Affirm Holdings, Inc. (AFRM) Earnings Call Transcript & Summary

Great. Well, thanks, everybody. Welcome to the Affirm fireside chat at the Goldman Sachs Communacopia and Technology Conference. I have the privilege of introducing Max Levchin, Founder and CEO of Affirm; and Michael Linford, CFO of Affirm, which is the leading U.S. By Now Pay Later provider. My name is Mike Ng, and I cover Affirm as part of our fintech coverage here at Goldman Sachs. We have 40 minutes for today's presentation. We're going to leave about 10 minutes for investor Q&A. So if you have a question at the end of the session, please raise your hand, and we'll get a mic over to you. First, Max, Michael, thank you so much for being here and participating in our conference. So Affirm is the leading U.S. Buy Now Pay Later provider with about $15.5 billion of GMV in the last fiscal year.

To start, could you help contextualize the market opportunity that Affirm's 2-sided payment network addresses and how Affirm provides value to consumers and merchants in a way that should continue to help it gain share of commerce over time?

We normally do these things. Does everybody hear me? No? How about now? Can everybody hear me?

This is a technology conference after all.

Fortunately, I'm a technologist. How is that? How about now? So as usual, I'll start and Michael will summarize and keep a bit of short answers. So if you look at where we came from, where we are today, it will give you a pretty good idea of where we're going. We started almost 11 years ago basically, helping people buy things that require consideration; so couches, Peloton bikes, mattresses, all sorts of things that require you to make at least a basic mental budgeting calculation. The idea was that we could be less expensive; more importantly, less confusing; safer than a credit card, all stemming from a study that I read almost 15 years ago now, describing how millennials are opting out of credit cards because they are confused by the fine print and they have no time for it. About 5 years ago, as sort of whole momentum of the genre picked up, we decided that this actually works just fine for less than a perfectly considered purchase. So we shifted into things like apparel, beauty products, anything that sort of -- we basically went down from several thousand dollars down to several hundred dollars, and then down to several dozen dollars. The announced product that we've talked about a lot, Debit Plus, which is our foray into a physical card, it's really much more than that. It's our assertion that we deserve to play in every transaction type. Building a network means you want ubiquity of coverage for merchants and consumers. There also has to be an ubiquity of transaction types. If at any given time, you tell a consumer, "Hey, we're really useful, but not for this," you're ultimately damaging your story, damaging your brand, damaging kind of the comprehension of, this is a pool for everyday spend. And so the -- by that logic, it should be obvious that we're going after the totality of consumer spend in the U.S., small number. The good news is that BNPL, even under today's definition, is massively underpenetrated. It's still single digits, and so we have trillion-ish to address and going pretty strong so far. The way you add value -- so really describe the consumer side of it, but it bears repeating, we are, until Apple's announcement in the space, are the only player that doesn't charge any late fees. Certainly no origination fees, no deferred interest, all the sort of schlocky things that our industry is infamous for, we opted out by design on this idea that if we build an honest product, treat the consumer right, we'll build a brand and a preference. And that's borne out to be true both in consumer references, also consumers' willingness to pay our bills, which is going to become progressively more important as we head into a downturn. And then on the merchant side, the simplest explanation back from the earlier days when I did a lot of the merchant telling myself was merchant keep discounting. Like when you go into a store and it's a 10% off, it's not meant for you, it's meant for anybody who's passing through. A much, much, much smarter way of doing this is individual pricing, and discounting in the form of 0% loans. A huge percentage of our volume is in consumer face no interest because the merchant is subsidizing it. It allows them to preserve integrity of price, improve conversion, improve consumer satisfaction because they know they're not paying any interest. In the case of the firm, they're also not paying any fees. So we help merchants convert. We also help them build the brand. At this point, we're big enough where we accrete to most merchants brand building versus in past where merchants validate us. I'll stop there. Michael says it is good enough.

So Affirm has done a really great job of diversifying business and product mix. Peloton a couple of years ago with 30% of revenue, now it's less than 10%. There are a couple of things I want to explore here. First, on underwriting, there obviously have been a lot of questions around the robustness of underwriting standards among all BNPL providers. I would think that Affirm's history with Peloton and longer duration to your point earlier, high AOV loans has probably led Affirm to be more diligent in that process. So could you talk a little bit about Affirm's underwriting practices? Why that's important? What are the implications for origination volumes as we think about going through a cycle?

