An Honest Conversation w/ Affirm CEO Max Levchin

Thank you. Great to be here.

That's exactly right. I'm not sure there's a ton more to add to the punchline and that is thoughtful, responsible usage of credit is basically investing in your future self, your future success and can take the form of anything understood as borrowing for your education. Vast majority of Americans, yours truly included, had to borrow money to afford higher education. We all understand and things are changing pretty rapidly with AI. But even in the world of post-AI world, my guess is being well trained to do a job will pay better. And so when we go to college, when we go to vocational school, we borrow money because the future earnings are all but assured. And that's kind of the canonical use case. There are many others in terms of investing in a productive version of yourself, including borrowing money for a small business, startup, etc., etc. The kind of more varied use of credit, if you will, is just affordability where you're borrowing money for a car to get to your job or perhaps to just tool around. You obviously have to borrow money for a home. These things are expensive, but they are kind of the American dream. The storage of wealth and you definitely should be investing in your real estate holdings if you can. And so borrowing money is typically the only way to go about it. And then probably the most banal but most frequently used case for credit is simply smoothing out your personal cash flow. For a lot of folks it is not a guarantee that the suddenly necessary or even pre-planned thousand dollar purchase is something that just comes right out of savings account and it's no big deal. And for many, savings account does have that thousand dollars but it is scary to drain it because what if an unexpected hit comes in next and there's sort of oft-repeated line about Americans are a couple hundred away from real financial trouble in case of a medical emergency etc. And so usage of credit when done right just provides that sense of control and certainty of your personal finances even at the sort of couple hundred, sometimes under $100. And so that's where Affirm plays obviously and we really see our role and our mission to provide this sense of clarity and control for each consumer as they navigate their ups and downs of everyday life.

It's a great prompt and obviously I'm very passionate on the issue. So, I'll try to keep it compact. Interrupt me and ask me to clarify if I'm not clear since this all makes so much sense in my head. I think it makes a lot of sense to our roughly 40 plus million consumers. But the basic difference between revolving credit and not is you're essentially signing up to a product that gives you no sense for when you're going to be out of debt. Most of us sort of consciously or subconsciously subscribe to Benjamin Franklin's point of view that it's generally speaking not a great idea to be in debt. And yet, as we established, it's actually a really, really useful tool to have credit for anything from cars to everyday purchases. And so how do you sort of slice this thing into two parts in a way that makes sense for both of those approaches? And it's really simple. You have to know when you're done paying off whatever it is you borrowed. That gives you the sleep at night, the I know my schedule, I got control. Like all those positive things come with the notion of start here, this many payments and there. That is how Affirm works. But before I get there, that is not how credit cards work. So credit cards basically say swipe. You know your rate. You kind of know some fees might apply, but what you don't know is when are you going to be done? If you look at the design of credit cards, it's actually deliberately made obscure. If you look at the fine print, which 99% of us never do, but if you flip your paper statement, there's a lot of sort of nine-point font in grayscale telling you here's all sorts of things that could go wrong, including when your rate might get adjusted, etc., etc. But the most important thing that it doesn't say, it doesn't say anywhere is when you swipe a card, you'll be done paying the swipe off by date X because that doesn't exist. You have a notion of a minimum payment where the card tells you, hey, you got to roughly pay off 3% of your balance every month otherwise there'll be problem. The reason for it is very, very simple. Not passing any judgment here, although I obviously am. The bank that gave you the card has the following math to go through. Here's a total balance. You got interest which is what accrues over the course of the time it takes you to pay it back. Whatever you haven't paid back there is interest calculation goes in that goes right into the principal the total amount you have to pay back. So the longer the balance stays high the more money you'll owe to the bank. The reason the minimum payment exists is basically some degree of both regulatory pressure and the banks need to believe that you're still going to pay something. But in general, they kind of want you to stay with a big balance accruing a bunch of interest and just paying it off for as long as possible because that maximizes their profits. It's a business model, but is a very, very stressful product for the end borrower. Affirm was born from this realization that most of us actually have a very stressful relationship with our credit cards. We don't know when we're going to be done paying it off. And so for thousands of years, there was a notion of an installment loan where you would say, I'm going to borrow X dollars. I'm going to pay it off in Y months. There'll be some interest or, you know, Z, whatever, whatever variable you want to use. At the end of the period, I'll be out of debt. Maybe there'll be some fluctuations. There might be some rescheduling, whatever. But generally speaking, everything going well, I will be out of debt for this particular transaction in this many months. That is literally what we built. And before things like smartphones and online transactions, it was very hard to scale that kind of borrowing, which is, by the way, this is exactly what you do when you buy a car, exactly what you do when you buy a home. So the installment loan is alive and well. We even sometimes use that in buying big things like refrigerators or HVAC systems and really expensive stuff. Doing this for a $200 purchase, it's a lot of work. Just swipe your plastic and move on. Because of the internet, because of smartphones, because of modern technology, we can scale this down and give it to you in a form of a little menu that shows up everywhere you go these days. We're available in everything from Amazon on down in e-commerce and now in store where you just say, hey, I want to borrow $500. I want to be done in 6 months. Here's my monthly payment. Off we go. And that is the fundamental difference between borrowing with revolving or not revolving. The fancy way of thinking about it mathematically is revolving is essentially constantly refinancing your debt, which most of us know constantly refinancing your debt is probably a bad idea because you're staying in debt. Affirm and buy now pay later in general, you don't have that option. And that's a good thing. You actually are given a constraint that keeps you clear and in control and sleeping well at night because you know when the transaction is done being paid off.

