’t just come from changing their shopping habits from a traditional credit card and moving it over to this buy now, pay later product. But we actually saw declines in delinquent accounts and in collection inquiries. So that was really interesting because it kind of said there might be some consumers that aren’t doing so well with maybe they are putting a lot of purchases on buy now, pay later, and we can’t tell. But on the whole, some are using this product, potentially, to improve their standing and to pay down maybe existing trade lines that have that revolving higher interest rate. Because what we saw was that those negative factors were declining and they were looking on the whole more healthy. In a lot of factors, not just on utilization.

Yvette Bohanan:

That’s really interesting because what it’s saying is people who… You’re disabusing us of a lot of myths here, longstanding beliefs, right? You’re saying if you’re… The myth would be like you’re subprime, so you don’t know how to manage your finances and you got yourself into trouble or whatever. And the reality is you are trying to make ends meet and now you have this other tool at your disposal to sort of offset things, help pay down this, not add to a higher interest rate for a small discreet purchase. And people are actually, instead of saying they don’t know how to manage their finances, they’re doing some very savvy things over here with financial management, cash flow management, if you will, at the personal level, to improve things. Is 12 points over 12 months, that seems like a lot to me.

Kristin Carlson:

Absolutely. I mean of course there can be big swings in credit scores, but when we’re talking about so many consumers in our study sample, we’re talking about hundreds of thousands that fall into that group. 27% of all buy now, pay later users fall into the subprime and regular buy now, pay later group. And that as a whole, if they’re moving that much, sure there could be consumers in there that aren’t doing so well and we can’t tell that and they are taking out too much. That could be buried in there. But on average on the whole, there’s enough consumers in that group that are actually probably finding ways to manage their money better through buy now, pay later, just because of how simple it is to set up and the fact that it’s 0% interest. That’s a huge way to get ahead if you are a revolving card person.

Yvette Bohanan:

Oh, yeah, it’s very attractive. Did you have other anecdotes for that particular cohort? Or you said you had subprime and prime in the regular space.

Kristin Carlson:

That’s right. So the second story with all of this was, and for me the most interesting, was trying to figure out, why would a prime consumer go to buy now, pay later very regularly? Remember that’s 10 or more uses in a 15-month period. I thought that was really interesting. And what we ended up finding was that frequent buy now pay, later use by that group, the prime consumers, which were 10%. So 10% of all users fall into the prime and regular user group. Well that’s actually a leading indicator of financial distress. We ended up seeing that over time that group had a decline in their credit score by on average five points. And at the same time we saw pretty big increases to the number of accounts that were 30 days delinquent, the number of collections inquiries. And then when we looked at different credit channels, so bank card utilization, retail card utilization, installment loans, all of those showed increases, which is not a good sign.

So there were all these factors, again, the same ones that we had pulled for that subprime group, but the prime group showed much worse over the same period. So it was pretty clear that buy now, pay later is kind of either a rescue option or it’s a prime consumer, but something must have happened in their life and they’re trying to use buy now, pay later as sort of a last stop in getting more affordable credit and digging themselves out. Again, this is on the whole and there could be consumers using it for different purposes. But we have to look at that macro picture and what it’s telling us is that you have a group of credit prime consumers, but something is going wrong and buy now, pay later is sort of a bellwether of what’s happening in their financial life.

Yvette Bohanan:

That’s interesting. So let’s go back for a second. So there’s this notion of improving credit score, but is there a way that you can actually measure if ultimately this payment method, buy now, pay later has positive outcomes in financial stability? Does it help people not just get a better score but get to where they want to be in life? Do they get greater financial stability through this?

Kristin Carlson:

Yeah, that’s a great question. So the way that we could get an understanding of that aspect was through our public record history and all the data that we pull in for that. And so we looked at evictions, bankruptcies, liens, subprime credit inquiries. We got a sense of… We wanted to see if buy now, pay later, was that a financial intervention in someone’s life that could sort of change the outcomes of these really negative events that can happen in someone’s life? And we looked at the rate of these events happening in buy now, pay later users compared to our benchmark populations, the banking and retail card applicants. And in general we saw that before using buy now, pay later, after using buy now, pay later, the consumers that use buy now, pay later are inherently more risky. And they do have two to three times the rate of bankruptcies in their past, collections inquiries and then more events with evictions and liens and other such things that we’ve been tracking.

