Nubank Aktie

Good evening, ladies and gentlemen. Welcome to Nu Holdings' conference call to discuss the results for the first quarter of 2026. A slide presentation is accompanying today's webcast, which is available in Nu's Investor Relations website, www.investors.nu in English and www.investidores.nu in Portuguese. This conference is being recorded, and the replay can also be accessed on the company's IR website. This call is also available in Portuguese. [Operator Instructions] [Foreign Language] [Operator Instructions]

I will now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer at Nu Holdings. Mr. Souto, you may proceed.

Thank you, operator, and thank you, everyone, for joining our earnings call today. With me on today's call are: David Velez, our Founder, Chief Executive Officer and Chairman; and Guilherme Lago, our Chief Financial Officer. All financial metrics discussed and presented today reflect our managerial P&L framework, which we introduced in our fourth quarter 2025 results. These managerial measures are important to how we manage the business, but are not financial measures as defined under IFRS and may not be comparable to other companies. The full reconciliation to the most directly comparable IFRS figures is available in our managerial P&L reconciliation report and in the appendix to this presentation.

We are aware that consensus estimates across the sell side reflect the mix of IFRS and managerial frameworks, and we encourage everyone to use the reconciliation report as a reference point for aligning models going forward. Unless otherwise noted, all growth rates discussed today are presented on a year-over-year FX-neutral basis. Today's discussion may include forward-looking statements, which are not guarantees of future performance and involve risks and uncertainties. Actual results may differ materially from those expressed or implied. Please refer to the forward-looking statements disclosure included in this earnings presentation for additional information.

With that, I will now turn the call over to David. Please go ahead, David.

Hello, everyone, and thank you for joining us today. For several years now, our results have followed the same earnings-generating formula: a growing, more engaged customer base, monetized at higher ARPAC with a scalable, low-cost platform translating into outsized earnings. The first quarter of 2026 was another clean expression of that model.

Our customer base now stands above 135 million customers. In Brazil, we surpassed 115 million customers and solidified our position as the largest private financial institution in the country. In Mexico, we crossed 15 million customers, becoming the third largest financial institution in the market. And in Colombia, we delivered another solid quarter of net additions and are getting close to 5 million customers.

Despite typical first quarter seasonality, consolidated monthly activity rate held at 83% and expanded sequentially. In Brazil, we're approaching 100 million monthly active customers. Customer growth, combined with ARPAC expansion, which has expanded sequentially every quarter since we began reporting and now sits at around $16 per active customer, compounded into record revenue, reaching $5 billion for the first time in our history. The higher revenue translated into strong operating leverage in the quarter, leading to a record low efficiency ratio below 18%, a result that reflects both structural progress and some timing benefits that Lago will unpack shortly. This is happening despite our laying the foundations for our international expansion and accelerating an AI transformation that I will come back to in a few minutes.

On the credit side, 3 things: seasonality, growth, and mix drove higher provisions. This reflects our ability to continue gaining market share with compelling and resilient unit economics and do not suggest any signs of asset quality degradation in our portfolio. Understanding this difference is key for those following high-growth, credit-led fintechs.

We delivered a quarter 1 historical high net income of $871 million, compounding at more than 80% a year on an FX-neutral basis from 2022. With that as a backdrop, let me start with our biggest market, where we still have a long road ahead of us. Brazil is, by any measure, one of the most attractive banking markets in the world. Across just the products and segments we serve today, the addressable profit pool already exceeds $100 billion in annual gross profit and is expected to keep showing healthy growth for years to come. As we expand our product shelf and deepen customer engagement, the profit pool becomes even larger.

Even after a year of meaningful share gains, it's still day 1 for Nubank in Brazil. Our share of that pool stands at roughly 7% even though we're already the largest private financial institution in Brazil by customer base with the strongest brand and the highest customer satisfaction scores.

And in our second largest market, the runway is even bigger. The opportunity in Mexico is in many ways where Brazil was a decade ago. The profit pool of the products we want to serve consumers with already exceeds $40 billion in annual gross profit and is growing faster than most major banking markets in the world.

The banking system in Mexico remains structurally underpenetrated. Cash still dominates everyday transactions. Less than half of adults hold a formal credit product, and a meaningful portion of the population still lacks access to banking. Our share of that profit pool is still below 1% today, a fraction of where we are in Brazil and a fraction of where we believe we can go.

What makes this opportunity particularly compelling is the dual dynamic at play. We're not only taking share of the existing pie, we're also helping grow it, bringing simple, digital, transparent financial products to broader segments of the population that have historically been left out of the formal banking system. That combination is what gives us such a long horizon ahead, and the proof of that thesis is already starting to show up in the numbers.

The same earnings-generating formula I described at the start of our remarks is now unfolding in Mexico, only earlier in its curve. In 4 years, our customer base there has grown from just over 2 million to 15 million today, roughly 7x larger. ARPAC has nearly doubled even as we have onboarded millions of newer, less mature customers. Our efficiency ratio has come down by 78 percentage points.