So there's a lot of rich veins to explore there. So I'll be ended through it a little bit. But feel free to guide me, and Michael, interrupt me. So the reason we started at the high ticket items was very much by design. I'm a big fan of moats as in things that prevent competitors from just cloning you and selling your product at a discount. Huge moats that you can build if you have the time and the data and lots of reps is underwriting. It matters less in shorter term transactions simply but if you are lending money over 6 weeks, you have 4 opportunities to default. You're lending money for a Peloton bike, on average, there's 39 opportunities to default. And so you've underwriting decision and you say, yes, you're betting the person who has the means to date, and they still have the means 39 months from now, and they will be willing to prioritize your bill if something changes about their ability to pay back with no stress. And so even in the best of economic times, there's always an opportunity to scrub underwriting simply by taking more risk. We made it harder on ourselves by saying we're not going to have compounding interest, we're not going to have deferred interest, we're not going to have late fees. So we basically stripped the model down to -- you've got to be really good at underwriting or you lose your s***. And we did that for a very long time. Feels like the microphone only somewhat functions. We did that for a long time before we entered lower ticket items, they are easier to underwrite. We also make less money because merchants recognize that you're taking less risk and, therefore, are less willing to compensate it for it. The fundamental risk management approach we have at Affirm is we have to get it right today. We have structured the product so that they're, generally speaking, very short duration relative to credit cards, relative to any other form of lending. About half of our loans are paid up in 4 months, about 80% in 8 months. So these are very, very short-term assets that allow you to focus on the front book as the jargon has it, not so much in the back book. But still means you can't screw up your underwriting decisions, but it gives you a lot of flexibility. That allows us to focus on 2 things, very broadly speaking, the quality of the model. So this is my favorite topic, so I'll try to keep it short, but this is really something I can spend several days talking about probably. A great model is one that doesn't decay, doesn't leave its ability to rank risks in the times of economic change. So there's no world in which when people lose their jobs, they will still pay you on time because people eventually need to find money. And if they don't have it, they do have it. The best you can hope for it, the best you want in one of those model is when you look at a collection of applicants asking to borrow, you want to have pretty good certainty that today, tomorrow and X months from now, your model will rank order the risks roughly to what the outcomes will look like. As models [ decay ] here, that's when you sort of find out who's actually batted underwriting, the need to be a good underwriter when everybody pays you back. When that becomes a little bit shakier if your model still rank orders your risks really well, that's a great model. That's where we spent the last 11 years convincing ourselves through all sorts of [ experimentation ] through the short recession of the brief COVID recession. We feel very, very good that our risk ordering is quite robust. It hasn't deteriorated. It's been performing to spec, and we continue getting better and continue launching new models literally every few weeks at this point because it's just something that we really are quite engaged in. The other part of this management is we set cutoffs. You look at the model and you say there's always a gray zone of people that may or may not keep their job through the recession. And what you're doing is you're trying to make a prudent, yet economically viable decision about making it a loan today, even as short term as ours, given that you have conviction your risk model ranks well within the grade zone of people who may or may not become viable at the risk. We feel very good about our policies. I think the major change that occurred to us in the last, call it, 9, 10 months. So we started doing this even before the year began is we review our policy and our posture of those managers much more frequently. So our risk teams meet every week multiple times reviewing all the metrics because we have [ abundance ] of data coming in. As a management team, we used to look at our risk situation and the performance of the book and sort of all the leading indicators and the back book performance probably once a month. As of a couple of quarters ago, we've shifted to a much higher gear. So all of us are paying much greater attention in part because, frankly, it's fun, maybe a very strange fun. But we -- that's what we bet the entire idea of this business. We will be the best underwriters in the space. As the tide received, we'll be swimming naked. We will -- we're making sure that our pants look extra good. But we feel quite good about our chances of survival, but it doesn't need abate the need to stare at all the metrics and tune policy much, much, much more frequently than we were a year ago.

On a related topic, you guys have also been doing a great job of merchant diversification. And as Affirm diversifies beyond Peloton, it's also getting a lot of exposure to different industries. One category that's been really promising has been general merchandise, and it's been supported by Amazon and Walmart and Target. So could you talk a little bit about your merchant and industry diversification? Then as an adjacent question, could you talk a little bit about what Affirm's points of differentiation are that will cause an enterprise merchant to go select them over other Buy Now Pay Later providers? And I'm sure underwriting plays a big part of that.