So that was one of the ways that we tried to see, well yeah, does buy now, pay later step in and help them stabilize in a really substantial way? But from what we’ve seen, I can say that it’s not a big enough impact to really change their trajectory. I will mention that when we talk about those factors like evictions or liens, those are pretty rare events. So we’re talking about a huge number of people and we’re just comparing the rate between banking and retail applicants compared to buy now, pay later. The rate is where we get those differences of two to three times. But ultimately very few consumers have any of those risky factors in their history. And it really just comes down to maybe they’ve mishandled some credit in the past and it put them in a subprime credit tier, but they don’t necessarily, they haven’t been evicted or something else.

So the whole population does have more risk. But I always like to point out that when we looked at measures of stability over time, that buy now, pay later users are on the whole, most of them don’t have those really risky signals. It’s just the rate when we compare the two of them. I hope that made sense. But yeah, we’ve definitely looked at how buy now, pay later might be a sort of an intervention in somebody’s financial life and it just really didn’t seem to be that much of a difference, at least over a six-month period. Perhaps if we looked out farther and we continue to do this work, we might see that the adopters of buy now, pay later that are frequently using it, finding it to be an easier tool for them, perhaps it does stabilize them in some way. So that can certainly be future research.

Yvette Bohanan:

Let’s see. My head is spinning a little bit, when you start referencing the additional data you brought in from public records like this data set, this monster data set, in trying to definitively link up all of these records. Wow. So I’m really curious about what you said where this actually led you, which was trying to build a model and can you become predictive, right? And it makes perfect sense that you kind of led towards that just because of the way this data starts falling out, but also the work you guys do just in general at LexisNexis Risk Solutions, make sense that you’d want to get predictive here somehow, or ask, can we be?

So can you predict? I’m going to put my merchant hat on for a second. Who is likely to actually choose buy now, pay later? And then what kind of customer am I attracting, right? If I offer this, what’s happening? And honestly, there’s typically no risk to the merchant here, just to be clear for offering this. All the risk is onto the provider, whoever that is. Can you get predictive? Can you know who’s going to choose it? When I go to the checkout page, can you tell if I’m about to click one of those buttons?

Kristin Carlson:

Absolutely. We built some predictive models to get at that, whether you would choose buy now, pay later, or a traditional credit product. And while we focused on how accurate can we make that model, how well can we predict that outcome? What we were really interested in was understanding what are the driving factors in that model that distinguish between a particular consumer choosing buy now, pay later or not. So those factors were, while the accuracy and the prediction power, that’s something that we can continue to work on. Models are a long… You have to really build them out for a long time to get the kind of accuracy that you want. But on first pass we were able to get something that was predictive enough that we felt we could look at the driving factors and I’ll tell you what those are.

The first one was age. So you would’ve guessed this probably that age is going to be a huge part of a consumer choosing buy now, pay later. They’re more likely to be young. And so age is certainly a factor. I mentioned that it’s sort of this 53, 47 split by age 35 where you have 53% younger than age 35, 47% older. And that fact alone, just how young on average buy now, pay later users are sets them apart from banking and retail card applicants. They are older. So on average it makes sense. But age is really, like I said, you can’t chalk everyone up to being young. They’re also in the whole age spectrum. So the first aspect was age. We also found some aspects of some transiency in how often they move and some aspects related to the socioeconomic background. But one of the more interesting factors to us was their inquiry history.

So we track when a consumer inquires with a particular credit product and the type of product that they inquire about is what is another distinguishing factor. So we found that alternative lending or basically non-bank lending, so places like online that offer personal loans and they don’t have a brick and mortar location, but there are plenty of them, right? All these fintech startups. When a consumer makes an application for a fintech loan, that is a pretty big deciding factor that they’re also going to have an appetite for buy now, pay later. So in the two years leading up to their first time using buy now, pay later, they’re 72% more likely to make two or more of those inquiry types compared to our benchmark populations. So that was so interesting, I didn’t expect that from when we set up the modeling, but that was one of the top features that really distinguished between groups.