The very basic answer is because planning is hard and planning against exponential curves is like basically impossible. So a fun anecdote true story 15-ish years ago when we just started Affirm we were still I've been in payments internet payments my entire adult life. So kind of a professional if you want to call it that. And our first CFO Rob and I were flying from San Francisco to New York to try to explain our newly created business to some potential investors. And we were deciding sort of how do we explain this thing? So like credit cards are kind of easy like everybody knows how they work and this thing is like a throwback to basic installment lending. By the way, we don't compound interest in the principal. We're not going to charge late fees. We're going to make it super user friendly. So we were supposed to make a presentation. So deck page number one was here's how Affirm works. Borrow $500, decide on an interest amount of let's say $30 more over the course of the next six months. $530 divided by six off you go that's your family done. Next page like here's what it cost to do a credit card. So like well we didn't this is in the days when they didn't have an especially great Wi-Fi connection on the plane. So we're like well instead of looking it up on some online calculator I'm just going to do it by hand here because it's you know we're mathematicians we're credit card rates whatever. So in the next six hours, the entire flight to New York separately doing this initially on napkins that Unreal pad, they're like, all right, well we don't seem to agree on a number. If you borrow $500 with a credit card and you have a minimum payment of X and a rate of Y and you're not going to be late, so there's no late fees to compute over. What's the total cost? And so at the end, we landed without a number that we could agree on. And by then, we were both modeling it in Excel. And then we finally got on Wi-Fi and looked it up online and neither of our numbers was right. And that number came from some random online bank that was willing to pre-calculate for you. So the short version of the answer is no one knows what the true cost of credit is. And with Affirm, you do. And the reason it's hard to plan. You have family, you have health issues, you have job issues, you have all the things that kind of you have to deal with because they're real stressors. Why would you want to add a stress of I have no idea how to think about my money? And yet it is one of the most important topics people think about all the time. And so just the reason it's so important is because it's a stress. It's a weight you don't have to carry. You can just use something like Affirm or please use Affirm obviously and feel easier and better about yourself and about your money and not worry when will I be done paying it off? What's the true cost? You really cannot know the true cost of a card swipe. Cannot even you know people who are supposedly professionals cannot figure it out. How would normal people do it?

It's an old mousetrap, but it is a better one.