So you can think of the merchant strategy really, really simply, it's a march to everyday utilization. So we started out in kind of an hysteria of connected fitness, and in connected fitness, everyone's got some connected fitness device. But 10 years ago when we just looked at it, it was very strange that you might want to have something like this in your living room instead of your gym. But that's -- I think you buy once every 5 years. To increase frequency, you have to go towards the everyday purchases. So we've marched down the AOV and up the frequency ladder very, very deliberately. We added a massive number of users even last quarter. And yet even with these new users, our overall transaction frequency still went up from 2.7 to 3 even, I think. That shows that we are successful at our teaching of our consumers that we're not just for Pelotons and Casper mattresses, we're actually available for all sorts of things. General merchandise being kind of the thing that you buy every week. They're still the completely underpenetrated, largely off-line market, which in itself is underpenetrated, but largely underpenetrated market of groceries, both online and off-line. That's an everyday or at least a couple of times a week purchase, and so that's another really major area of opportunity for us. The reason -- and by the way, the efficient way of distributing into those verticals is obviously partnering with giant platforms. It's very, very hard to convince every local merchant that sells you grocery that they should take Affirm. But if Walmart decides to do it, that covers a vast majority of the United States. There's a Walmart 90 miles from literally every single person. And so general merchandise push has been all about ubiquity. The reason we have a right to win those deals is at least trifold. One, we are actually better at underwriting, which means that we can say yes to more consumers. And we have lots of tools of sticking with, yes, even as the economy worsens, for example, we can ask a little bit more downpayment, we can change the term length that we're willing to take the risk on. But generally speaking, we have been improving or increasing approvals even as our posture or point of view on credit has worsened. Two, probably something that's truly critical for a lot of new merchants, the really big platforms, very few of them are in the business of selling you a bunch of $10 items. Most of them are selling some $10 items, some $50 items, some couple of $100 items, and a few $ 1000 dollar items. You're coming in to buy some groceries at Walmart, you may or not be in the market a set of tires. We're the only player in the space that has been able to port our model without the fees with the approvals with old and goodness of Affirm to every AOV. We do not have limitations on -- well, only 6 weeks and only $150. Like we work at $150 and $1,500 and in some cases, $15,000. So being able to go up and down the entire [ AOV ] stack has proven a very powerful proposition for our partners. Within that, when you're selling a huge number of what essentially is financing programs, it starts to look like integrations across multiple different products. Probably one of the most important reasons people, merchants choose us is we are technically exceptionally strong. We are a technology company founded by [indiscernible] science majors. We build this thing like an engineering stack, not a slap together financial services product from the '80s. It really helps you when you're coming in to partner with the likes of Shopify or Stripe or Adyen, Amazon and more traditional players like Walmart who value ability to deliver engineering solutions on time with robustness scaling through Black Friday, Cyber Monday-type events. All that -- I don't think there's anyone else in the U.S. or worldwide that has nearly the quality of engineering. Our underwriting function incidentally lives within engineering, which would sort of tell you how we think of those things as well.

It makes a lot of sense. Yes, the engineering the broad product portfolio, the adaptive check out, can certainly see that. I have one for Michael. This is one on guidance. So Affirm has guidance for the quarter, GMV of $4.2 billion to $4.4 billion, full year $20.5 billion to $22 billion. Could you just talk a little bit about how you're pacing towards those targets? And what do you see as the biggest swing factors that could affect your ability to achieve it, underperform, outperform?

Yes. So we're obviously not going to talk about the quarter here. But I think we're really pleased with the traction that we have with our largest platforms, and they're going to play a really big part in this year's business. Things are moving so fast in our industry. I think a lot of folks have forgotten that we launched with Amazon in November of last year. It's not even 1 year old. It's still an infant, and we have a lot of work to do to continue to grow that further. I mean the team is busy at work with a very long list of optimizations that are small tweaks to the user experience that can have huge impacts on end-to-end conversion. And similarly with Shopify, I think we have a lot of opportunity to continue to grow our share of their total process volume. And so we have a lot of wood to chop just with these kind of scale partnerships. I think that obviously is an opportunity for us to continue to grow our user base alongside that. And I think one of the biggest pieces of upside for us is really engaging those users throughout the year. And as proud as we are of driving that frequency measure up, we really do think that we have a lot of upside to that as we continue to grow our low AOV business.