And on the bottom line, we have moved from a $30 million quarterly loss to our first quarter of IFRS profitability, a milestone that arrived ahead of our own internal plan. Underpinning our operations in Latin America, including Brazil, Mexico and Colombia, and where we believe will further accelerate our impact in the region for years to come is the AI technology shift I referenced at the start of our remarks. Our ongoing AI transformation is a core priority of Nu. Some companies see AI as a productivity enhancement tool. That is useful, but it is not the real opportunity in our view. AI transformation is something different. It means we're designing from the ground up how financial products and services are manufactured and possibly distributed.

There is a parallel here to the bet we made when we started Nubank a little over a decade ago. We did not digitize a branch. We built a bank without branches. We're applying the same logic to AI. We're not just adding AI to banking, we are rebuilding banking around AI. This transformation is already underway and unfolding in 3 phases at different stages of progress. The first phase, AI assistance, is largely complete. We're reaching close to 100% utilization of AI tools among our employees across all functions of the organization. This enablement is driving productivity gains across the company, with engineering throughput up over 50% year-over-year, weekly token consumption nearly 10x higher than at the start of the year and testing cycles 90% faster.

The second phase, workflow reinvention, is in motion. The principle is simple. AI executes, humans hold judgment. Customer journeys are being rebuilt end-to-end, and new AI-native customer experiences will reach our customers this year, deepening engagement and expanding monetization. A number of teams at Nubank are already working on products and features that we had originally planned to launch only in mid-2027.

The third phase, the AI-native bank, is still early but the foundations are visible. AI Private Banker functionalities such as financial insights, payments, credit advice and debt resolution across the app are already serving more than 15 million monthly active users. nuFormer, our set of proprietary foundation models, are in production today for credit card decisioning in Brazil and Mexico and for unsecured lending in Brazil. We're now able to use real-time AI valuation for every personal loan request, priced and approved individually based on its predictive net present value in under 1 second. These capabilities have been a meaningful driver of the significant expansion in our credit portfolio over the last 12 months, enabling us to grow limits with resilience, not just speed.

And we believe Nu is uniquely positioned to win in AI-accelerated world, anchored by 3 structural advantages. First, our scale, first-party data, 135 million customers transacting on our platform every day, generating one of the largest, cleanest and most differentiated financial data sets in the world. Second, our proprietary technology stack, cloud native with core banking systems built internally, data unified across the company and the ability to move from experiment to production in days rather than quarters.

Third, our talent and culture, our world-class bench of employees for more than 50 nationalities with offices across 6 countries, all working under a single AI mandate, and one we keep reinforcing with the recent appointment of Carl Rivera as our new Chief Product Officer. AI is not an experiment at Nubank. It is reshaping how we build, how we decide and how we serve, and we're still very early in what this transformation will eventually deliver.

Taken together, this is the model we're running in 2026, deepening Brazil from a position of leadership, scaling Mexico and Colombia through their inflection points and making AI compound through every layer of the company, including investing further in our internationalization plans.

With that, I hand it over to Lago, our CFO, to walk you through the financial highlights of the quarter. Over to you, Lago.

Thank you, David, and good evening, everyone. Beginning with our consolidated credit portfolio. We ended the quarter with $37.2 billion, up 40% year-over-year on an FX-neutral basis and up 7% quarter-over-quarter. Growth was strong across all products, especially when the first quarter seasonality is considered. Credit cards, for example, grew 36% year-over-year on an FX-neutral basis. Unsecured lending grew 53%, reaching $10 billion in total portfolio. And secured lending grew 38%, keeping pace with the rest of the book and holding its 8% mix even with the setback from FGTS loans last year.

Now turning to deposits. Total deposits reached $42.4 billion in the quarter, up 22% year-over-year on an FX-neutral basis. Deposits in Brazil declined modestly due to seasonality, while Colombia kept growing. In Mexico, the deposit outflow reflects 2 specific dynamics. Number one, a sharper-than-expected reversal of seasonal year-end inflows. And number two, our deliberate decision to optimize cost of funds and very low loan-to-deposit ratios.

Now our consolidated cost of deposits closed at 88% of the interbank rate, slightly higher sequentially. Even though we saw improvements in the cost of funds in both Mexico and Colombia, this was offset by Brazil, reflecting the reversion of the fourth quarter seasonal effect. Year-end inflows tend to lend in short-tenure balances that carry low cost of funds. And in the first quarter of the year, these balances naturally migrate into longer tenure yield-bearing products. Now we remain very comfortable with our current balance levels and with our cost of deposits. We will continue to manage this franchise to build resilience, deepen customer engagement and preserve its attractive economics.

Moving on to our P&L. Net interest income reached a record $3.25 billion in the quarter, up 12% quarter-over-quarter on an FX-neutral basis. This expansion was driven by strong revenue growth across the franchise, combined with our credit portfolio expanding faster than our liabilities. This mix shift continues to optimize our balance sheet, lifting our net interest margin, or NIM, to 21.1%. Credit loss allowance, or CLA, closed at $1.79 billion in the quarter, up 33% quarter-over-quarter on an FX-neutral basis, mostly driven by 3 very specific dynamics already mentioned by David.

Number one, seasonality. Number two, portfolio growth. Number three, portfolio mix, which I will unpack in the next slides. As a result, our risk-adjusted NIM came in at 9.5%, down 100 basis points sequentially from 10.5%. We expect risk-adjusted NIM to move back towards the level we operated at during the second half of 2025 as the dynamics of first quarter normalize over the coming quarters.

