$NU：Nubank 是1994年的 Capital One，还是2006年的 Capital One？| Vanshap Capital

Evan, how’s it going?

It’s going great, Andrew. Thanks so much for having me. I appreciate it.

I’m excited to talk about the stock today. The stock we’re going to talk about is pretty popular in FinTwit circles, I would say. When I was prepping for this podcast, one of the first things I do now is have my AI pull all the fund letters it can find from Fiscal AI’s fund-letter database. It said, “Hey, there are 87 different write-ups of this company.”

But I think you have a really interesting angle on it, and you’ve done a lot of work and thought about the historical parallels. Anyway, the company is Nubank. The ticker is NU, and I’ll toss it over to you. What is Nubank, and why are they so interesting?

Thank you again for having me. I’m excited to talk about this company. We’ve owned it and been studying it for almost 4 years now.

Basically, Nubank is a digital bank. It’s one of the largest in the world. They have about 140 million customers as of yesterday’s earnings report. Their biggest market by far is Brazil, although Mexico and Colombia are coming up nicely as well.

It was started back in 2013 by 3 founders. The largest of the founders is the CEO, a gentleman named David Velez, who is actually from Colombia. He grew up there, and out of his frustration with opening a bank account in São Paulo in 2012, Nubank was born. There were armed guards, bulletproof doors, months of paperwork, and so on.

The idea he had at the time was a branchless bank using mobile, which was coming up very quickly in Brazil, a zero-fee credit card, and, most importantly, a fanatical customer culture focused on NPS and really pleasing the customer in many different ways.

Over the last 13 years or so, we would argue that the results have been absolutely staggering. They have about 60% of the adult population as total customers. Most of the referrals they get—about 80% to 90%—are organic, so their marketing costs are extremely low versus the incumbent banks.

Most importantly, they’re extremely efficient. We can get into it, but the efficiency ratio is about 20%, compared to the incumbent legacy banks, which are more in the 40% to 60% range. Over the roughly 4 years since we’ve been studying the company, it went from being barely profitable to extremely profitable.

Because of the operating leverage, which we’ll get into, ROEs are now, as of yesterday, in the mid-30s. To bring it forward, particularly this year, there were some macro concerns, some credit concerns, and some transitions that the business has gone through that caused a very significant decline in the stock price. That’s why we think it’s particularly timely to discuss the company today.

No, look, great overview. I’m laughing because you said the founder tried to open a bank account in São Paulo, and there were armed guards and all this paperwork. I’m one for efficiency, and whenever there’s something inefficient, I hate it and I want to fix it. But this is just me, a mere mortal—not a man who started a multibillion-dollar banking startup that’s taking over all of Latin America.

When I go out and see armed guards, I think, “Okay, armed guards are part of the problem. I’m not touching that.” But this guy starts it.

There’s lots to touch on there, and lots that we will touch on, but I guess to start, let’s frame it at a high level. The market’s a competitive place, and this is a finance startup, right? They give lots of details. I believe I was reading an article that said, “This is the company that, out of all the fintechs I’ve ever followed, gives the most disclosure on all sorts of different KPIs and everything.”

This is a company that gives a lot of disclosure. As I mentioned, there are about 70 fund letters mentioning it as a top holding, so it’s a very visible, big company now. What are you—and I suppose the other funds—seeing that you think the market is missing that makes this an alpha opportunity?

It’s a great question. To your point, there are all kinds of analysts on the sell side and buy side covering the name. I would say that we still meet quite a few investors out there who are unfamiliar with it or categorize it as an emerging-market, early-stage, not-as-profitable, risky bank—however you want to say it.

Again, acknowledging that it’s definitely become more well known over the years and is now a roughly $70-billion market-cap company, it’s still, in our opinion, not fully understood.

Most importantly, I would say it’s the view that we have over the long term. There’s a view out there, aside from the short-term noise, that the company has already gotten 60% of the customers—or, sorry, 60% of the population—in Brazil. Obviously, you’re getting to the point of maturity there in terms of how many customers you can get.

I think the market is missing that, while they’ve spent the last 5 years getting those customers, they’re going to spend the next 5 years deepening those relationships. Again, we can get into this, but there’s a whole maturation of those customers over time. If you run those numbers out in Brazil alone, you come up with a far more profitable enterprise.

Then there’s Mexico, which I think, correctly, the market has been extremely skeptical of. Until recently, we thought that business had the potential to be a similar size, if not bigger, than Mexico over time. They mentioned yesterday on the call that Mexico is benchmarked back to where Brazil was in 2020. For instance, revenue per customer—the ARPAC per month—is double where it was in Brazil at the same point in time.

You have that, and then we can get into Colombia, which is obviously smaller. I’m sure at some point we’ll touch on the U.S. opportunity. To summarize, we’re taking the 10-year view, whereas many of these other folks may not be taking the same time-horizon view.

Let me pull on a few points there. I want people to keep in mind the deepening of the customer relationship point that you mentioned. With banking, Robinhood, and all these things, you get the customer relationship, and this is why they give you incentives. You get one of those emails saying, “Sign up for this credit card and we’ll give you $150,” or, “Set up direct deposit and we’ll give you $200.”