Great. And Michael, could you talk a little bit about how you expect the shape of the year to unfold? And any comments as you can make to GMV revenue less transaction expenses, EBIT?

Yes. So we have a seasonal business that was controversial a year ago. I really hope folks understand that retail is a seasonal business, and we definitely have a seasonal business. We do a lot around Black Friday, Cyber Monday and through a holiday season, where a lot of considered purchases happen. And what that means is our Q2, our fiscal Q2, calendar Q4, has some pretty acute seasonality factors that shake out by the end of the year but are really important for the quarter. We do expect more GMV, sequentially higher GMV and that seasonality to show up in the total processed volume. But the revenue and associated unit economics, we call that revenue less transaction costs, will be lower on a percentage of GMV in Q2. And as a result, your adjusted operating income will also be lower in Q2 seasonally. A lot of it has to do with the surge in GMV and also the timing of it. A lot of shopping happens in the December month in the quarter. And when you back in, in the quarter, a lot of volume results in a lot of the costs and transaction costs showing up in the P&L, like provision without all of the revenue there, and that happens and has earned over a longer period of time. And it just creates a little bit of back-endedness to the P&L, and we would expect that to be the same thing that we saw last year. We would expect the same kind of trends to play out in this year, and that has a lot to do with the big categories we continue to expect to see growth with, were also growers for us last year, and we would expect the same underlying trends to play out again.

That's great. Just while we're on the topic of guidance and the numbers. So the full year guidance implies revenue less transaction cost margin of 3.7%. Some of the questions that we've gotten post earnings is whether or not there's enough provisioning in those numbers given that charge-offs have gone up a little bit, though, obviously, you're growing. ITACs has gone down a little bit lower on the spectrum. The flip side allowances have gone down. But obviously, you guys have a very unique product that affects that. So maybe you could help just clarify some of the expectations around credit provisions for the year and talk about how you feel about loan performance at this point in the cycle?

Yes. So this is a -- it's a bit of a curious question for us internally. It's hard, I think, for folks outside to really think as we think. And I think maybe the most important insight is we don't think about our business on a portfolio outstanding basis. That's the rhetoric you hear from a lot of credit card companies, companies whose business model is to get you to buy something and then get trapped paying it off forever. Our business model is really built around the unit of our business as a transaction. So we think about for a transaction, what's the economic content in it. A lot of folks looking at the charge-off number, for example, are trying to denominate it by the current portfolio outstanding and they ignore the fact that it goes back to originations that probably happened in Q2 or Q1 of last year, where we had really strong unit economics on a horizontal basis. And we seek to make underwriting decisions and credit approval decisions to ensure that we have enough economic content at the time of the origination. The way the provisions roll through is a lot more mathematical and mechanical than it is judgment based. We don't sit around as a management team and think about, hey, we need to increase or decrease the amount of the allowance. Instead, we use our models, which as Max talked about, are still really good at sorting risk to help us understand what the probability of loss content is on the loans that are on the balance sheet. And we look at that off balance sheet, too, because it's obviously very important to our capital partners as well. And the math tells us how we need to think about the total amount of allowance needed. The last thing to mention again, because our unit is the transaction and because the velocity is so high in our business, it is just not the case that most of the provision, which is the unit economics hit is tied to what's left at the end of a quarter. It's actually what the flow-through was throughout the quarter, and so the loss numbers that I think some folks have anchored around are just not where the business is at. And so I think that concern comes back to somewhat of a disbelief that we're able to run our business with really healthy loss numbers. And yes, it's the case that if loans deteriorate, if the delinquencies rise in such a manner that we call it flow rates that people go from being 4 days delinquent to 30 to 60, then we would take additional allowance that would show up as additional provision. And if the opposite is happening, if consumers are on time and making their full payments or making payments consistent with the expectations at the time of origination, you're not going to see any increase or decrease in the back book associated with macro environment. You're going to see it just be a function of the front book or decisions you make going forward. And as Max mentioned, we're very mindful right now. We're definitely 2 hands on the steering wheel, really thinking about every day with how much risk that we're taking.