Two reasons. It's really convenient, and it's really profitable. The profitability of a revolving credit is easy to imagine. Like if you remember your high school math, the cost of credit with Affirm is basically or total cost of the whole loan with plus whatever you might pay Affirm is you can think of it as a line. Like it sort of it starts with your first payment, ends with the last payment. It's a line. People are pretty wired to think linearly like 1 2 3 4 like I kind of know what the next number is going to be. I can predict that. Credit card cost is an exponent. In other words, it's a thing that looks like a curve. 1 2 4 8 for a lot of people like is it 8 or 6 the next one and so we're wired to think it's going to be 6 but it's actually going to be 8 if we're talking about an exponential curve here and so on and so because it's hard to figure out because it is stressful and I think people who issue credit cards revolving credit cards anyway understand that it's stressful but because most of us have a hard time figuring out whether it's a 6 or an 8 or 16 there's a lot of profit in between those numbers and so once you taste that profitability, it's very hard to go back. You have shareholders, you have people saying, wow, that's an amazing mousetrap you invented there with so much profitability. So, it's not hard to see why the incentives are on the wrong side of that story. It's also the case that revolving credit is a more flexible model. I'm not actually into revolving credit as a concept. I just think that you need a lawyer present at all times because you need to read the fine print. You need to know the penalties. You need to know how the minimum payment factors into this thing. What happens if your rate changes? So as a corporate borrowing mechanism, if you look at like giant companies borrowing money from each other or from banks, sure, use revolving credit all you like. That's a great idea. They have massive lawyers armies on staff. As a consumer, you just have you. So simpler is absolutely better. And by the way, it's also cheaper. And so in the case of Affirm, you will pay less just because there's not that exponential curve. It's hard for those that created these exponential curves to say, cool, I'm going to get out of this business. I'm just going to start charging these linear terms. And so I think the world by the way is changing for the better and I'm sure we'll get into it because it is the topic of the moment. But there's a couple of things happening secularly that are pretty conducive to the positive change. So part one, we're all getting smarter. AI tools are there. They're available. They're really inexpensive. A lot of them are free now. You sort of have a financial adviser available to you at all times from the likes of ChatGPT and Gemini and everything in between. And so over time you'll get better and smarter about just managing your money, thinking about your money. Maybe you'll still be stressed, but at least you'll have transparency into what's really going on. So, I think we're all getting smarter about money. Two, we took a really maximalist approach where we said no late fees, no compounding. We cap out the total amount of cost you could possibly pay us back. So, if you borrow $500 and interest is $30 and then something went wrong and you've rescheduled a payment, that $30 doesn't change. It's still $530. Obviously, we're leaving money on the table. It's open to abuse, so we have to be thoughtful about it. But generally speaking, we picked this because we wanted it to be super simple. We wanted it to be really, really easy to understand. When we started, I met a bunch of people who run banks that issue these revolving credit cards, and 100% of them told me, you're insane. Like, this is far less profitable. It's never going to work. You're an idiot. Like, just get over yourself and issue a credit card already. And you'd recognize some of the names of these CEOs, by the way. These are like, you know, well-known, very thoughtful, very smart people. Fast forward 15 years, it's not just us. More than half of buy now pay later volume in the US does not charge late fees. We were the only ones at start. It's now more than half the industry. And so people are generally realizing that if you treat the consumer right, they are loyal to you. North of 95% of our transactions come from repeat customers. So there's something in this product that really makes sense. And so as we sort of went around doing our thing the only way we would do it, people started to copy us. And that is a good thing. Like I love the fact that more and more people are shifting to no late fees. More and more banks are offering some version of buy now pay later, really pre-priced credits they have upfront pricing and many are now saying we'll not have any fees so it's all kind of shifting towards this model but I think for folks that primarily make money on revolving it's just going to take a very long time to kind of get off the drug of the profits that these things generate now.

A couple of different reasons actually. So, one of the things I realized at the very beginning is a lot of the fees and a lot of kind of the stuff that we all have come to hate about credit cards is actually there because the provider doesn't get a chance to say to a consumer, hey, wait a second. That's a bad idea. You're borrowing money. Maybe you really want this couch or this bicycle or whatever it is you borrow money for, but it's just a dumb financial move. You're making a mistake. Put that card down. Let's rethink this. In part because these cards are so convenient. Like I touched on, you just swipe the plastic and it happens. There's nothing to sign, nothing to really sort of upfront, nothing to think about. That creates for a wonderful engine of the economy, if you will. But it sure allows you to overspend. So one of the key early design criteria was what if we made purchasing conscious? What if we made you ask yourself, is this a good idea? And what if we gave ourselves a voice? That would allow us to make it cheaper for both sides. If you get into debt, if you get into a purchase that you shouldn't have, we should lower things like return rates because you might not be trying to get out of a thing you shouldn't have bought in the first place. We will definitely lower defaults and delinquencies because we would tell you we understand your financial situation reasonably well. This seems like a dumb move. Like at least let us ask you more questions. So sometimes if you use Affirm like really extensively, you'll see we won't just say oh cool okay let's go with transaction. We'll say things like hm this seems like something we need to learn more about. Can you please tell us your income or can you please let us see into your bank account so that we can understand your cash flow better so that we can tell you whether it's a good idea or a bad idea. None of this is available in credit cards because the transactions are literally one touch and you go and so we wanted this interaction to give ourselves the right to communicate with you and to ask you sort of like is this a good call? That is also why our delinquencies and defaults are so much lower than the standard credit card world because we do have a chance to tell you, wait a second, this is a bad move. We're going to say no. If you want to do this thing, you should make a bigger down payment. You should maybe save up. Maybe you just shouldn't do it at all. And it seems like exactly kind of thing a consumer product designer would be like, oh my god, you're telling someone no. It turns out that people love this. They really enjoy the interaction of no one likes to be told no. But if you're like, sorry, it's a hard no. Your card's been declined. Versus, hey, here's kind of why you should reconsider. By the way, there's probably a way to come into this item that you want, but you should save up and make a bigger down payment because then you can afford the monthly payments. And so, that was a key design criteria. As we thought of this, we realized that we wouldn't need late fees. We wouldn't need all the crutches because they really primarily exist to cover up for the oh man we should have never given you a line of credit that long and that's why it works as well as it does.

