Affirm’s Max Levchin: Why ‘Buy Now, Pay Later’ Beats Credit Cards

0:04As much as we all can laugh and mock the, oh gosh, they're financing burritos, when you put your credit card down at your favorite burrito, you are financing that burrito too. You may be personally paying it off at the end of the month, but half the country is paying interest to their credit card provider. That burrito gets right on top of that pile and immediately starts accruing interest. The sort of, oh gosh, a new way to borrow money. Let's stick with the old one because it's better is pretty disingenuous. This week on Bold Names, serial entrepreneur Max Levchin. You might know him as one of the co-founders of the company that became PayPal.

0:45And for the past decade or so, he's been at the company Affirm. Max started the Buy Now, Pay Later company to try to make credit more transparent and therefore less risky. From the Wall Street Journal, I'm Tim Higgins. And this is Bold Names, where you'll hear from the leaders of the bold name companies featured in the Wall Street Journal. Today we ask, what will the future of credit look like?

1:13Max, thank you for joining us on Bold Names. You have been in the fintech space for most of your adult life. You've been heading up a firm since its creation in 2012, really becoming one of the darlings in the buy now, pay later space. That buy now, pay later, it's not necessarily a new idea, right? Installment payments were really made popular in the Great Depression and kind of lost favor as we saw the rise of the credit cards years later. I'm curious, how did you think you could innovate in this space and what was motivating you? Why go into that area? Two personal events actually shaped the foundation for the founding of the company.

1:54So I came to the US as a teenager from what was Soviet Union. And as a semi-independent adult, I tried to buy a car and was declined, even though this happened right after we took my first successful startup, a.k.a. PayPal, public. And so I was independently wealthy. You had a lot of money, millions of dollars. Very successful. I did well. I did all right. And yet an auto dealer in L.A. told me, your credit score is so low, I'm not going to sell you this car unless you pay cash. which is what I did. But the reason I had such a low credit score wasn't entirely just due to the fact that I was a recent immigrant and had no history.

2:38I'd managed to get myself in real trouble getting a credit card in college, not fully understanding how it worked, missing minimum payments, getting into a 90-day delinquency, and basically just completely wrecking my credit history and score. I'd sort of become more and more motivated to build a product that 18-year-old fresh-off-the-boat immigrant Max would understand and actually feel good about using. That's where our farm came from. At the time, this wasn't a big idea, but now, because a lot of the work you all have done, buy now, pay later is becoming a force out there. The market size projected for this year is around half a trillion dollars, according to Deloitte.

3:19They think it could be a trillion dollars, close to a trillion dollars by 2030. You've got more than 24 million users. How do those transactions work? It's pretty simple. Every transaction for a firm always begins with an active thought of, I'm going to buy a thing and I'd like to borrow money to pay for it. With a firm, before you consummate the transaction, you actually see the payment schedule. You understand what it will cost you. You confirm that you do plan to use it and then the transaction completes. You get the thing that you were buying and then we will bill you typically 30 days after this event.

4:00What's the typical transaction size and what's the typical thing people are buying? The average transaction size for us is about$270 plus or minus and it's just sort of trended down as people have been using the product in more and more places. Obviously, the average transaction out there is much, much lower. We're not especially commonly used for things like lunches. That's one of the big criticisms, right? That people are buying burritos with these things. I do not believe Affirm is used to finance burritos. We are primarily used to finance things like bicycles, couches, electronics. I mean, it's easy to use.

4:36I bought a couch a few years ago using Affirm. The idea of using cash for such a purchase really It kind of hurt my head, but ignoring my wife's desire to have a new couch probably would have meant that I was sleeping on the old one. So I turned to a firm. But really the big draw was the direct-to-consumer seller, which was kind of emerging about seven or eight years ago, becoming very popular, had this very neat offer, right? It was zero interest. I think it was over a year's time, and I could see the payments every month. I think it even came out automatically out of my debit card or my checking account.

5:14What was the pitch to merchants early on to make this work? Why were they willing to kind of go into business with you? I think you almost reproduced a pitch just from your personal history. So you want to sell it as a merchant to this consumer who is hesitant to put their credit card down for whatever reason. You know, maybe it's the 18-year-old Max confused about how credit cards work. Maybe it's a much more experienced consumer who says, look, I don't love revolving. Revolving is expensive, and I have no idea when I'll be out of debt. Revolving means they don't pay it off at the end. It keeps going, turning over month to month.