The relationships tend to be very sticky, and that’s their customer-acquisition cost that they plan on making back. As you said, they’ve grown quickly and spent a lot to grow. They’re hoping, “We spent to acquire all these customers, and we’re going to get them to deepen the relationship.” They took the credit card, and then they’ll get the bank account or whatever.

Your write-up—and anyone who's followed this company—has a lot of parallels to the Capital One story. I think people point to that positively because Capital One has been a good story in banking. I'd love for you to go through the parallels, and then I'm going to follow up with some questions on the bear cases that could emerge from that.

Yeah, sure. I guess a few things there. Maybe it's helpful to give some background on David Vélez, if that's all right, because I think it's a core part of the story: David Vélez and the co-founders. As I mentioned, David grew up in Colombia and had a very entrepreneurial background. His father, I think, started or co-owned a button factory, as we understand the story, where David worked in quality control, checking the buttons.

He apparently bought 1 cow and grew it to a herd of 6 cows before he sold the cow—or the cows, I should say—and went to Stanford. I won't bore you at this point with the rest of the story. The point is that he had a very entrepreneurial background. When he was starting Nubank in 2013, or 2012, thereabout, he brought in Sequoia, which was an investor in the seed round. I believe they invested $2 million.

Another big investor, to your point about Capital One, to close the loop here, was QED, the fintech firm, and Nigel Morris, who was a big backer of the company at the beginning. From then on, a lot of folks from Capital One came and were involved on the data science side, on the board, obviously at the governance level, but more importantly in the operational side of things. The parallels between the 2 companies are really, as I say, the focus on data science.

Like Capital One, Nubank would start with a very basic credit card and a very low limit, with a huge reliance on data, algorithms, and models constantly updating. As the customer matured and behaved, they would increase those limits. I would say that's really the genesis of where they came from and the similarities between the 2 organizations over time.

Okay, perfect. For those of you listening, I'm sure most of my listeners are domestic, and most people know Capital One. The Capital One story is that these were the smartest guys in the room. They started in the 1980s and figured out segmentation in credit cards, as well as direct mail: if you have somebody switch their credit card balance, that's actually a really undervalued thing. Credit cards are super profitable, and that's how Capital One got started.

The reason I think Nubank is interesting is because Capital One had a killer stock after the IPO in November 1994. The stock was up about 13× over the next 12 years. It crushed the market. In 2001, I'm just looking at my Bloomberg and glancing at it, they were approaching 4× price-to-book value and all that sort of stuff. This was a great growth company, trading at a huge valuation with all that profitability.

Then what happens? They find this loophole, but two things happen. A lot of the companies they're competing with, because they're offering a financial product, start doing balance transfers, too. Chase and all these guys get involved. It gets a little bit more competitive. People start hiring the Capital One executives and saying, “Hey, come over here and bring all that data science to us.”

The market becomes a little more competitive, and they kind of grow and start taking so much of that market that they have to expand. They buy a bank and start doing deposits. What happened since 2006? I'm just looking at it: from 2006 to today, the stock does fine. It's up about 3× over 20 years, which is fine, but the market is up about 8× over that time.

It crushes the market for the first 12 years, but then it kind of grows into a mature fintech, and it doesn't really do that well. It's trading around tangible book value today, and if you held it, you're kind of disappointed. I'm sure people can see where I'm going with this, but when I look at Nubank, I see this company with 30% ROEs and great growth over the past few years.

In Brazil, I can't remember off the top of my head, but I think about 60% of the population has a relationship with them, and they're into the teens in terms of market share. I can say, “Hey, that all sounds great. Let's put international aside for a second.” But when I look at that, I say, “They're Capital One in 2006 at this point, right?” The future is a lot slower growth, ROEs coming down, and expanding into a lot of areas.

That's my first worry when I look at the story. I threw a lot out there. Hopefully, it was a fun little story and trip down memory lane for anyone who remembers Capital One, but I'd love to hear what you think of that as the pushback.

Yeah, it's a fair pushback. Obviously, our thesis going forward does not rely on so many more customers, particularly in Brazil. Mexico, the US, and Colombia are different stories. To frame it up, I think we look at the Brazilian profit pool—and I'll somewhat answer your question by touching on Capital One—but I'm not as deeply familiar with that business as I am with Nubank.

The profit pool generally in Brazil is about $100 billion of gross profit in the banking system, and today Nubank is at about 7% of that. If you look by different product types, for instance, they just got into payroll lending, which is actually the largest part of the credit business in Brazil—payroll loans—and they are now at about 1% of that. If you go down the different verticals, or different product lines, not only in Brazil but in the other countries, it's hard for us to see how they don't grow materially from here, frankly.

Again, I'm sure we can get into it: barring some downturns and crises, there will be competition. There is competition from other fintechs. There are a few thousand fintechs in Brazil alone, I think, so there are other players. I would argue that, at the end of the day, Nubank has reached the point where it's so well known and the brand and reputation are so strong that, when it goes into these other verticals, it has a distinct competitive advantage.

Again, we can touch on the data advantage that they have. There is no question that they've picked the lowest-hanging fruit in Brazil, but our argument is that there's tremendous opportunity not only in Brazil, but also in the other countries where they're just getting started.