That's right and every transaction is underwritten by machines no humans are involved. We actually find that we are much better as software engineers in terms of underwriting and by we here I mean the industry and so the throwback to the 1950s where a fat banker shakes your hand and says you sir look like a good risk cuz you look just like me. Like that's where all the discrimination came from. People are like, oh, you know, I don't want to lend to someone who doesn't look like he's from this neighborhood. It's a great thing that we left all of that behind. Machines are really, really good. They're also very good at saying things like, I will not consider race. I will not find out what race you are. I'm not going to guess what race you are because it's an illegal basis. Shouldn't use that. Can't use that. And so it's much easier actually if you have the capacity to build these models and train them to build a really good very fair model that will pass your fair lending exams and all the sort of usual regulatory hurdles that you have to clear if you lend. So we have a giant infrastructure that does we've been an AI company for a lot longer than AI companies were cool and for a little while I was bothered by the AI revolution. Been on the forefront of this thing forever and it's obviously this particular tide is lifting all the boats and has been lifting us for a very long time but it's all entirely machine driven. We have a large group of AI and machine learning engineers that are extraordinary. We continue building completely different kind of models that the industry has never seen before and deploying them and we do it responsibly. So we run it right alongside older models and make sure that the new model approves more people with better financial results and doesn't discriminate and all that good stuff. And so it's probably the very core of what's happening at Affirm is just a bunch of math nerds who are obsessed with credit and also have this extremely clear point of view on what is and isn't moral and that's allowed us to build a pretty great company.

So before Affirm, before the current thing, my current thing was PayPal, which I co-founded basically straight out of college. I got to the US just a year and a half before college. So I was completely sort of fresh immigrant. I came from USSR at the time. And so my exposure to capitalism or American credit system was, let's say, non-existent. So I got to college, I got my first credit card on campus. That was great. I didn't make my minimum payments because I didn't read the fine print. But who does? You know, you're busy doing homework. And 90 days into not making any payments, I got an extremely nasty notice saying, hey, like one, your credit card is not going to work anymore. Two, you owe us $600, so get on it. And by the way, your credit score is trashed. I'm like, well, I can go earn $600, which I did, and paid it off and didn't think of it twice. Fast forward six years, we just took PayPal public. I was independently wealthy and I decided that it's time to upgrade my Honda to a Mercedes. So, I flew to LA with my then girlfriend now wife. Went to a car dealership and said, I'd like to buy this really cool looking Mercedes convertible and guy said, great. We're going to talk to the finance department. They're going to get you a car loan. It's going to be great. 5 minutes later said, actually, can't buy this car. In fact, young man, you should get out of here. And like, wait a second. What's going on? Said, well, we looked up your credit score and even though we know your name because PayPal just made the rounds of news and LA Times did this big spread on, you know, PayPal founders make good yada yada yada, your credit score basically says you're a deadbeat. We don't know what you did to yourself back in the day, but with a FICO score of 600 or 500 something, you can't buy this car. Like, you know, maybe you can go rent one. And so I first of all I was embarrassed to like the color of my face was several shades darker for a couple hours. And I eventually had to figure out a way to pay cash for the car. And that stuck with me where like and the unfairness the fact that the mistake I made in college which I rectified very quickly was six years ago and my credit score hadn't recovered and I look like a total idiot in front of the girl I was trying to impress at the time and anyway. So that sticks with you for a lot longer than you might expect. Another five or six years went by and I would still look up my credit score, which I became very conscious of, and it would still not update and it would still be like, yeah, you know, you've never been late in the last 10 years, but yeah, you're not that good of a risk. And so after PayPal, I was hanging out with one of my earliest compatriots there. And he and I just got into this long debate of like how in he was very aware of the embarrassing car purchase story. And so we sort of got talking about like why didn't we bring the kind of math expertise and willingness to go the extra mile and figuring out human behavior through AI and modeling which

We did at PayPal by the way plenty. Why don't we bring that to credit? And we're sort of like, ah, we don't really have a good answer. Maybe we were too young, didn't understand credit at the time. We were just too focused on payments. And one thing led to another, he and I co-founded this company. So that's the full story of Affirm. There's no better story, no better motivation than a personal thing that sits in the back of your head and still embarrasses you.

One sort of asterisk or a postscript into this thing. Part of why it took me so long is even when you're young, you don't have the empathy for others as much as you do 10 years later. Affirm started when I was in my late 30s. And in my early 20s when this happened, I was like, 'It's just me. This sucks.' Like this was so embarrassing, but obviously I'm the only person to ever happen to. In my late 30s, I knew enough people who had a version of the story because I would get triggered when somebody would say, 'I got declined for this or I was told my rate's going to be that.' I'm like, 'Oh, let me tell you my story.' I was so angry. So, by the time I got to it, I was like, 'Oh my god, I know all these people and this is a terrible broken system. I got to do something.'