So that was one. And then we also set up a model that looked at whether we could predict if a person would become a super user or a regular user of buy now, pay later. That was our secondary model and we had different factors that came out, but I think you would also almost be able to guess them because they make so much sense. The top factor for predicting if someone was regularly going to use buy now, pay later was credit score. And it makes so much sense because the lower your credit score, we saw how it rank ordered with the usage profile of a consumer. If you have really unaffordable credit being offered to you and then you’ve got this buy now, pay later option, there’s no hard credit checks, 0% interest, you’re increasingly likely to choose buy now, pay later with lower and lower credit scores. So that was very clear.

And then secondly, those inquiry types, again, if someone’s making fintech lending, personal loan inquiries or also retail card inquiries, those were the top, second and third most impactful differences between whether they would just barely use buy now, pay later, or really take to buy now, pay later. So these models helped us say, “Yeah, we’ve clearly got a lot of insights here, let’s see if we can put them to the test. How do they shake out when we try to predict?” And on a first pass model bill, we did pretty darn well for being able to predict whether someone would use buy now, pay later, and how often.

Yvette Bohanan:

That’s really cool. Can we turn the question inside out real quick and say, can the fact that you’re a buy now, pay later user be predictive as an indicator for use in other credit decisioning? For other products, not for buy now, pay later, but does it feed anything reliably, do you think?

Zach Tondre:

Yeah, I mean I can take a stab at that too. Just I like what Kristin was, when she was just alluding to the inquiry traffic and how predictive just a simple inquiry for a type of loan can be. And I think the same thing goes in reverse, right? I mean if we’re seeing a certain velocity of inquiries, and so if an online lender is looking at a velocity of BNPL inquiries or inquiries around a variety of different lenders, that’s pretty indicative of credit risk. ‘Cause it just kind of shows you this state of how desperate a consumer might be for funds at that point in time. I mean, so really high levels of inquiries within a short time period are big red flags when it comes to credit risk, but there’s a lot of other factors in there. When we think about credit decisioning, we really think about first, how stable is this consumer? And then what is their ability to repay the loan and their willingness to repay the loan?

And so all of those things end up factoring in and while just the simple inquiries on a BNPL could be an indicator of credit risk, the other interesting things that Kristin’s talked about, kind of this artificial inflation of scores as the total available credit goes down or yeah, goes down, things like that. There’s going to be these compensating factors. It’s like now you’ve got this traditional credit score that’s going up and maybe this alternative credit score, that’s the only one that has visibility into the BNPL transactions, it’s going down. And so do the two wash out? Has that consumer’s ability to repay or willingness to repay really changed? Or have they just switched their appetite for a different type of product? And I think that one of the things that we think about all the time, and going back to the idea of what is this doing to a consumer’s financial wellbeing, I would say, well, it kind of depends, right?

It’s like anything else. Like medicine, it can be really good when it’s used in the right doses, but can be really bad when it’s abused. And with BNPL transactions, it’s so interesting because the way that I look at it is, it’s kind of like that it’s item that sits right by the checkout stand at the grocery store that you weren’t shopping for, but you go to checkout and it’s just right there. And it’s like, “Oh, you know what? I would like to push this payment out over six weeks and so I’m going to go ahead and take it.” And so it’s placed kind of in the right place, but also I think that it does take some more thoughtfulness on the consumer’s part because they’re deciding purchase by purchase, what they want to push out into the future versus their credit card where they’ve got this whole available balance and they don’t have to do a lot of thought to just hammer that whole balance to max in a quick period of time. Whereas the BNPL transactions are kind of this slow steady burn of eating into this credit.

And so I think it really is going to depend, but absolutely there’s factors I think that will correlate between traditional lending products and BNPLs. But I think it’s pretty complicated at this point in time. And until we’re able to actually start seeing kind of outcome data of how these super users end up performing in other ways, in real life situations where lenders can start building new models that incorporate this data, we won’t know exactly what that’s going to mean for the credit scoring market as a whole.