Most importantly for our thesis is David Vélez, and the fact that he and his co-founder still control the company and own about 20% of the stock. Particularly in banking, as Munger used to say, I think you really need to trust the banker. Whether it's an insurance company or a bank, you really need to have someone you can trust.

The fact that he's still in his mid-40s, controlling the company and making sure that the culture and growth remain intact, gives us tremendous comfort and optimism going forward.

That is just wild. You look at this, and as you mentioned, this is a $70 billion market-cap company that he started after Capital One. Capital One is the model. I mean, Capital One today is, I think, less than $70 billion in market cap. This is just an incredible growth story.

But let me go to valuation. They reported this morning, and the stock is up about 10%, so that's probably pretty nice for you. Oh, Capital One's market cap right now is 140 billion. I was looking at—I had their 2007 market cap on it. We've learned this before, but I obviously still think you believe there's a long way to go. People can listen to the whole podcast, but as we're talking, the stock is trading at about $15. That puts them in the high teens to about 20× price-to-earnings on an LTM basis.

This is a business that earns a 30% ROE. My worry is that I talk to a lot of investors who are US-domestic-based, and whether it's a bank, a tower company, or whatever, they'll point and say, “Hey, look at this great company I found. Peers in the US trade at 10× EBITDA. This trades at 9× EBITDA, and it's growing faster, so it should trade for a higher multiple.”

I always think to myself, “Well, maybe you're right, but domestic”—I had this with VEON recently—“emerging-market companies should have a higher cost of capital.” So what is the right answer?

Here you have a company that's earning a 30% ROE, trading for 18× to 20× earnings, and growing quickly. If this were the US, you'd say, “Hey, this should trade for an infinite price-to-book multiple with those metrics,” right? We're talking about how much of the TAM they can capture before they start slowing down.

Brazilian banking—I don't know. What's the right cost of equity for a Brazilian bank? If it's 15%, they're still earning a ton, but that's way higher than a US bank. If it's 45%—which, if this were some emerging market beyond Brazil, I don't know—they'd actually be destroying capital.

What’s the right cost of equity? How do you frame that valuation? How do you think about all of that?

It’s a great question. I think at the core, maybe some of it is philosophical. When you’re investing on one of these leading edges of a disruptive business model, as we believe Nubank is—not solely in Brazil and these other countries—I’m not sure anyone has the right answer to the question.

I would say we would argue that this is more of a tech company than a bank. Obviously, if you go back and look at the history of banking with branches, they’re not even in mortgages, basically. If you look at what a traditional bank balance sheet looks like, it otherwise doesn’t look much like Nubank, although obviously there are loans and deposits.

I think it depends how you frame it up. The way we frame it up is we look at 2029 or 2030. We think the company can make about $2 in earnings. To answer your question on the multiple, we use roughly a 16-times multiple, which I think is fair.

You can also look at, for instance, Itaú in Brazil. Itaú, coming back to today, is one of the 5 legacy banks. It has an ROE of about half or less than Nubank’s and is also trading at 7 times 2029 earnings. So, to go back to your question, I don’t think anyone knows exactly the answer. The company and we ourselves would say it’s more of a tech company than a bank.

I think time will tell how true that is. So far, I think it’s been a fair argument. As for your question about Capital One and whether they can continue to grow at this rate and produce these kinds of returns, I think that argument will be more true 10 years from now than it is now. But we shall see.

No, it’s interesting, because every other word you say there, I find myself either nodding and thinking, “Yep, I agree,” or going, “No, I think that’s crazy.” It’s just, is it an emerging-market bank? That is the key debate here, right?

Let me try to frame it a different way. The 2 public-market companies that come to mind when I hear this as loose parallels, aside from Capital One and the question of whether this is Capital One in 1994 or 2006, would be Mercado Libre, or MELI, and Kaspi. I think MELI is a little bit of a different business. They have more of a retail focus as well, but they do have a fintech segment. And Kaspi obviously owns Kazakhstan.

Both have been catnip for investors in a similar way to this. You can look at fund letters and see smart investors discussing them. Both haven’t worked out as well as I think they would have. Both are still growing nicely, the ROEs on both are nice, and I think the answer there has been, first, there’s always this regulatory axe that people are worried about with an emerging market. We can talk about the Brazil-specific Nubank one in a second. But I do think there is something to the idea that people just aren’t going to give you a big multiple. They’re always going to worry about that because of this cost-of-capital issue and everything.

How would you look at Nubank if you’re investing today? You’re doing it because you think it’s going to be alpha. Kaspi and MELI haven’t performed poorly over the past 5 years; they just haven’t generated tons of alpha. What do you think is different here versus those 2 peers? And if you can also say, “I haven’t done any work on those, but I do think the framework is at least helpful.”

Coincidentally, the 2 companies you mentioned are our other 2 large holdings, so I know them well. I guess, if I can, in the most respectful way, push back on your assertion—

No, please. I’m just asking questions.

On Mercado Libre, I think, just to touch on that, we again think that’s a phenomenal business. You mentioned the 5-year look-back, and that is absolutely true. I would say that the valuation, in our opinion, back during the COVID-19 period in 2021 was extremely rich, and now we would say it’s undervalued. If you go back to the IPO—what is it, 27 years ago?—I think it’s compounded in the 20s.