Great. Well, let's shift gears a little bit towards partnerships. Affirm extended its partnership with Shopify. We've seen some really good momentum in Split Pay. Could you talk a little bit about how much more room and opportunity there is for the Shopify-Affirm partnership? How are you thinking about expansion into other products beyond Split Pay, which Shopify like long-term 0s and interest-bearing?

So I think on the GMV basis, you can -- I think we disclose exactly where we are. But if you squint, it's not hard to do the math it's pretty early. There's a lot more to do there. And both by way of Shopify still growing very rapidly as a company, as a merchant base, but also because Split Pay within Shopify is gaining traction still at a pace that's faster than Shopify is growing. The -- we've long promised to roll out more than just Split Pay to Shopify. That's now the case. So Adaptive checkout exists on Shopify. It's definitely the case, not just Shopify, but across every one of our partnerships, that these things take a while to deploy, not because of a technical reason, but because you have to optimize both the conversion. The end-to-end conversion depends a lot on everything from time it takes for us to make an underwriting decision all the way to user interface placement. Adaptive on Shopify is a first of its kind in a sense that it's an API-driven adaptive. So normally, when we deliver it out to some merchant, we say here's a box of pixels, we will help you optimize the perfect financial transaction. It's the efficient frontier cost to you and conversion to consumers. In case of Shopify, it's the same product, but it's driven fully by our APIs and their user interface. And so the optimization is actually a dual company affair. They have their design standards. We have our underwriting standards. There's a lot of things that have to go right. And so before that is in full and total deployment, we will take the necessary time to make sure that it's financially accretive to both sides and obviously accrues to the merchant and importantly do not push consumers into more depth that they can handle, but they had a nice benefit of very, very, very short duration. It's easy enough to tell for making too many loans or to maintain too much risk. Adaptive goes quite into the long term. The next thing on the horizon is things like long-term zeroes, et cetera. And so with every partner, this is true [indiscernible] more than Shopify. We take the necessary time, especially as we go into longer terms because we know that at our scale, the cost of screwing it up is just prohibitive. And we know the world is watching, so we're quite keen on delivering good risk numbers. But vast majority of our large platform partnerships are in absolute infancy, where we just have so much more with the job, as Michael put it.

Babies.

One thing that I think is really exciting that I don't actually think it's enough airtime is BNPL as a tool for lead generation and also brand-sponsored promotions. Could you just expand a little bit on some of the product ideas when you're thinking about this opportunity and when that will be available to the public? On the last earnings call, you also talked about eventually building out a self-serve platform, which sounded really interesting to me. What does this all look like?

Actually, the 2 are intertwined, so I'll talk about both [ BSP ] and -- a quick look with you on terminology. So lead gen is a little bit of a dirty word, in my mind at least, and it's because no merchant wants to know that their buyer is going to get sold off as a lead to potentially competing merchant. That's an easy way of running it out of your contracts and the very least the spirit of the contract. Nobody wants to be clear that there -- and we have some competitors that are unabashedly, hey your shopper is my marketing lead, so back off, I'm going to sell it. I think it works in really select pockets of merchants that don't expect to be in business for too long, while they don't care if they even comes back to them because they might be nothing to come back to. Most of our customers, most of our partners are businesses that have been around a very long time, and they expect to be for a very long time. And so we're very, very careful not to turn into a lead gen business. What we do have is a lot of opportunity for engagement and discovery. Our consumers come to us to pay their bills. And in the case of Peloton, it's 39 monthly opportunities to speak to the consumer. It's obviously a lot of communications that we send out, telling them that their auto pay is about to trigger, telling them that they might be delinquent. They have all sorts of comms and services that we own and operate. A large percentage of our volume originated on our own services where consumers come in to pay their bill. They say, "Oh, you have a whole catalog of all the merchants." We've now introduced very, very nascent, but still pretty cool, if I do say so myself, search box, where you can actually search for merchants that -- what an amazing idea. People like want to search for prices that have Affirm supported on the site. It was not a thing for quite some time. And now we said maybe now that we have thousands, maybe we should tell people where to shop or at least help them find where Affirm is accepted. And sure enough, we're now generating more than a million of these searches a month just in our own app. And that's a huge opportunity to not so much say, hey merchant X, hey you should pay me for that lead. But to say, hey, merchant X, I am sending you someone who is fully pre-approved. They know their limit. They wanted to see you because they searched for your brand. It's worth a lot to that merchant that's a brand new customer with spending budget that has been given to them, [ courtesy ] Affirm that they're quite excited to go to work. And so that's a very interesting business opportunity for us in itself. Now the really interesting barrier of this idea is when the consumer then says, "Oh, yes, I would love the 0% loan, but the merchant that I'm going to is actually a broad line merchant. They're not going to be manufacturing their own things. They're just reselling cool brands. Something that our scale allows us to do is we can go to these brands and say, "Hey, if I send a shopper as they search for a retailer and they choose to buy your item, would you be willing to basically subsidize their interest?" For a lot of brands, there's a great opportunity to introduce someone to a brand that they haven't tried or just a chance to push a new model or a model that they no longer want to stock. So there's a really amazing potential for marriage of consumers that are keen on buying something specific to a brand at a retailer that wants to give them an extra special deal. And we do this today already at places like Walmart, where you can get everything from tires to TV, where the brand is actually subsidizing the interest of the consumer. But we think there's a lot of opportunity for that, specifically within our own app as consumers are actively telling us what they're into. So that's a major area of focus for us. Super nascent. I might joke about who are the [indiscernible] that searches a thing. It's only half joke. We found out that our consumers were interested in searching on the app. So we're there.