With that, let me now turn to the 3 dynamics I mentioned that drove CLA this quarter and walk you through each of them. Starting with the first reason, seasonality. As you can see on the chart, our 15- to 90-day ratio is highly seasonal. It tends to peak in the first quarter and then resume its trend through the rest of the year. The first quarter 2026 print of 5%, up 89 basis points from year-end, is consistent with that seasonal pattern and broadly in line with what we saw in both 2024 and 2025.

On the right, 90-plus NPLs, our late-stage delinquencies continue to ease, closing at 6.5% in the first quarter of 2026, 10 basis points lower than the fourth quarter of 2025 and well below the 7% peak we reached in the third quarter of 2024. Both metrics came broadly in line with our own internal expectations for the quarter.

And I want to pause on that phrase because it's not incidental. The goal of our credit operations, it's not to minimize NPLs at a point in time. Instead, it is to optimize for resilient NPVs. NPLs only capture the cost side of the equation. They say nothing about the revenues we generated from the customers who perform. Pricing risk accurately is what really reconciles both sides. It is the mechanism by which attractive returns and predictable losses coexist. When the first quarter unfolds as our models anticipated, that is not a coincidence. It is an evidence that the pricing discipline is working well.

Now before we move on, I want to address directly a concern. We know it's top of mind for many investors. Brazil's household debt service ratio. We track this ratio closely, but the data tells a more nuanced story. The debt service ratio in isolation has limited predictive power over delinquency outcomes. What actually drives credit performance is a much broader set of income and employment dynamics.

Employment in Brazil remains strong, and the income tax exemption for earnings up to BRL 5,000 per month is a meaningful structural tailwind for a large portion of our customer base, directly improving disposable income and debt service capacity at the segment levels where we operate the most. And critically, as you will see in the next slides, our portfolio has a particularly short duration, which means that if we ever did see unexpected asset quality movements, we can react fast and we can react consequentially, and we can do so at a very granular level. Looking ahead, the Desenrola program is an additional tailwind expected to take form in the second and third quarters of this year.

Now to the second reason, growth. And what matters here is not only our credit portfolio, but our total exposure, a broader measure that includes the on-balance sheet credit balances and the off-balance sheet credit card limits we extend to our customers. Both things expand our IFRS 9 provisioning base. On the left side of this slide, you will see that total exposure reached $70.7 billion in the quarter, up 44% year-over-year on an FX-neutral basis. Every single dollar of incremental exposure carries upfront provisioning regardless of whether the customer ever draws on it.

And that brings me to the third reason, which is mix. On the right side of this slide, you will see that the incremental exposure we added this quarter tilted further towards credit cards and unsecured lending, which together accounted for 98% of the new exposure, up from 88% in the same quarter a year ago. Secure lending's contribution now stepped down, mostly reflecting the changes in FGTS loans at the end of 2025. And then both credit cards and unsecured lending, they carry higher expected losses than secured lending, which mechanically just lift the marginal provisioning we book.

Growth means a larger exposure, and mix means that the base is tilted towards higher yielding, higher losses products. Both things pushed the upfront expected credit losses build higher even before any change in underlying credit quality. Bringing it all together, the 3 drivers we just walked through: number one, seasonality; number two, growth; and number three, mix, are exactly what shape the moves in NPL 15-90 and in ECL allowance this quarter. There was no sign of credit portfolio degradation.

Let me walk you through each of those bridges. On the left side of this slide, the NPL 15-90 moved from 4.11% at year-end to 5% in the first quarter, an 89 basis points increase; 65 basis points came from seasonality, 17 from intentional risk expansions, 4 from product mix shifts and the small remainder from other effects. Now none of these drivers reflect the systemic deterioration in underlying credit quality.

Now on the right side of the slides, you will see that ECL allowance moved from $5.3 billion at year-end to $6.1 billion in the first quarter, an $800 million increase. The numbers here are worth pausing on. Why? Because portfolio growth alone contributed $423 million, more than half of the total build, which simply reflects the upfront lifetime loss provisioning we book under IFRS 9 as we expand the credit book. Seasonality alone contributed another $267 million, consistent with prior years. Together, growth and seasonality account for 86% of the entire allowance increase. Intentional risk expansion contributed $69 million; product mix, $16 million; and other minor effects, the small remainder.

Now not one of those components reflects deterioration in underlying credit quality. These moves reflect the deliberate scaling of our credit portfolio. As we said before, we manage this business not to minimize NPLs or cost of risk in any given quarter, but to maximize the long-term resilient risk-adjusted returns. We see that discipline at work in the cohort unit economics of our 3 most relevant unsecured credit products. Across all of them, revenues consistently outweighed funding costs and expected losses, leading to return levels that are best-in-class for retail banking.

With a significant buffer, these portfolios remain NPV positive even at substantially higher levels of expected losses. And the short duration of these portfolios is worth pausing on. Why? Because it means that if we ever did observe an expected asset quality movement, we can react fast and we can react decisively. And we can do so at very granular levels well before they become a systemic issue.