I had to stop it at 5 years, because if you go to 10 or 15 for MELI, it looks really good.

There we go. So, our view is that the next 5 or 10 years will be back on track. As it relates to Kaspi, I would say it’s somewhat similar. The valuation got quite high just prior to the Russia-Ukraine war, and the multiple also contracted from, I think, a peak of 19 times down to a bottom of around 6 times.

In our opinion, both of those businesses, including Nubank, are firing on all cylinders. There have been short-term issues that we could get into another time with Kaspi or something like that. But I would say those business models, over the long term, are remaining intact.

Our view is that Nubank has a similar competitive advantage with the data they have, their scale, and so on—basically, constantly pleasing customers, adding more and more products, going into more and more geographies, and so on. The short-term summary is that the short term is a question mark, but our view is that all 3 of the businesses we’re discussing should have a tremendous opportunity over the next 5 to 10 years.

Let me go back to one more thing you said earlier. You said, “Nubank in our mind isn’t a financial company so much as a fintech,” like an internet company. People can look at the metrics they publish, and they publish lots of metrics—more than most companies. You can look at their better loan underwriting and all this sort of stuff.

A frequent thing I hear from people, just in researching this, is that these guys are actually too conservative on their loan underwriting. They would be more profitable if they actually increased their loans and did some more bad loans, which is not something you hear a lot, right?

But I did hear you say, “These guys are branchless, with better technology.” It does remind you, as you mentioned earlier in 2021, of SoFi and a bunch of these branchless banks and fintechs that were coming along and saying, “We’ve got better tech. Consumers love us. We’re simpler.” Robinhood has probably worked out okay, probably the best of them. But that was the pitch, right?

What happened? The young people flocked to that. But guess what? It’s really competitive when you’re giving out a loan, because what do people want when they want a loan? They want a sleek interface and all this sort of stuff, but they also just want the lowest rate. If you’re lending at a 30% ROE, maybe it’s because you’re more efficient, but there’s always going to be an incentive for someone to undercut you to stay in business.

With their market share, why isn’t this on a path similar to Capital One in 2006? They’re so big they can’t take much more share without the banks really pushing back aggressively. The banks look at that and say, “All right, it’s existential for us now. We need to cut costs.” Or even the regulators could step in and say, “Nubank, you’re the 5th-largest bank in Brazil. We need to cap your lending because you’re becoming a systemically important financial institution that’s kind of too big to fail.” There’s a lot there, but I’d love to ask about that forward look.

I’m trying to pick it apart. As it relates to focus, you mentioned SoFi and some of the other fintechs in the US, and Robinhood. Obviously, those have done fantastically well. I would point out that they are not nearly as focused as Nubank, having all different kinds of businesses and transforming and so on. That has obviously worked for them, but in my opinion, it is not the same focused economic model that Nubank has.

As it relates to pushback by the regulators and so on, it’s important to keep in mind that Nubank has brought a lot of its customers out of the unbanked population. As you can imagine, in Brazil—and it’s also the case in Mexico and Colombia, and actually even here among the Latino population in the US—many folks have no credit score and no access to the banking system. Their ability to use their data to underwrite, even in a small way at the beginning, is very powerful.

For that reason, and for many reasons in Brazil in particular, they have a very fanatical following. So, the idea of the regulators or the population going against them anytime soon seems unlikely. But anything is possible.

I would also say, as it relates to the loans, you mentioned the efficiency. The fact that they are, however you want to frame it, at least twice as efficient, if not more—for instance, I don’t think I mentioned that they have about 13,000 customers per employee at Nubank, whereas the legacy banks have about 1,300. Again, on that metric, it’s almost 10 times as efficient.

What that allows them to do is underprice—not on every product, but as it relates to the private legacy banks in Brazil. They’re able to underprice them because of that efficiency.

So, again, over time, I’m sure there will be competition. You see it in the different markets, but our argument would be that even that competition is going to be difficult for them. More likely, those folks will be eating away at the legacy banks—the legacy banks in Brazil, Mexico, and the other geographies.

That is crazy: 13,000 versus 1,300. Though, you also don’t know whether there’s mortgage servicing on one or the other, but that is just wild on the face of it. Let me ask: Brazil is their largest market?

Yep.

Let’s put the other markets to the side for a second. Are there any risks around Brazil—macro, regulatory, anything—that you worry about with this company?

Absolutely. Just to be quite clear about that, it’s Brazil. As you and many of the listeners know, there’s normally a crisis there every few years or whatnot. There has not been one in a handful of years, so obviously we’ll see over time. I would say there are a few concerning things.

One would be, for instance, that 90-plus-day NPLs in the banking system are at 15-year highs. Part of the reason for that is that the Selic rate, the interest rate in Brazil, was up in the mid-teens. They’ve just recently started to bring it down, but as you can imagine, that stretched the customer to a large extent.

The average customer in Brazil is heavily indebted. If you speak to Brazilians, they’re getting credit card offers in the mail every 5 seconds, along with calls for new credit cards and so on. There’s no question that the consumer in Brazil is stretched from a leverage perspective.