Right. I mean it does feel like there's some sort of gap in merchant discovery and diligence outside of maybe some deal sites and aggregators and sort of things like that. So...

Yes, I think there's a keenful and thoughtful way to do it, and there's a way to sort of turn it into a kind of a marketing hammer and our DNA is definitely very far for a marketing hammer. So we will be very thoughtful and careful. And ultimately, we aim to provide confidence and control to consumers. Our job is to say, yes, the TV you really want is, in fact, available to you on the terms we can really understand at this retailer that you're looking for.

Great. I do want to open it up to investor questions if there are any. So if you have a question, raise your hand. We'll get a mic over to you. And while we're gathering up those questions, maybe I'll sneak one more in. Let's talk a little bit about the competitive landscape. U.S. BNPL competition is intense. Some very large VM providers have been very aggressively expanding into the United States. Some things have changed. There's Apple, as you mentioned before. So how do you think about competition in the U.S.? Is that a good thing because BNPL awareness needs to grow? Is it a bad thing because it just makes it tougher for unit economics? What's your general view?

So at the moment, the space is massively underpenetrated in the U.S. So in terms of -- the game has not come anywhere near close to zero sum. So that on the margin, that's just a good place to be in. We're still very much educating the consumer. The business is really cool, and it converts really well to the merchant. So that's the backdrop. Competition has changed towards a lot more rational thought, which is a nice, favorable change from my point of view on behalf of the providers. We no longer are facing off with people that are willing in small barrels of cash and offering merchants bribes to select them because they can't compete on value. The competition around value is something we're very comfortable in because we are the only engineering shop. I think pretty soon and maybe now, we're certainly not waiting, the competition will become about can you innovate, can you build new things that are profoundly different and add value to both consumers and merchants in ways that they didn't expect. BNPL is a 10-year-old idea. And we -- you can -- I think after paying the term, I'd like to believe we reinvented the genre after people who came before us in the U.S. Didn't matter who made it. For 10 years, that existed more or less in this form. My primary time as a product manager is spent on Debit Plus, which I think is the coolest thing that we've built and will be majorly important for our consumers and for merchants and for all sorts of other constituents because it is the everyday spending device. It is something that once you understand how to use it, which we still have to fully sort of breach the comprehension gap because it's such a new idea; lot of consumers kind of go, "I like, I love it, but I don't understand, exactly what it is " until we sort of do the right thing, explaining it to them. Once people start using it, it has near-perfect retention. So we've clearly hit something profoundly important for this customer. I don't think it's for everyone. I'm not sure my mom is ever going to switch her Debit Plus from her Black Card. But I'm not trying to sell it to everyone. I'm trying to sell it to the generation that wants confidence and control in their finances. And so building new products, delivering new ideas to both the consumer and merchant is where I want the competition to be because I have a lot of confidence that we're quite good at that. I filibustered long enough. But if there are questions from the audience, Michael will answer them.

