Affirm’s Max Levchin & Axios’ Dan Primack

Thanks for inviting me.

So first of all, an important disclaimer: Affirm consumer is a very large group now. So it's 24 million people who transacted last quarter alone and tens of millions of people who have transacted over the last couple of years who are still paying us back the loans that they took out. And so we have a great insight into what they're doing in terms of their shopping demand and their capacity and willingness to pay us back. The reason the disclaimer is in order is we are unique among lenders. We don't charge late fees. We don't allow revolving. We have a fairly rigid view of if you're delinquent, you can't use the product again. So it's a very sort of you behave like an adult and we'll treat you like one, and if you can't play by the rules, we're not interested in transacting with you, which has a little bit of polarizing from my point of view, a positive selection bias where people who think that's a great thing. I want to live in the world where I have total transparency into my personal indebtedness and I know when I'm going to stop paying my bills for whatever thing that I bought. I love that. And people who are like, I hate it. I want to put it all into a bucket and revolve on it and pay forever. You know, lots of interest and late fees are awesome, which is not...

There are many people who sort of say, 'This will never happen to me.' And the convenience of just dropping a plastic and seeing what happens afterwards is actually quite nice, which, you know, I'm not here to tell people how to live their lives. I obviously think the former is a better way, but the consumer that is in the former self-selected category of this more disciplined approach to personal finance, they're doing great. They are paying their bills. So our last quarter we printed basically the same delinquency rate as we had the one before and the one before and the one before. If you wanted a measure of our point of view on the economy, you would measure things like approval rates. So people who are obviously people we lend money to. If you think we're good at underwriting, which we think we are, you would look at how good we are by examining our delinquency rates, which we publish, things like losses, etc. All of those stayed completely stable to the prior quarters and to all the sort of forecasts that we published. If you would want to ask like, sure, those are the people you let in. How about the ones you've asked not to come?

That would be implicit in the approval rates. If you look at our approval rate, it's basically the same.

We did not. But one interesting thing from the Affirm Economics Bureau, which I made that up on the spot, but like we do look at all the macro elements we possibly can get our hands on. We do know enough about our borrowers to know whether they work for the government or not. And so of course the very first thing I did after a week of shutdown, like this thing isn't ending anytime obviously soon. What's going to happen to our repayment streams? And so we sampled that group of users every day. We saw no deterioration of ability or willingness to pay. So their ability to pay the book, the loans that they have from us back did not deteriorate at all. What did change is that group was not a lot but just enough less interested in shopping. So they applied for Affirm transactions less frequently. Not a ton, like they didn't fall off a cliff, but they were clearly holding back.

Not enough to make it interesting. Sort of across the board. And Affirm is useful in like the best shorthand of describing what Affirm is, it's the best way to pay for things when it matters. Like if you're considering a purchase and it's expensive or it's like just uncomfortable and you want total transparency and no late fees and all this sort of goodness that we stand for, we are the best tool. Like by and large, you will not find a better deal anywhere. And so a lot of these people, I think, whatever they were contemplating in the world of couches or bicycles or prams for their kids, whatever it is, like I'm going to wait until I know when this thing is over. And interestingly enough, from what I this is like 20 basis points. It's kind of a range of change. I think they all came back shortly after the shutdown ended. So this is not a permanent damage to the economy. And I actually talked to somebody who is a professional economist and said I think I'm seeing this where they sort of tap the brakes and then they hit the accelerator right out. Like that's always what happens in...

So majority of our underwriting decisions are very much bottoms up. So we take great pride in underwriting every transaction. So part of why we're able to run the business that we run as successfully as we have been is we underwrite every single transaction fully. It's it's not there's no shortcuts, there's no skips. And so we look at everything we can. Your sort of credit data available from your credit bureau, whichever one has the best record of you, but also we'll go as far as to ask you to confirm your income, pull in your bank account data if you're willing to share with us. So we just get a better and better picture of who you are and how you behave as a financial specimen. And because of that, we actually don't care so much about how the macro world looks. We care very much about how your personal micro world looks right now.

Um, so we don't have a integrated relationship with...

Um, it's something in between. So, we think that it is not our job. In fact, we take great pride in not judging. So, someone's luxury is someone else's necessity and vice versa. Like many many years ago, we had this heated debate where should we be underwriting jewelry and like it's luxury like we want to help lives. You know, jewelry is such a and you know, nerdy software engineers debating whether jewelry is necessary indicates that none of them have ever proposed to their girlfriends or boyfriends yet because it very quickly becomes a necessity. And so that was a that was silly back then. But fast forward to a conversation around like, well, somebody wants to buy $1,000 worth of nylon rope. That seems like a really unnecessary luxury unless you're a climber and which is like absolute necessity because it will save your life. So pretty quickly after the nylon rope debate, I said like we will never debate...