There’s also a presidential election coming up in October. Our view is that the market is pricing in another Lula win. Obviously, Lula is the president now, so things have been going just fine for Nubank. But to the extent that there’s political turmoil of any kind later in the year, that could cause a problem.

We could go into some of the mitigants, and I guess I’ve touched on a few of them, as to why we think that even in a more macro-stressed situation, Nubank should do just fine. But yes, for sure, it’s an emerging market at the end of the day, and it has a history of some very volatile periods.

No, that’s part of it. I want to start talking about some of the emerging markets—Mexico, where I think they’re furthest along, then Colombia, maybe even the U.S.—but if I put those emerging markets and everything else to the side, you kind of mentioned that when you run this out to about 2030, you look at it valued at 17 times 2030 EPS, which does not assume any multiple expansion, right? So, you’re valuing it all on growth. But when you talk about that 2030 number, how much value are you baking in for Brazil versus the expansion markets?

We don’t count the U.S., and obviously the company thinks there’s a pretty decent shot that they can earn significant market share. I have to go back to the numbers, but Brazil is obviously still a very, very large part of it. We’re assuming that Mexico continues on its trajectory. Colombia is a small part but, again, is earlier stage, similar to Mexico.

We sort of continue to model all 3 countries at their current rates. It’s not like one becomes more overweight. From a profitability perspective, Mexico just crossed over breakeven in the last few quarters. Obviously, that will start to become increasingly meaningful for the company as operating leverage starts to shine through.

But it sounds like Brazil is the scary one. Obviously, Brazil is the largest, and Mexico is growing, but it sounds like the vast majority of your forecast—maybe you’re not even doing it by geography at this point—but the vast majority of your value is still coming from core Brazil?

Yes, that’s right. Particularly under the assumption you’re mentioning, we have ARPAC—which is the contribution margin, effectively, per customer per month—going from roughly 17 up to, I believe, the high 20s. The mature cohorts in Nubank are already in that range, and the incumbent legacy banks are more in the 40s.

There are some good reasons why they structurally should be higher, because they have additional services that Nubank doesn’t currently offer. But yes, that’s the correct answer, and that summarizes your point.

Let’s go. They’ve got 3 expansion areas: Mexico, Colombia, and the U.S. Mexico has just crossed profitability, while the other 2 are drags because they’re investing in the startup businesses. If you want to say anything specific about any of the geographies, you’re welcome to, but my overarching question is this: With a banking-focused fintech going into separate geographies, I know their whole thing is, “Hey, we’re going to be the first one that actually does this all on a core technology platform,” but I can’t remember a single banking-focused fintech that has successfully gone from its geography into any other geography.

Every other geography has unique rules, unique relationships, unique regulatory regimes, and so on. I think that has served as a massive barrier. Citibank is famous for this, right? In the ’90s, they said, “Hey, we’re going to be the global banking center.” Terrible underperformance for 25 years: write-off after write-off after write-off after write-off after write-off.

You can take it forward to pretty much any fintech you want—all the buzzy ones from 2021. I can’t remember a single one that was successful in the expansion. So, my question would be twofold. Mexico is already past breakeven, but why do you think this could be successful? And given the upside that they have in the Mexico and Brazil markets, does it even make sense for them to spend time diversifying away from this?

Yeah, all great questions. We could spend an hour just on the philosophical part of that. But I would make one important point that I didn’t touch on earlier: Part of the reason for the recent underperformance in the stock is that there were a handful of management departures, particularly people who were very highly regarded by investors, including the CFO.

One of the reasons, in our opinion—or maybe the main reason—those folks left is that David has been trying to take the company in a global direction. Part of that is obviously finding people who have built global businesses. Rob Livingston, the new CFO who joined the company recently, for instance, was the CFO of Visa North America. He had spent a lot of time—I think 18 years or so—at Capital One. There are folks like that. The new chief product officer just came from Shopify.

To answer your question, the risk there is not lost on the management team, number one. As it relates to the U.S. and going beyond the current geographies, they’ve mentioned that they’re going to cap the investment for the next few years at 100 basis points to the efficiency ratio. They’re very aware that they don’t want to sacrifice the golden goose by going into all these geographies and risking the company.

At the same time, you look at Mexico, and investors were saying the same thing about that 5 or whatever years ago. I guess that comes back to our earlier discussion about the philosophy. Going back to Citibank, as you mentioned, in the ’80s or ’90s, how was that managed? Admittedly, I’m not an expert on what happened there, but I would say that a big part of it comes back to the management team here and the thoughtfulness they bring.

At the end of the day, they acknowledge that. The chief product officer acknowledges, for instance, that the app will be different in each market and so on and so forth. Not only that, but the credit metrics themselves and the data are going to be different. You have all kinds of different data coming from the government, not coming from the government, and so on and so forth.

The answer is that I think they’re going to do it in a very thoughtful, calibrated way. If it’s not working, they will stop doing it—that’s our hope—and deploy capital or return it to shareholders.

I do believe that, at the end of the day, the business model of going into a given geography and having a far more efficient, far more attractive product for the customer is still, although it’s now been around for 12, 13, whatever, years, relatively underpenetrated. Digital banks around the world are still a small minority of the banking system, and our opinion is that over time they should be the whole thing. So, that’s really the question.