5:49They accrue interest upon interest upon interest. Interest, of course, compounds into principle, and on it goes. But I do need the couch. Offering a payment plan that is super simple, super transparent, the costs are spelled out, cannot change is pretty attractive. Merchants would place those offers on their site and promptly see increments of 20, 30 % more sales. And so the product sold itself. The experience you had, the 0 % one, is especially compelling. What really, obviously, no one invented sort of free money in the future. What's really going on here is the merchant said, if you paid interest, it would cost you this many dollars.

6:29How about I use that as an incentive for you to transact? I'm not going to give you a 10 % discount or 15 % discount, sort of a standard retail enticement. Instead, I will turn those dollars into essentially making that deal interest-free. Part of what we were hitting on, what we hit on 12, 13, 15 years ago when we started thinking about this product was this distaste for 0 % but an asterisk. This idea that 0 % in consumer finance is definitely not a 0 % was a huge motivator. But I really wanted to prove to the world that there is a way to build a product where what you see is what you get. And so both merchants and consumers really latched onto this idea of honest financial products.

7:14And we were there. So how do you make your money? Exactly as it sounds. in the world of merchants paying your interest, they will pay us slightly more and we get to keep that. In a world of consumers paying interest, we will tell you exactly how much of that interest we get. It's not free to run this business. We are not a bank. So we have to borrow money or finance our loan book somewhere. So there's cost to that. There's cost of servicing. Obviously, not everybody is able to pay their bills perfectly on time. We don't charge late fees, which means we lose money, which, you know, by the way, That aligns us perfectly with our consumers.

7:51We don't want to align money if we suspect there's a chance for you being late or not being able to pay you back. But bad things do happen to good people sometimes and so on. But it's a very simple business. Every loan has to be profitable. That means we have to be good at underwriting. That means we have to be honest and tell folks that should not be borrowing money right now. Hey, you shouldn't borrow. We're not going to be willing to give you a loan. And if on average, every loan you make ultimately does pay you back with a little extra, it's just a matter of time until the extras add up to more than your fixed cost, at which point the company becomes profitable and you can adjust it for gap principles.

8:28So things like stock-based compensation and amortization of various non-cash expenses need to be folded in. But at some point you hit the scale where it is profitable. And that, you know, it also kind of requires you to be pretty good at math if you're running a lending company. And so we are naturally pretty good forecasting. A key part of that math problem is your gross merchandise volume. I think that's a technical term. It's really the total dollar amount of transactions process on the platform has been growing dramatically, right? And you got more merchants on there. You've been getting into a debit card, and I want to get more into that.

9:01But the business has just really been taking off. one of the we've talked about the importance of being partnered with merchants one of the big wins early on was with walmart but then you know another kind of challenge for 2025 was one of your competitors klarna ahead of their ipo they got a deal with walmart you're being phased out there and i'm curious how that affects the business as you do that math problem is this going to be a major concern going forward You know, I think the numbers speak for themselves. We have all but been phased out at Walmart and we've had a wonderful relationship with them and have nothing but great things to say about the management team there and certainly the people we worked with up and down the development group.

9:46But it's certainly their set of decisions to make. Just given what we have been able to show growth-wise on our own here, I don't think it's proven to be a major concern for the business itself. I think Walmart, I think you guys have said that Walmart accounted for maybe about 2 % of your adjusted operating income at one point. That's about right. But one of the things that you sort of alluded to and probably sort of dovetails neatly to this question, we do have a debit card that we launched. The debit card is pretty cool. It's a really special product. It basically has two modes. It allows you to say, I am going to use, back to the burrito question, if you have the Affirm debit card, you can connect it to your existing checking account, wherever you bank.

10:40We could also connect it to an Affirm account that we're happy to create for you. But it works with just about any regular bank where if you swipe the Affirm card and you say, not going to do anything, just swipe, it will charge that transaction against that existing connecting account. So if you're buying a burrito, I strongly encourage you to use that mode. If you're going to borrow money, if you'd like to borrow money, you go into the Affirm app and you say, hey, my next transaction at RetailerX is going to be a loan. You go through the same set of disclosures. You do exactly the same thing I described before, where you see the plans, you choose the length of time you want to pay us back.

11:15You see your interest rate, if any. You agree, and the card is ready. And then you can go to any retailer, including Walmart, where even though the direct integration is changing, your debit card will still work. That's another way we are available to just about anyone who wants to use Affirm and qualifies for our product. After the break, some buy now pay later users are getting in over their heads on debt. Max says it's still better than relying on those credit cards. When you put your credit card down at your favorite burritoria, you are financing that burrito too. And immediately starts accruing interest for which you will pay and pay and pay, potentially until the day you die.