That is very dark.

Good nylon rope does not rip. I understand. But I'm not a climber and I find nylon rope to be...

Yeah. Anyway, but the the best we can do in terms of deciding where we are willing to put pass judgment. One of the obvious things that you can say look if the size of the transaction is relatively speaking really really low and you still need to split it up into four installments or six installments something is going wrong like we may believe that this is a perfectly repayable thing for you but we don't understand what motivates you and so we shouldn't do that until we have a better idea because then we'll be able to build a better product. So at some point we said look paying six for a $30 transaction is nonsense like it just does not make sense given the reality of today's American prices. Like yeah that said for a lot of people it could be the boundary condition between now and the next paycheck. We typically don't lend in the credit strata where a $30 delta is like a really really important thing but we definitely are open to the idea of helping everyone we can but the product should be right. So we built this thing called Pay in 30 where you can you still pay the same $30 but you push it off by 30 days. That said all of this is a little bit academic. Affirm stands for I need to plan my couch, my bicycle. So these are not $30 transactions. They're more like $300 transactions. And so that's where we end up with a vast vast majority of our transactions. Therefore, we haven't felt the extreme need to go figure out how to be in the very low AOV. We're very open to it. We have a debit card product that switches mode in real time as you choose to be either pay now or push it off by 30 days or pay over six or 12 months or whatever. And that will work in McDonald's if I and I'm much more frequently found inside of a coffee shop than McDonald's. But I use my Affirm card probably eight times in the last three days in various coffee shops. None of those transactions were remotely touching the $3 mark. But the flip side is they all settled against my bank account in real time. And so this the card serves all those use cases really really well. And we're quite happy people won't use it in McDonald's.

The short answer will not do justice to the situation. There are absolute winners and losers. And actually, it's not I'm not breaking any interesting news here, mostly because a lot of these merchants are publicly traded and you can see even the stock market reactions to their quarterly earnings are like 20% up, 20% down for a bunch of them. And there's almost no rhyme or reason to it, but it's primarily a reaction to growth. And some of the merchants we serve are super growthy. They're like doing really, really well. And some are not doing well at all. And so I think last quarter from memory, sporting goods and outdoor sales, if I remember correctly, basically flat year-over-year. I don't want to drop any names, but names in that domain are not doing well in the stock market. And by the way, they're not doing well, period. On a flip side, general merchandise, like diapers all the way to couches, great. Travel still strong. Like every quarter now I've been like this must be the last quarter of 30% plus growth in travel. Not true. Since COVID people have just been gorging themselves in travel and doing just fine.

We've gotten really good at starting small and growing well. The most important kind of activation energy leap forward a merchant has to take is it's some integration. So you're adding our API, you're settling through a completely new settlement pipeline. We have lots of partnerships. So we'll work with you if you're on Stripe, if you're on Adyen, like all the various backend partnerships we have, but at the limit you're still adding something to your site or your store, you know, whatever the integration is, you have to believe you're going to have incremental sales. And if we don't nail the first several months of this incrementality, whatever's on the roadmap to do more, it's going like, 'Ah, wait a second. Is this working?' And so this is now fortunately a theoretical problem. Back in the day, we would launch a merchant be like, 'Huh, the last five had 30% pop in sales. This one isn't. What is going on? Is American consumer no longer interested in Affirm or is this merchant screwing it up or are we screwing it up for them?' We've now, so last quarter we had 420,000 merchants that transacted with our 24 million consumers. We've gotten very, very good at saying on day one you're going to have the set of products available. Here's the pop in sales you're going to see. Here's the incremental value you're going to get for every dollar you spend paying us for the privilege of accepting Affirm. And it's only going to get better from here. Every quarter, you're going to see more and more and more. And gotten really good at coming back and saying, 'All right, here's the next thing you should do. We didn't push it on day one because you didn't necessarily believe. Now that you're seeing it so incremental, you should do more.'

I'm so sorry. We didn't really talk. I saw you walking by and you were on your way somewhere and I was on my way and I thought, 'Oh, look, there's Max.'

That's how I like to have my Vegas experience. Fast and out.