The nice thing about it, I believe, is that they bought back shares in the first half of the year. Given that the stock is up to 15, and I think they bought them in the low 12s, the nice thing about having ROEs in the 30s is that if your stock is trading for 15 times earnings and your ROE is 30, you can have enough capital to be profitable, buy back stock, grow the core business, and grow the emerging markets. You can invest in everything when your ROE is 30. It’s kind of nice. That works.

Let me pull on one thing you mentioned there. Again, I’m just trying to think this through, but if I told you, “Hey, I’ve been looking at this bank that’s doing great ROEs, and they’re letting go of all their people with local-market expertise and banking backgrounds and hiring people with global fintech backgrounds,” that sounds nice—we’re going to take over the world—but I’m also really worried that I’ve seen this before.

You hire the people and say, “Oh, we’re making this big push,” and then say, “Oops, we lost control of our core business.”

So, how do you think about managing that when, as you said, the reason the stock might have been weak in the first half of the year is that a lot of the core guys are leaving? Nobody likes to see the CFO leave at any type of fintech or banking platform. A lot of the core people are leaving for people who, yes, might be more well-known globally and might have more global experience, but they don't have the local banking and core Brazil experience.

Yeah, great question. I would say—and I didn't mean to overstate this—that the folks who left were at the very top, and we don't get any sense that the core business, the actual operating people, or even the people below them were departing. In that sense, it's not like the whole Brazilian team is walking out the door or something like that.

They've restructured it so that, I believe, there is now a CEO of each country who is a local or at least intimately familiar with the business, and then there's a LatAm CEO. So, it's more of a reshuffling and a restructuring. One could argue on the other side that, as I understand it, there were too many layers of direct reports coming up to David Vélez from the different businesses, and now that has actually been streamlined.

I think the direct reports have actually shrunk down. His argument would be that now he's closer to the business and keeping a closer eye on it. Again, those folks who are, for instance, doing the credit underwriting in Brazil are still doing the credit underwriting in Brazil. It's not like they've hired some team from Iceland to do that or something like that.

So, I guess time will tell. But I think that, so far, what we've heard—and again, in the last podcast that the company put out with Rob Livingston, the new CEO, David Vélez mentioned, “What are you going to be focused on?” We very much liked Rob's answer, which was, “I'm going to continue to focus on Brazil, make sure the focus of the company is on Brazil, and not on the next 10 markets where we're going.”

Again, I think time will tell, but we're optimistic that this is being done in a thoughtful, prudent way.

Slight pivot. David Vélez is a very—I mean, look, he started a $70 billion bank in his mid-40s, a highly regarded executive, and owns a ton of this stock. He just joined the board of OpenAI. So, I guess I'd ask you 2 questions. How much of Nubank, in your opinion, is a valuation bet?

We mentioned a high-teens multiple for a 30% ROE business that's growing quickly and has a lot of room to grow, versus a business bet. We mentioned KPIs that are off the charts compared with all their Brazilian competitors, that it's way more efficient, and that it's probably the right to win, versus a jockey bet.

David Vélez is super well regarded. He started this business and has deep ties throughout a lot of the tech circles. How do you think about this overarching question?

Yeah, it's a great question. The way we frame up the world is that we're always looking around for what we call “customer fanatics.” I guess the simple answer to your question is that we like to think every investment we make is, at the end of the day, a jockey bet.

I would say that, in the case of Nubank, it's particularly a jockey bet—not only because of his control of the company, but also because of the way he treats shareholders. A few years ago, he had a compensation package that I think would have added another few percentage points to his ownership if he had hit certain metrics, and he decided to get rid of that, I believe, for more or less free because he didn't think it was, at the end of the day, fair to shareholders.

There are other similar moves that support, in our opinion, that view. On the multiple, as we touched on, I don't think you really need multiple help here. I think David Vélez just continuing to execute in the way that he has—and, as we mentioned, he's still quite young—matters.

The longer we're in the investment business, the more important the jockey becomes. Questioning the jockey's every single move is worthwhile, obviously, as it relates to going into certain countries and other geographies. But so far, the way they've built this business and the results clearly lead us to believe that he is an exceptional talent.

While one can question each move, it's probably best, at the end of the day, unless one has a good reason not to, to let him continue to build the business. I'm not sure if I fully answered your question there, but I think I answered it.

Let me ask 1 more question about him. I don't know anything about him. He obviously is super highly regarded, and people can hear that throughout the podcast.

There's 1 interesting thing: He just joined the board of OpenAI. You can correct me if I'm wrong, but I thought that was interesting because, on the 1 hand, it speaks to being plugged in and connected. I'm sure there are boards that are more prestigious to join right now, but I don't think there are many. That's a really cool thing, and it speaks to his connections. Obviously, you mentioned the Sequoia connection and all that sort of stuff.

On the other hand, to say, “Hey, you've got this growth fintech; you're the CEO, and you own a ton of it”—when I look at a lot of the customer-fanatic founders you mentioned, like Mark Zuckerberg or Jeff Bezos, I think Mark Zuckerberg and Jeff Bezos do not sit on outside boards, right? Jeff Bezos is retired, and he still doesn't sit on any boards outside Amazon. I'm sure anyone would want him.