11:58Stay with us.

12:10One of the key things, key parts of the credit card industry is credit score. It affects if you can get a credit card, what your interest rates will be and stuff like that. One of the things that a firm is really different from their credit card industry is the data used to create that credit profile, who you're going to underwrite. I'm interested in that because you're using AI for that. You've talked about testing a new metric, cash flow. Why are you experimenting with this? Why is that data so important for your business? That goes all the way back to the origin story again. My credit score had not changed.

12:43It just hadn't caught up. In fact, it hadn't caught up with you yet, right? Well, it also got dinged by my credit card experience in college. And so I was entirely not well represented by my credit score. If you look at my credit history, you would see had one credit card managed to go very delinquent, not a good risk. And yet I certainly was great. I had a company I started, which was well-funded. We took it public. And so a question that's always been in my head, how can we do better? The short answer is, well, if we knew we, the credit scoring universe, that Max had done pretty well for himself, that he had a solid salary, that he had, you know, some reasonably good stock option participations in this company, perhaps I would have been eligible for all sorts of credit products.

13:30And as we started Affirm, we basically asked ourselves, can we reinvent the credit scoring system if we had focused on the real person under the credit profile instead of just relying on a good old score that, by the way, really calcified around the 80s. And so that's where our own, the idea of building our own credit scoring system, process, multiple approaches, we definitely use a tremendous amount of machine learning and artificial intelligence in it. We also do very much love the idea of building a credit score on top of your personal cash flow. So it's a little bit more than just looking at what it you're bringing in every month.

14:10It's also trying to understand what it is you're spending every month. The data used to be very hard to come by these days with modern technology. It's a little bit easier to come by. With your permission, we'll look at it. Very often, it allows us to shape a view of your personal finances that is vastly different from what the canonical credit scores will tell you and therefore give you more credit. And so that's a great tool for us to continue growing our business. I mean, one of the concerns out there with this industry, buy now, pay later, just in general, is that it seems to target customers who conventional credit cards would see as high risk.

14:45It's appealing to people who perhaps are novice users, young people. There's concerns that maybe they don't, they're getting in over their head, right? I think of So lots of headlines out there. My colleagues had a story recently. The headline was, a generation is turning to buy now, pay later for Botox and concert tickets. You know, there's some concerns out there. You were once one of these young people who didn't understand credit. What do you think about these kind of cautionary tales that are just so plentiful right now? So as much as we all can laugh and mock the, oh gosh, they're financing burritos, when When you put your credit card down at your favorite burriteria, you are financing that burrito too.

15:25You may be personally paying it off at the end of the month, but half the country is paying interest to their credit card provider. That burrito gets right on top of that pile and immediately starts accruing interest for which you will pay and pay and pay potentially until the day you die. And so the sort of, oh gosh, a new way to borrow money, let's stick with the old one because it's better is pretty disingenuous. For most people, it is not better. For most people, it's actually, I just don't understand it, but it's always been here. And so maybe I could just stick with that. That's what these articles, that's what these stories say.

16:01The reason by now Paylator is a better alternative is because it is transparent. Now, you can worry, well, they'll use this and this. It's all going to pile together and then we'll just have more debt. That's a much more sophisticated way of expressing this concern, which is why what we do at Affirm is different in the very specific ways that it is. The reason we don't charge late fees is not just because it makes for a great story, although I think it is a genuinely good one. It's because it aligns us with the borrower. Credit card industry makes almost half the profits from late fees because they don't actually mind if you're late.

16:37In fact, they would prefer you set on a$10 ,000 pile of debt, revolved forever and ever, paid minimum payments, and created a cash flow stream for the lender. With Affirm, you are not allowed to revolve. You cannot stay in any more debt than what we have planned out for you on any given month. If you are behind, you are delinquent, and we will not lend to you again until you cure, as they say in the industry, until you basically catch up. So clearly a key part of the company, a clear part of your success is that AI helping you pick who you're going to lend to and kind of managing that relationship.

17:17I think really a firm is being an AI company long before it was cool to be saying AI, right? How do you see AI kind of changing the shopping experience in the next decade, right? I mean, we've seen this big change over the last generation. Keyword search at Google or Amazon, kind of the consumer gets on there and kind of clearly says what they're looking for. Something comes back from search, right? The threat of AI is that a new way emerges of shopping digitally, at least. And how are you thinking about that? Where do you play in that space? So that's an incredibly interesting area. And the short answer is no one knows.