We are not a loan book lender waiting quarters and quarters to see the impact of a credit policy change. We see it in days, and we act on it immediately. That is what grounds our strategy. Beyond the unit economics, we also hold considerable buffers in the balance sheet. Our total coverage stands at 16.2% of the portfolio, roughly 2.5x our entire 90-plus delinquency balance. And we are adding to that buffer each quarter. Our gross CLA against the new 90-plus NPL formation closed at 153.8%, which means the provisions we book are running ahead of the new NPL forming. That is balance sheet engineered for resilience, and one that lets us grow the franchise from a position of strength.

That balance sheet resilience that I've just mentioned flows through the gross profit line, which closed at $1.88 billion in the quarter, up 27% year-over-year on an FX-neutral basis. This quarter's mix reflects the elevated CLA we just walked through, which directly reduced credit's contribution and brought float to roughly 40% of the total.

Beneath that quarterly seasonal effect, a multi-quarter trend of genuine diversification continues. Our credit business, our float business and our fee business have been scaling and balancing each other. And our model allows us to build a more diversified gross profit base and ultimately, a higher quality earnings profile overall.

Now turning to efficiency. With net revenues outpacing operating expenses, we continue to deliver operating leverage in the quarter. Our efficiency ratio improved this quarter to 17.6% on a reported basis and 16.6% at the core, which excludes our return to office investments, our international expansion and our investments in AI infrastructure.

The first quarter came in better than expected for 2 reasons working together. Number one, revenues accelerated faster than we anticipated, driven by both ARPAC outperformance and continued portfolio growth. Second, OpEx came in below plan. And here, it's worth pausing to discuss why. Roughly 1/3 reflects structural efficiency gains that are durable and compounding, mainly AI-driven improvements in operations and collections, software platform consolidation and hiring discipline. Now the remaining 2/3 reflect timing items that will normalize in the next quarters, including real estate and marketing phasing. So the 17.6% efficiency ratio should not be extrapolated as our run rate.

But even accounting for those normalizations, we expect our consolidated efficiency ratio for the full year of 2026 to land at approximately 20%, broadly in line with where we ended 2025. And while our core efficiency ratio continues its natural downward trend, we remain confident in the attractiveness of our investments in return to office, U.S. expansion and AI infrastructure.

The positive effects of operating leverage and financial leverage continue to flow through the bottom line. Net income reached $871 million in the quarter, the highest ever for our first quarter and up 41% year-over-year on an FX-neutral basis.

Now I want to be direct about our effective tax rate, or ETR, because we know it may be a focus. The 8.7% IFRS rate this quarter reflects structural changes we have been making to our global operating and corporate structure. It is not a one-off, and it's not an accounting adjustment. It is a recurring structural feature of how we operate. The first quarter rate is naturally lower than our full year rate because it reflects some of the seasonal patterns we discussed earlier in this call.

For modeling purposes, we expect our IFRS ETR for the remainder of 2026 to converge towards the 15% to 20% range. Our managerial ETR, which we believe is the more economically meaningful comparison, should converge towards the 30% to 35% range, which is broadly in line with peers in the region.

Now the broader point is this. We are absorbing intentional investment headwinds in the OpEx line, and those are being more than offset by structural improvements in our ETR. The net result is a net income trajectory that remains durable and compounding, which is the right lens through which to assess the earnings power of our business.

Now to wrap it all up, this was another quarter that demonstrated the durability of our business model. Number one, a growing and engaged customer base. Number two, an expanding credit portfolio, growing profitably and resilient. Number three, a more diversified gross profit base. And number four, one of the strongest balance sheets in financial services.

With that, I will pass it over to David for his closing remarks.

Thanks, Lago. Nubank is incredibly well positioned to continue strengthening its place as Latin America's leading digital bank. While our consumer base is large, our total market share is still small. And that gap represents a long and visible growth runway in our core markets. This remains our #1 priority.

But we continue to have conviction that the digital banking thesis we started to execute in 2013 is a global thesis, not a local or regional one. First principles reasoning shows our advantages [ travel well ]. Our cost structure is 20x to 30x more efficient than the incumbents that still own 90% of the world's banking market.

Our technology gives us the agility to move fast in any environment. Our differentiated approach to credit gives us the tools to compete and grow within a segment that represents over 70% of the world's consumer banking profit pool. And our consumer obsession allows us to build relationships with fans, not just customers, creating one of the strongest and most authentic consumer brands wherever we operate. That is why we are excited to be expanding our model to the U.S. deliberately and at a measured pace, treating it the way we treat every new market, as a call option.

We invest a relatively small amount of capital and resources while we protect our core. Once we see product market fit, we're ready to scale. To be precise, the maximum OpEx headwind we expect from U.S. investment in each of 2026 and 2027 is less than 100 basis points on our consolidated efficiency ratio. And this is inside the 20% efficiency ratio level Lago mentioned before. For a company at our scale, that is quite affordable.

Beyond that, any additional investment is explicitly contingent on clear evidence of product market fit and a credible path to profitable scalability. Even in a scenario where we do not find product market fit, the cost to you as a shareholder is less than 100 basis points on our efficiency ratio, temporary and fully absorbable without touching the trajectory of our core businesses. The upside, if we do find product market fit, is a second Nu. We have seen this movie before in both Mexico and Colombia. The asymmetry between a bounded downside and an uncapped upside is at the center of our investment thesis in the U.S. and potentially, the world. And it does not change our long-term trajectory on efficiency.

With that, let's open it up for questions.