Steve Jobs sat on a few, but I don't think many, and I think most of them were direct complements to Apple when he was there or Pixar when he was there. So, how do you think about David Vélez joining OpenAI in the context of it seeming like that might be a bit of a distraction? I just think it's a weird thing for a fanatic with all of his net worth tied up in this business.

Yeah, I would agree that there's a risk of that. Yesterday, one of the analysts asked him about that on the conference call, and he basically said that it's to the benefit of Nubank. I think there's probably a large element of truth to that.

Being on the board of one of the largest LLM and AI businesses in the world would provide a lot of insight. The other thing—and we could dig into it now—is how much AI is helping Nubank become smarter.

So, when you're competing with legacy players and you're the fintech-enabled player, you have to think that AI is going to be massive for them. If you exclude the idea that the government is making sure 5 players get cybersecurity protection, you have to think AI is going to be massive for them because they're just going to be so much faster and so much more efficient. Sorry to cut you off. Please.

No, no. Maybe to transition, I guess you're right. Normally, if we saw what we would consider a customer fanatic—and, to your point, the vast majority of his wealth is in the business, and he's obsessed with pleasing the customer and so on—then the question is, why would he be spending whatever hours per month on this?

To touch on some highlights, something like 60% or more of customer inquiries are now handled by AI, without human interaction, on the first pass. So, there's a tremendous cost savings there.

David Vélez has talked about this becoming an AI business, in the sense of AI helping on the financial-advisory side and on all kinds of things. They talk about how just because you have a bank account doesn't mean you know what you're doing as it relates to that. There's huge value in AI simply helping the customer.

On the business side, I think they're using Cognition's models, and they've talked about the Devin AI agents. They recently put out a blog about 12x efficiency gains in engineering hours and 20x cost savings on the implementation of what's going on inside the business.

They also talk about how credit scoring and the data used to take months to process, so they would update their model every few months. Now it's down to something like a handful of days. Basically, AI is changing all parts of the business, including customer service and credit.

They can obviously, therefore, write much more credit, with much more security in terms of understanding the risk of that. To bring it back, we would normally be skeptical, but I would say that the amount of impact AI is having on the business is so incredible that, in this case, my guess is that he heard about this opportunity and said, “This would be silly to pass up.”

The insight that he's getting from this relationship and his integration as part of the board would be of huge value to Nubank itself.

Let me ask you a broader question on AI as it relates to fintech, because there's nothing to me that's more of a commodity than a financial product. If I have a credit card and I'm looking for a credit-card loan, you offer me 19%, you offer me 18%, and I go with the 18%. I don't really care about anything else.

Checking accounts, all this sort of stuff. Now, maybe that's a little bit too far because if I need an ATM, I want ATMs close by or something. But really, financial products are about as commoditized as you can get. With a mortgage, you really don't care; you want the lowest mortgage as long as you can send a check in.

I've worried about all of them. I've thought about the US banking system. It's built a lot on zero-cost deposits, right? People probably remember back in March 2023, there was the regional banking crisis and the whole flight of cash. Everybody was worried that all the cash was going to shift to high-yield money markets and destroy the entire business model. That didn't really happen, but I've worried about AI. The best place I can think of AI being, in terms of consumers using it to maximize their lifestyle, would be: “AI, optimize all my finances. Get all of my checking accounts into the highest-yielding things with US FDIC insurance. Any place where I'm paying interest on a card or mortgage, go find me a cheaper one if you can, and do that constantly.”

So, if I brought it back to Nubank in the short term, I bet they're going to be so much faster than a lot of their peers at adopting AI. But in the long term, that results in supernormal profits for 6 months, a year, 2 or 3 years—I'm not sure. In the long term, if everybody adopts it, I'm worried the profit pool is going to zero and that 30% ROE we talked about—everybody just competes it away because AI does everything.

That's a more high-minded, maybe longer-term thought, but it's something I've thought about with US banks. As you said, it made me think about Nubank. So, I'd love to ask how you think about that.

No, that's another excellent question. I would say there's a lot of points there, so I'll try to remember them all and get them in sometime.

AI's impact on the entire financial system—is it a 2-minute sound bite?

Exactly. So, I guess, number one, at your point about commodity, I would agree that, at some level, that's true. I would say that if you want to make that assumption, then the most important point in a commodity business is to have the lowest cost, right? To be the most efficient—the last guy standing.

As we've discussed, I think, at least as of now and for the foreseeable and potentially very long-term future, no one is going to be more efficient than Nubank. In terms of efficiency, and therefore being able to offer, to your point about credit cards, 1% lower rates, they should be able to offer the most efficient rate because of the scale they have and because they're branchless. Again, that's a whole other conversation about how hard it is to close these branches, change these legacy banks, and how slow that process can be.

So, I think the number one point is just the fact that they're so efficient and therefore the lowest cost. Again, if you want to make the assumption of a commodity industry, I would push back on the pace of change. There are, what, 5 banks in the US that control a massive percentage of the banking system, and they pay almost nothing on deposits. Yet still, people—including myself—keep money at these banks.