17:57it's going to be a really, really interesting ride from here to the next iteration of product search, which is exactly what you're hitting on. The thing that's both amazing and threatening to some players and exciting to others is this idea that we are rapidly getting to a place where you could talk to a robot that is capable of doing web searches and reading reviews written by humans and other robots to describe actual product experiences. And so the notion of, I am going to buy a TV, but I want it to fit my room and I kind of want to visualize what it looks like when I hang it on my wall. And also I really want to know if it supports my favorite sound system that I already have or a sound system that I'm thinking about.

18:43And on and on and on it goes. Used to be this painstaking process. I'm a bit of an electronics nerd. I would wake up on Saturday morning, sit down and do my research and decide what my next sort of electronics upgrade will look like, I think it's going to become a lot easier. It's already a lot easier. I can have a deep conversation with the chatbot of my choice and tell it, hey, go on the web and research who's got the best standards compliant TV. And by the way, I would like it on sale and I want it to be shippable and kind of want it before the Super Bowl party. And so on and on and on and on.

19:12That is going to become, already is becoming a really important new way. Within that, I actually think we're going to pick up an incredible boost for Affirm and products that live in that same ecosystem of transparency is better. As you are chatting with the chatbot, the chatbot's responsibility, from our point of view, of course, and we're working very hard to make sure this responsibility becomes a reality, to tell you, by the way, this TV is available to you in 12 installments at this exact price. I already know what a firm thinks of you as a borrower. I have gone to a firm and found out exactly what this will cost you on a monthly basis.

19:55By the way, they have no late fees, no deferred interest, no gimmicks. The cool thing about these tools, which is sort of what gets me out of bed every morning, I've always thought it should include at least 5 % of the time spent on figuring out what is the plan to pay for this. AI can take that burden off our hands and we'll be there for it. Just ahead, Max explains why he thinks the next big frontier for AI isn't just about processing power. We're generating more training data than we know what to do with. We're just not storing it. We're not capturing it. That's next.

20:42I'm curious, just hearing you talk about all these ideas, what's it like to be an entrepreneur in this era of AI? I heard you talk about a framework that you used several years ago when you were thinking about the Internet and social and how this was going to be taking off, perhaps. And you kind of use this framework of humans are wired to sin. The Internet is becoming social. Social interactions are fundamentally sin-driven. So I wrote down seven deadly sins and that I'm going to start a company or fund the company for every sin. So I'm curious, what's your framework for what AI might be able to create for businesses going forward?

21:22You know, I'm still working on the exact framework and amazing memory and or archaeological skills to uncover my startup for every sin framework back in the day. In the world of AI, I think the new framework actually may go back to what I thought leaving college. So when I graduated with my meager bachelor's degree in computer science, I was struck by the fact that when I entered school, storage was still denominated in floppy disks, which was sort of a stunning thing. If you look back 30 years now, like a notion of a disk with a megabyte of data on it is such a joke. We have little tiny pieces of metal that have a terabyte on it like it's no big deal.

22:16And so the thought that I had in my early 20s was data is this never-ending commodity. It's a thing we will never run out of. In fact, every time we move, every time we breathe, we're just generating infinite amounts of data. And that's why we need so much storage. And for a long time, that seemed like this sort of abstract notion, you know, nothing to do with anything. In the day of AI, now we are constantly talking about how we're running out of training data. We're not running out of training data. We're generating more training data than we know what to do with. We're just not storing it. We're not capturing it.

22:48And so data is everything is probably the shorthand I have for the moment of today. We're not yet fully aware of what makes data more or less valuable. And so maybe we should just be capturing all of it. Well, Max, this has been fun. I hope it's been valuable. I hope you'll come back. Thank you for the time. Thank you so much. We reached out to Walmart and they didn't respond to a request for comment. And that's bold names for this week. Our producers are Danny Lewis and Alexis Green. Our video producer is Kasia Rousalian. Our fact checker is Aparna Nathan. Jessica Fenton is our technical manager.

23:27Jessica and Michael Laval are our sound designers. Jessica also wrote our theme music. Our supervising producer is Katie Ferguson. Our development producer is Aisha Al-Muslim. Chris Sinsley is the deputy editor. and Felana Patterson is the Wall Street Journal's head of news audio. For even more, check out our columns on wsj.com. We'll link them in the show notes. I'm Tim Higgins. Thanks for listening.