Yvette Bohanan:

That’s a good point. It’s still early days. It is still very early days, especially when you look at the market share these, but it’s such a rapid growth trajectory. Going back to this comment about rates of things and it’s the rates that are capturing everyone’s attention right now, the usage. Okay, can we go for two more minutes because I have two more questions and if I don’t ask at least one of these, I’m going to hear about it from some listeners. So Russ Jones always says in our workshops, he’ll say, “Payment methods, payment systems, payment networks are not necessarily all friends. They like to compete with each other. They’re like animals in the jungle, they’re going to eat each other, they’re going after volume.” And so I have to ask, do you you see at larger in any particular cohort that buy now, pay later is causing a volume shift away from other forms of payments in a very deliberate way, in a meaningful, statistically meaningful way? Did you see that?

Kristin Carlson:

Well from the research that we’ve done, which we’ve definitely looked at this question because it is a burning question of so many. We haven’t seen a substantial or meaningful shift away from retail cards, which is something that we were able to, we pulled in retail card applications as a good benchmark for understanding whether, is buy now, pay later disruptive to a very similar form of credit? Getting a accountable item at a store, that would be a great comparison for a retail card, right? So what we did was look at retail card applications over several years and we looked at the volume and we compared it to the volume of new buy now, pay later applicants. So someone that’s never applied before and we just looked month over month, what are we seeing when we compare retail card applications to newcomers to buy now, pay later? And we used that sort of ratio to see the points in time where buy now, pay later might have been eating into the retail card space.

But what we found was that only during the few months in the pandemic when we had a lockdown, so between March of 2020 and June of 2020 is when we see a huge switch between from retail card applications all going into buy now, pay later. That’s really when things took off because consumers weren’t in stores and they were online and they were shopping and they saw that really easy button to click to split their cart into four easy payments. So when we looked at that time series, we saw that huge spike, but immediately fell right after June of 2020. Essentially things went back to normal for retail card applications, meaning we’re not seeing some kind of sustained disruption in the volume of credit card applications coming through. The rate of growth in that industry has remained steady. We’re not seeing some kind of decline. And since those are two really comparative populations to be looking at, if buy now, pay later isn’t eating into retail cards, well then we can kind of assume that they’re not eating into a traditional bank card product or something of the like.

So, so far the places that we’ve seen buy now, pay later growing, it’s not so much by taking business away from other forms of credit. But it’s actually that we see fewer and fewer newcomers to buy now, pay later, especially over the last few months. During the Black Friday shopping season for 2022, we looked at this all again to say, “All right, where is the growth in buy now, pay later continuing to come from? Is it newcomers or is it repeat users?” And it’s very clearly repeat users. They are the ones driving the adoption and the continued growth on these platforms. And even if you look on those particular buy now, pay later provider websites, they’re really doing the kinds of things that would show they want to keep you using their platform. So it makes sense. Everything lines up from our data standpoint and just from the vetting that I’ve done looking online at how these platforms are positioning themselves. But that’s certainly what we’re seeing is not a massive shift from a traditional form of credit into buy now, pay later. That’s not the case.

Yvette Bohanan:

Oh, that’s going to be good news to a lot of listeners here and not so good news to some others, but they might already know it. That’s kind of interesting though. That brings me to my final question. What questions would you like to answer next? And I guess both of you, you probably have different questions in mind, but what do you want to go after next here?

Kristin Carlson:

Yeah, my burning question on all of this is to understand more about the performance of these buy now, pay later loans, meaning how often are people paying them off? And what are the size of the buy now, pay later loans? We don’t have that data and it would be really informative to know if someone isn’t paying off their exercise equipment, that might be a different case than if someone isn’t paying off their pizza that they purchased with buy now, pay later. So those are very different ways of understanding all the nuance in this group.

I think the summary in a lot of the research is that you can’t chalk up the buy now, pay later group as, oh, they’re all young and subprime and inexperienced with credit. No, we’ve totally debunked that, some of these myths and actually found that no people spread the age range. They are very experienced with credit, they really need credit. And some are finding buy now, pay later to be potentially very useful for getting ahead on things that are costing them more to run for credit that they have. So I’m really interested though to understand are people using buy now, pay later more in store as that happens? Is it more attractive for a merchant to offer buy now, pay later in store and compete more with online? And just understanding the products that people are buying. There’s so many questions with this research and I’d love to know more about what’s really driving the adoption for the groups that we’ve talked about today.