As it relates to AI suggesting that people pull their money out, again, I think if you are offering the best rates and so on and so forth, there's an element—and I don't think that this will change, probably even in our generation—where, even with FDIC-backed accounts and so on, you want security. You want, obviously, good customer service, and if those things are being achieved—the low cost and customer service—I don't see why you would change.

Even with someone's AI suggesting that you do so, I don't see any time in the future or the near future where that would change. I would say those combinations of variables, along with the customer service that they have and the low rate, mean I don't see why someone would go somewhere else.

Let me ask one last question, then I probably need to run to the bathroom, to be honest with you. You mentioned Mexico; I get it. Colombia, I get it. You know, the one expansion where I look and say, “Ooh, is this kind of too much?” is moving into the US.

If you just told me, “Hey, there's an emerging-market player that's moving into the US in fintech,” I'd be like, “They're too— they've got too much of an inflated ego.” The US is very, very well served. I'm not saying there aren't underbanked people, but there are lots of very juicy startups that are really attacking these problems, lots of really good banks with great offerings, and it's just kind of like, “Hey, why are they moving in here?”

They've got so much on their plate and so much opportunity to expand. And when I see that we're moving to the US, it kind of brings me back to the earlier question where I said, “Hey, I haven't seen fintechs expand successfully.” I might believe it if you told me Mexico and Colombia, but when they move into the US, I kind of say, “Oh, maybe these guys just think they're god-kings and don't understand the competition they're going up against.”

I know people have mentioned Banco Santander moving into the US. I don't think that's been a screaming success. Barclays—they're in, they're out. The US is just really competitive, and there are really good banks, whether it's at the large scale with relationships. People say, “Oh, I'm going to move in,” but it's not easy. You have to hire a team. You have to go acquire all those customers.

Maybe if you're in a very inefficient market, like it seems Brazil was or maybe Mexico, okay. But if you're in a market that's as deep as the US, with this many competitors, it seems like you're asking for failure. So, I just want to get your thoughts on that expansion because they talk about it a lot, too.

They do, yeah. Again, I think that is a source of anxiety for investors today. I would say a few things. One, as I said, they've decided to cap, correctly or incorrectly, their investment for the next few years. So, it's not like they're betting the company.

I also should note that, as I said, in our valuation and our view of the future, we're not counting anything, nor are we subtracting anything, to your point. So, obviously, if the management team decides to dump billions of dollars here and not get a return, then obviously that would be a reduction in our valuation estimate.

And so, it's a massive market. Something like 1 in 3 Hispanic folks in the US are unbanked, basically. They're not just going after the average person in the US; they're going after a very specific segment. Being here in Miami, the co-founder, Cristina, moved here in the last year or so. They're making a huge push here, I believe, and probably in Texas and California, where you have a huge immigrant population from Brazil, Mexico, and so forth. Many of these folks are familiar with Nu, so I think that's part of it.

The other part is that, I don't know about you, but again, with Chase Bank across the street or one of the others, you go there and have to wait in line, and so on and so forth. These branches take forever to do this and do that. While maybe the customer service is better than it was 5 years ago, I think Nubank thinks that they can provide far better service digitally, with a much better user interface and so on, and potentially, over time, at a more attractive cost.

So, we shall see. Again, I think the verdict is out. They seem to be doing it in a very thoughtful way, bringing in people who know what they're doing, and it's a very, very large market. I think they've said that, over time, targeting 5 to 10 million customers would be a good scenario. Again, they're not assuming that they're going to become a 10% market-share bank.

We shall see. It'll be a very interesting experiment. To your point, others have tried, and it has not necessarily gone well. But at a high level, we're quite optimistic about how they're going about it. If you dig into the niche that they're focused on, I think there's a pretty good shot that they can make some progress here in the US.

No, I certainly hear you. You know, I just like the Robinhood app; it's really damn good. I remember looking at Robinhood when they were really down, and I was like, “Oh, man, I don't know.” But the Robinhood app is really good. For all the stuff that Chase has—the ATMs—and Capital One, I mean, Nubank's model of Capital One and Chase—these apps are pretty darn good.

Yes, if you go in person, it's slow, but going in person is also there for when you need to prove your identity or when you need cold, hard cash. Most of the other stuff can be done online. So, if you're saying, “Hey, I wouldn't compare Nubank to the in-person experience; I would compare Nubank to the app,” I'd say that the apps are pretty damn good across the board.

Anyway, I am getting old, and it's Friday, so I had some Diet Dr Pepper today. I have to go hit the restroom. But anyway, last thoughts you have on Nubank or anything you want to leave listeners with?

No, I think we've done a good job covering the highlights.

I think it’s a name that a lot of people recognize and that has grown a lot. Our view is that it has a lot more room to grow by deepening the relationships in Brazil and Mexico and so forth. As we discussed, we’ll see what happens with the U.S., but I would say that, considering the valuation has come down materially, you’re buying into a situation with a customer franchise at a very attractive price that, over the next 5 or 10 years, could and should produce very attractive returns going forward. But time will tell.

Perfect. Well, Evan, thank you so much for coming on. I really enjoyed this conversation. Great discussion on both Nubank and the overarching fintech theme. This has been a really fun conversation, so we’ll have to have you on again, and we’ll chat soon.

Thanks, Andrew. I look forward to it. Have a great weekend.