[Operator Instructions] I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer.

Thank you, operator. Could you please open the line for Mr. Jorge Kuri from Morgan Stanley?

Congrats on the results. And really much appreciated the incredible detail around delinquency and credit losses and provisions and expenses and sizing the U.S., I think that's going to go a long way in helping people understand the story better. So thank you.

My question is on an announcement, and I really don't have any questions on the quarter, sorry about that, but I did see an interesting announcement in the press that you're launching an SME-specific product in Brazil. And so I wonder if you can maybe talk about it, what type of products, how they're different from what your competitors offer? What is the edge, the moat that you guys are using in SMEs? And dream the dream, what size can this business be to you overall?

Thanks, Jorge, for that question. I actually think this is probably one of the most underappreciated opportunities we have at Nu. We should be speaking more about it, but we're not. So thank you for asking the question. The reality is we've kind of silently have built the largest SME base in Brazil with our 5 million SME customers effectively built with 0 customer acquisition cost. Since Brazil has a very large number of SME, a significant percentage of employment in Brazil, something like upwards of 70% is -- operates in small businesses. A very large percentage of our 110 million customers have their own businesses. And so we were able to cross-sell our SME product to them, and that took us at a 0 CAC to build this base of upwards of 5 million customers.

We've invested increasingly in building better the product savings account. Initially, we now crossed over 2 million credit cards for the small business. We recently announced what you probably saw is now a number of new lines of credit, both secured and unsecured. Some of them are using some of the government-available programs where entrepreneurs are able to use certain government guarantees to get loans.

And we see a blue ocean in that space, really, the kind of comparative advantages that we have on the [ FS side ] on the individual side applies to the SME. This is a very underserved segment. And I think we're also -- while we began at the base of the pyramid with a very small micro entrepreneurs, we've been slowly going up the base and starting to serve companies that are -- have more than 10 to 15 employees.

So I'll leave at that for now, but we have an ambitious plan on that space. We -- it's -- there's a lot of scarcity in the entire environment. It builds a lot of loyalty with our consumers since they get to back both their businesses and their individual in the same place. And it brings pretty significant advantages on our flywheels.

Operator, could you please open the line for Mr. Yuri Fernandes from JPMorgan?

Everybody, congrats also here on the presentation, very clear. I have one regarding asset quality. And maybe 1 week ago, there was a podcast with Lago, [ Jeremy and Tyler ] to discuss asset quality. And I think one of the highlights in the podcast was how to read a good or a bad quarter for asset quality, right, especially regarding coverage formation and all those metrics.

And here in this quarter, it looks like, kind of an introduction for the quarter, in my view. Because we had a quarter that the company did a lot of provisions, right? The coverage went up. The new NPL formation when you look at the amount of coverage was over 150%. So my question here is, is Nubank being a little bit more conservative and building more reserves, maybe just to show the market that you have a very strong balance sheet? Or no? Or are you seeing a worsening outlook? Are you seeing something that we are not seeing? I guess, Lago already mentioned that he is not seeing a worsening, and the presentation was clear on that, but just reinforcing this message and trying to link with the past week's podcast.

Thanks so much for the question. Look, we feel that our balance sheet is fairly robust, and we try to be extremely conservative in how we build our provisions over time. But I have also to say that the provision that we have been doing over the past quarters, they do not reflect any directional outlook that we have on the credit cycle of each of the markets in which we operate.

The way that we have tried to do credit underwriting and consequently, to do credit provisioning is one where we always assume that the future will be worse than the past, irrespective of where any of us think we may be in the credit cycle. And then for each and every single cohort, we have the stress test whereby that cohort has to withstand a fairly material credit deterioration and still be kind of an NPV positive.

And then the additional disclosure that we are providing that may be helpful to address your question, Yuri, if you go to Slide 18, you can see the unit economics of our 3 core unsecured credit products, namely credit cards in Brazil, unsecured lending in Brazil and credit cards in Mexico, right? They account for the majority of our unsecured exposure. And I would underscore 2 things. First, if you go through the unit economics, it's kind of a healthy unit economics in our view. But more importantly, if you take a look at the ratio between losses and the net margin, you can see that they can withstand a lot of risk worsening and still being NPV positive.

The second thing I would underscore is the duration, right? If you take a look at the duration of each of those portfolios, we operate intentionally with much lower duration than the average of the market. This is a feature, this is not a bug. Why? Because it allows us to navigate with lots of agility at a very granular level.

All this to say that we are provisioning conservatively as we have provisioned in the past. Nothing has changed. We are not provisioning more or less because we have a directional view on the macro or on the micro, but we continue to provision and underwrite with what we believe to be a fairly healthy resilient buffer across every single segment that we do.

There are two things that I believe, Yuri, you have written extensively in your reports that we have not yet taken into account in credit underwriting and provisioning. The first one is the income tax exemption or reduction that has been announced in Brazil at the beginning of this year. That basically benefits consumers with up to BRL 7,400 per month of income. It may very well be a tailwind for us. It's hard to calibrate the magnitude, but that has not been taken into account in the provisions and credit results for.

The second one is the Desenrola 2.0 that we briefly mentioned and you also wrote about it. That, I think can be a fairly important kind of a renegotiation tool sponsored by the federal government for our customers, which we believe can be, for Nubank, either neutral or positive.