Yvette Bohanan:

Zach, what do you think?

Zach Tondre:

Yeah, I mean think that my kind of interests are really driven off of what our customers are asking us. And so one piece of it would be a part that we already covered off on quite a bit, but what does BNPL usage by a consumer really mean to the credit worthiness of that consumer? And so how do the two translate? Is as $1,000 worth of outstanding BNPL payments the same as $1,000 in a credit card balance? Is it different? Is it more risky? Is it less risky? I don’t know. So I think our customers are wanting to understand that.

And then when we think about, Kristin mentioned that we haven’t seen this decline in card applications, either store card or retail card. But we know that every single transaction that takes place through a BNPL provider is a transaction that didn’t take place through that card. And so while there’s no fewer card accounts, that number of transactions has kind of dipped. And so from another place that we support our customers is on the marketing side. And so we want to understand as well, what does it take to drive that consumer to one payment method or another when they actually get out to get to checkout? And how can we inform that? How can we help our customers kind of figure out how to drive consumers towards one payment method or another? And so I think that’s another question that would be interesting to answer just because our customers are interested to know the answer.

Yvette Bohanan:

There are a lot of people. And before buy now, pay later, there were a lot of people that wanted to know the answer to that question. So if you can answer that question, please come back on a podcast because we all want to hear the answer to that question. Oh, you guys have been so, so helpful and so generous with your time today. If people want to hear or learn more about this body of work that you’re doing, where do we point them? Where on? Is it your website?

Kristin Carlson:

That’s a great question. Yes. So we are producing a number of research briefs, a series of five. We’re on research brief three, we have two of them published and they are on our website. I can certainly share the links if you have this in the show notes. So, happy to share those. And we’re also doing some external webinars, so anyone can sign up and listen in, get a synopsis with visuals to see more of the data. And of course check out those research briefs. I think it’ll be really helpful. And we’re trying to get the word out there about debunking some of these myths.

Yvette Bohanan:

Fantastic. We will make sure that we put the links into the show notes. And it’s that special time now where we have to wrap things up. So thank you so much for spending time with us on this episode and for all of the hard work that you’ve put into this research, it’s really going to benefit I think, and inform, a lot of people across the payments industry and hopefully regulators and everyone else looking at this problem to give us better insights into making decisions about this.

And to all of you listening, if you’re curious about buy now, pay later, we’ve got podcasts on this, you have this body of work with LexisNexis Risk Solutions, we have some PV posts, we have on demand modules, there’s a lot out there. So please check it out and look for more resources, both at LexisNexis Risk Solutions and glenbrook.com. And as always, thanks for joining us and until next time, keep up the good work. Bye for now.

If you enjoy Payments on Fire, someone else might too. So please feel free to share this podcast on your favorite social media outlet. Payments on Fire is a production of Glenbrook Partners. Glenbrook is a leading global consulting and education firm to the payments industry. Learn more and connect with us by visiting our website at glenbrook.com. All opinions expressed on our podcast are those of our hosts and guests. While companies featured or mentioned on our show may be clients of Glenbrook, Glenbrook receives no compensation for podcasts. No mention of any company or specific offering should be construed as an endorsement of that company’s products or services.

Recent Payment Views

Demystifying Payments Orchestration: Part 1

Demystifying Payments Orchestration: Part 1

Introduction For those of us who follow the payments industry, it has been difficult in the past few years to avoid the hype around payments orchestration.  The ubiquity of the term begs the question: what exactly does payments orchestration mean?  We see...

read more

Glenbrook Payments Boot CampTM

Register for the next Glenbrook Payments Boot CampTM

An intensive and comprehensive overview of the payments industry.

Train your Team

Customized, private Payments Boot CampsTM workshops tailored to meet your team’s unique needs.

OnDemand Modules

Recorded, one-hour videos covering a broad array of payments concepts.

GlenbrookTM Company Press

Comprehensive books that detail the systems and innovations shaping the payments industry.

Launch, improve & grow your payments business