Great. Well, that leaves an opportunity to have a follow-up on Debit Plus. The transactions per active, I think last disclosed was 2.7x weekly. Incredible numbers, both off-line and online. Strategically, what is Debits Plus? Is it an engagement driver, a profit driver, both? And is there anything you could talk about as it relates to timing?

It's engagement driver. It's ubiquity driver. It's a whole new set of challenges. It has things like insufficient funds, which is not normally the risk we have to run, but it's now a thing that we have to care about as a completely new frequency model, which means what used to be a velocity -- so a simplistic way of managing credit is, for example, if you have velocity control model says if you're overspending, maybe something is wrong, you need to slow it down. For giving somebody a daily spend instrument, you better get out of their way when they're trying to use it daily. And so velocity controls don't work anymore, you have to build a very different model. So there's a lot of things you have to get right, but the visible reality of Debit Plus consumer is if somebody who thinks Affirm can replace all other cards in their life, it's a de facto top-of-wallet product. It's a -- we're just about to start adding features to it that are beyond kind of the very basic BNPL that we launched it with. And it's a software programmable card, not a credit card. It's not a debit card, it doesn't exist. I spend a lot of time talking to both the regulators and various folks in which we are trying to explain what this thing is. It's like much better than debit, but not quite as toxic as credit cards. So we tried to build it out, but it's fundamentally a way for our consumers to just fully embrace Affirm way of life. The rollout thing, we have 14-ish million consumers that we think all want it, love it, deserve it. We have a massive wait list of people that raise their hand and said, give me one of these to me as soon as you have it. We don't get a lot of shots to get it right. If we deliver it and the compression is like, I don't get it. It was kind of cool on the brochure, but I'm not sure where to use it or it just doesn't work because we had a blip somewhere, we will all regret because this is the biggest opportunity that's in our very immediate future anyway. And so we're being very, very deliberate as we roll it out, but all the metrics that I'm seeing today about it make me very excited about what it's going to be.

Right. I'm sure the innovation road map and the product road map certainly helps with that competitiveness to differentiate yourself among other BNPL providers. Can we talk a little bit about the regulatory environment? In the U.S., obviously, there's a CFPB opening query around BNPL issues like credit reporting, data privacy, consumer merchant protections. Can you frame what the regulatory landscape looks like today and how Affirm is positioned to address them, perhaps even relative to some other providers?

Obviously, the CFPB inquiry is ongoing, so too early to comment on something we know not very much yet. We were highly engaged and excited, frankly, to educate the regulators about how we do business. We count ourselves in -- among the fortunate few, where we literally have absolutely nothing to hide. We respect consumer privacy to the extreme degree. We don't sell data. We are extremely careful to be more than just a little compliant with every applicable law, and then we take the extra mile to make sure we actually feel good about our decisions. You've seen our products. We disclosed our entire fee schedule on a spreadsheet full of 0s, and I think that compares us very favorably to the industry. So on sort of do you have anything to hide? Have you done something untoward? The short answer is no. We're quite excited about the way we've been running our business since inception. We never had to clean up something that we weren't proud about. There are a bunch of things that I think are going to be truly beneficial for the industry, most importantly, a unified push by the regulators to report performance of the shorter term. So we already furnished everything but as product, which is the BNPL paying for a 6-week loan, in part because the credit bureaus just have no idea how to ingest it, and there's a fair amount of bottoms-up pressure on the bureaus and other participants to ingest this data correctly, I think regulators coming out very clearly saying, hey, this is the way you do it, and that's what the expected behavior will be based on performance would be excellent because it will just set the record straight for a lot of people. And once again, I feel very comfortable in our ability to do better than average on expectation in that domain. In terms of the rest of the regulatory landscape, again, I think it goes a really long way that the closer you look at Affirm, the more you realize that we are profoundly on the consumer side. There's just not -- there's plenty of people that had a bad merchant experience. Merchant didn't ship something, and merchant screwed up. We stepped in and tried to help them out. But ultimately, if you look at the core of who we are and what we do for the end customer, be it the consumer or the merchant, but ultimately helping both, and we do it with great gusto for positive outcomes for all involved.

Great. Well, we're a couple of minutes over. That was fantastic. I wish we had more time. Would you please join me in thanking Max and Michael for their time, thoughts and insights.

Thank you.