No. So basically, I guess, if I can summarize, first quarter is usually seasonal, higher provisions. Following the pattern, I guess, you mentioned this on your remarks, maybe cost of risk moves a little bit lower. Margins are higher. So risk adjusted, maybe the trajectory should be more positive going forward. Do you agree with the summary here?

I do. And I think we even mentioned a bit in the opening remarks, if you go through kind of Slide 14, Yuri, you can see that the risk-adjusted margin contracted from 10.5 to about 9.5 mostly because of the additional CLA in the first quarter, which does not indicate any sign of credit deterioration. And therefore, once seasonality goes out, you should see risk-adjusted NIMs converging back to the levels where it was towards the end of 2025.

Operator, could you please open the line for Mr. Eduardo Rosman from BTG?

Hi, everyone. Look, I do see local Brazilian investors today being much more constructive than the foreigners regarding the investment story, right? They are less concerned about asset quality, more positive on the expansion into the U.S. So I just wanted to hear your thoughts based on your conversations that you have, right? Do you think that this is because maybe local investors were the ones skeptical at the time of the IPO? Or -- and naturally, you delivered a lot, right? Or maybe -- I don't know, maybe foreign investors, they are more concerned about AI, disruption risk and maybe because they never saw a digital bank really succeeding at scale in the U.S. So trying to understand here, based on the conversations that you've been having with investors, if you can share your thoughts with us?

I wouldn't go as far as segregating kind of local versus foreign investors or Brazilian versus non-Brazilian there. But there are, I think, a few topics that are top of mind for many of them. The ones that I would highlight first is kind of asset quality.

So I think when Nubank was founded 13 years ago, the bank had a fairly strong thesis and hypothesis on its ability to do credit underwriting at scale throughout multiple credit cycles in Latin America, which is one of the most volatile regions of the world. It was a hypothesis we couldn't prove at that point in time. You fast forward the move 13, 14 years, and I think we can both in Brazil, in Mexico and in Colombia, already now clearly highlight that we have developed the ability in terms of process, systems and talent to be able to do credit underwriting in a resilient manner at scale.

And I think the velocity to which Nubank has been able to gain market share has now encouraged or impressed some of them. So I think the credit underwriting capabilities of the bank and concerns with asset quality will always remain and they should remain because for any kind of digital bank that has been able to attack credit, we will always have credit risk first in our priority list. But I think at this point in time, let's say, across most of the investor spectrum with when we speak, that has been more of a common theme.

The second question that I would say that is more polarizing is on our international expansion, specifically to the U.S. On one hand, Rosman, you do have investors that are extremely bullish on our ability to basically break into what is simply the largest retail financial services market globally, right? And there are key and relevant pockets of pain points on consumers there that a digital bank franchise can attract.

On the other hand, you have investors, they are more skeptical about this. At this point in time, we have deliberately chosen not to fully disclose the go-to-market strategy that we want to have in the U.S., mostly for competitive reasons. But I think what we can know -- and David tried to address this in his closing remarks, provide the comfort to investors is that we will be very deliberate and we will stage the deployment of capital and the deployment of talent and never putting at risk, our ability to execute in Latin America. So it's more of an attempt to balance the downside that hopefully will allow investors to more clearly identify the asymmetry of this bet.

And finally, the third one that I would say that it kind of has an even more heterogeneous assessment is the role that AI has been playing and will continue to play in digital banking or in banking in general, right? So a lot of companies have been talking about their efforts to kind of use AI. We have the first time, in the opening remarks of David, prove that we have been able to use AI to deliver impacts and results, not efforts. So a material growth of our customer base and credit underwriting hinges on our success to kind of embed AI across how we manage the company. A material improvement in our efficiency ratio hinges on our ability to fully embrace AI. And there's a ton of additional things for us to do, and we are very confident that we have the capabilities to continue on that front. So 3 points, Rosman: asset quality, internationalization and AI.

Thanks a lot, Lago.

Only thing I'll just add to everything that Lago said on the internationalization is that it's interesting that every time we've launched a new country, the locals have been skeptical. When we launched in Brazil, the locals were very skeptical. When we launched in Mexico, the locals were very skeptical. And the capital came from the foreigners.

And so sort of the same thing kind of repeats, sometimes being a local is a little bit of a blessing. Sometimes, it's a little bit of a curse because if you're a local by definition, it's very hard for you to reimagine how things can happen differently. You are too consumed by the status quo. So I definitely do not want to minimize the challenge that a country like the U.S. is going to be -- is extremely -- will be clearly challenging. There's a lot of very competent competitors. But we think we have an insight, and we'll see how that goes. The good news is that if we're wrong, it's a little loss. If we're right, it's going to be a huge opportunity for us.

Operator, could you please open the line for Mr. Daniel Vaz from Safra?

Congrats on the results, and thanks for the insightful presentation. David, in the present [indiscernible] agenda for the next years with AI transformation, Mexico and U.S. expansion, and so on and so forth. But let me ask you about Brazil. How specifically the team is looking at Brazil, right? So your incremental exposure is again on unsecured products. And all your competitors are trying to focus in on exactly the opposite, like they're trying to grow in secured loans, private payroll.

So I guess my question is, how should we read that, right? So is it -- we realize that we can extract much more value and returns from these unsecured products compared to our peers, and we'll try to focus the most on it and dominate the market, especially the mass market. And as a follow-up, how should we think about the secured products like the private payroll loans? So if you can answer that, it's very helpful.

Of course. And it's a great question. And that's why I think we wanted to -- if you go back to Slide 7, we wanted to -- this is a slide I would like to use maybe at least once a year to kind of anchor people on the opportunity. And to remind everybody how early it is, this story, even for us in Brazil, even though we're already -- in terms of number of customers, we have over 110 million Brazilians, and we're the largest private financial institution in the country.

But in terms of profit pool, we only have 7%. And I think the answer to your question is really the growth opportunities everywhere. In unsecured loans, there is a lot of growth ahead. We only have about 8% market share, but we have something like 25%, 30% market share of new originations every month. So we have a disproportionate amount of market share gains every month.

In secured loans, we are tiny. We started later, as you know, we've been kind of around for about 1.5 years, 2 years. Operationally, it's much more complex, especially on the public, what is called public payroll, [ consignado ]. There's been a fair amount of contracts that we needed to sign with the municipalities, and there is a fair amount of integrations that need to happen.

But if you look at the growth rate we're seeing in secured loans, it's growing pretty significantly as well. And we think that the growth in secured will be -- will continue to be even in something like FGTS, which we launched about 2 years ago, we became the largest FGTS provider, which is fully secured in about 18 months. Obviously, the product was restructured by the government.

But anyway, long way to say that the opportunity, the growth opportunity is in both, and we continue to see both. On private payrolls specifically, we've discussed that we've been slower at growing that, and that has been by design. Here, I think we have just a little bit of a different point of view than a lot of other players in the market. We think that this product began with more risk than people anticipated because there was -- there were a number of different points in the chain. And especially with integration with DataPrev and some of the providers that were untested. There were a lot of flows like what happens when employee goes from company to company that was completely untested.

And so we just took a more careful approach. And I think yes, we'll see how that goes, but we're seeing 10% to 15% for payment default. That's a very high FPD. That's a very high risk for supposedly, a secure product. We also decided not to put interest rate too high. We don't want to be charging too high of an interest rate for these products because they are supposed to be secured. And we thought that there was a lot of regulatory risk. And in fact, there is now a conversation about capping pricing, which is going to hurt more, the players that were too fast, pricing very high.

So we think in the long run, this is a winning product. We think in the long run, this is going to be great for customers to be able to have that security. And in the long run, we will also -- will stand ready to win this market with the same advantages that we have of data, of consumer trust, of cost to serve also apply for secured loans. But here, we decided to just be a little bit more careful and go a little bit slower as we measure them.

So long answer to say, there is no preference necessarily here. This is a wide open market. We're very well positioned to gain -- continue gaining share across the board, even in credit cards, which has been our first product. And so that's why we just say that this is sort of still the first minute of the first half in Brazil.

Operator, could you please open the line for Mr. Marcelo Mizrahi from Bradesco BBI.

I have two questions. First one is regarding the efficiency ratio. So you guys were saying that to target at range to achieve 20% efficiency ratio. So now we are below this level, just to understand, just how we can predict that. So how to forecast that looking forward? First one.

And the second one is about the private payroll. If you guys have any update in terms of the view of Nubank looking to this product and the possibility to this product to bring more clients or even some impact that could bring on the NIMs on the margins?

Thanks so much for the question. I'll take the first one, and we can maybe refine David's last response on the private payroll loan. But the first one was about efficiency ratio. So I will draw your attention to Slide 21. And you can see that we have had kind of a positive trajectory on efficiency ratio overall.

Now last quarter, or last call, we did mention that we were making deliberate investments in 3 fields: return to office, internationalization, and AI infrastructure. And those investments would be kind of a headwind to our overall efficiency ratio. And therefore, we wouldn't be able to get the same level of efficiency ratio gains over time that we saw over the past 2 years.

We still believe this is going to be the case, but I wanted to kind of unpack the performance in the first quarter a little bit more. So in the first quarter, you can see that we got kind of a 17.6% efficiency ratio. It was slightly better than even us expected there, but I would underscore a few things. First, about 2/3 of this kind of overperformance in efficiency ratio in the first quarter was mostly due to timing. What do I mean? It would mean kind of operating expenses that would be incurred in the first quarter, but will likely be incurred in the subsequent quarters of the years. Examples, some marketing investments, some real estate, then it will be tied to the return to office.

Now about 1/3 of the overperformance is truly structural. They are mainly coming from some of the operation gains driven by some of the AI investments that we are making across the board, from BPOs to software consolidation to enterprise functions. And those will continue.

The second thing that I would highlight, as we mentioned in the last quarter that we would start breaking down the efficiency ratio in two. One is the consolidated efficiency ratio, which we can see is 17.6%. But also is the efficiency ratio that we would have had, had we not decided to make the investments in RTO, internationalization and AI infrastructure, which in this quarter will be 16.6%.

Now going forward, I think one should expect that our 2026 efficiency ratio will converge towards approximately 20%, which is largely in line with where we landed last year. And this 20% envelope includes those kind of strategic investments that I alluded, both the RTO and U.S. expansions that David touched as well.

Your second point was on private payroll. I think I will -- David has covered kind of a little bit of the strategic reason on our choice to speed it up more or less. The one point that I would just underscore, if I got your question correctly is we continue to believe that as the lowest cost manufacturer of this industry. We will be able to provide kind of this product at very competitive levels. And we are very bullish about this product specifically because it will allow us to have access to customers and data that we have been unable to do as we don't have a corporate business.

So if you are today an employee of a large corporate in Brazil, most likely, that corporate has a payroll agreement with 1 of the top 5 incumbent banks of the country. And we historically have had some limitations on the amount of data that we could get from those customers by virtue of not being able to offer that payroll services.

Now with private payroll, we can have access exactly to the data, by which I mean how much money you make, for how long you've been working at the company, what's your expected severance cost. So we basically closed entirely, the gap that we could have had on that specific segment against incumbent banks. So we will likely drive more customer acquisition, better credit underwriting, better cross-sell.

Operator, could you please open the line for Mr. Tito Labarta from Goldman Sachs?

Great job addressing a lot of the key concerns with the credit quality and expenses. I guess my question, a follow-up a little bit, I guess, on credit quality. I think part of the concern also is your relative exposure to the lower income segment. So the question is more, how is the high income segment going? I think that's still a big opportunity for you as well. Anything you can comment on that?

And also one follow-up on the secured lending side because I know you're not growing the private payroll now. And you still had the headwind from FGTS in the quarter, the full quarter impact. So I was a little surprised with the strong growth in the secured lending quarter-over-quarter. I imagine that's public payroll, but just to talk a little bit about that opportunity on the secured lending side? Not just private payroll, but I think public payrolls and other segment where you have a lot of opportunity to grow.

Let me try to address some of your questions in order. So I'll start with what you call the high income, which I would mention kind of in the 3 segments that we have in the bank, so mass market, super core and the high income. Look, we have been kind of quite encouraged by the progress that we have made across the more affluent segments. Both the super core and the high income, which I think all our banks would probably core more of a mass affluent than the high income, which, just to be clear, those are customers who earn anywhere between BRL 5,000 to BRL 12,000 super core and more than BRL 12,000, the high income.

If you take a look at not only the number of products that we have been launching with the new UV credit cards, the [indiscernible], the 3 kind of international seen, the cashback, [indiscernible], so there's a lot of new products and features that we have been launching. And all of those things have been translating in not only more customers, but also more engagement, right?

So out of the high income, we now have about 2 out of every 5 high-income Brazilians are customers of Nubank. The customer base has grown by about 24% year-over-year based at the end of the first quarter of 2026, with now monthly credit card kind of volumes up 42% year-over-year, assets under custody by like 36%. So we are seeing lots of traction there. In super core, 3 out of every 5 Brazilians are already customers of the bank, again, kind of TPV and AUC all growing between 35% and 40%. So happy with the traction that we have had over the past now 2.5 years.

Now back to your question on credit exposure, you're absolutely right that still the bulk of our credit exposure is in what we call mass market. That's a bigger exposure than what we have in kind of the more affluent customers. And then when you look at Slide 12, you will see the evolution of unsecured and secured. And in spite of the headwinds from the new regulations of FGTS, you can see that we continue to grow secured, as you pointed out.

But I would underscore one thing, Tito. As the duration of the FGTS portfolio is relatively long, so more than 36 months, it means that even if we decrease the origination, it takes some time for that to play out entirely in the balance, and that's the fact that you may be seeing. Now to counter the slowdown in originations of FGTS, we are seeing an increase in the originations of public payroll loans, and we do expect that we will also see a pickup in the originations of private payroll loans. So I would not expect that the volume of secured cards will suffer too much throughout the year, irrespective of the FGTS regulations.

To your question, Tito, on high income specifically. We don't disclose the numbers particularly, but the growth of -- the PV growth for high income for us is one of the fastest segments that we have, growing upwards of 40%. So way faster, way higher than what we're seeing in mass market. And a lot of the benefits of these new models that allow us to give higher exposure. If you see the big growth in higher exposure is coming also disproportionately for being able to give better limits to high income population, which historically has been something that we hadn't really gotten, right, since our models were very much focused on mass market.

So from this exposure growth, there is a disproportionate amount of high income. And obviously, that's good news because this is a segment where we have a very large -- 2 out of 5 Brazilians with high income are already customers of Nubank, and we have a significant opportunity to continue growing that share and diversify the customer base that we have.

So thank you, everyone. We now have approached 60 minutes of the call. So we are now concluding today's call. On behalf of Nu Holdings, our Investor Relations team, I want to thank you very much for your time and participation on Nu earnings call today.

Over the coming days, we will be following up with questions received tonight, but we were not able to answer. And please do not hesitate to reach out to our team if you have any further questions. Thank you, and have a good night.

The Nu Holdings conference call has now concluded. Thank you for attending today's presentation. You may now disconnect.

Good evening, ladies and gentlemen. Welcome to Nu Holdings conference call to discuss the results for the fourth quarter of 2025. A slide presentation is accompanying today's webcast, which is available in Nu's Investor Relations website, www.investors.nu in English and www.investidores.nu in Portuguese. This call is also available in Portuguese. [Operator Instructions] [Foreign Language] [Operator Instructions]

I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer at New Holdings. Mr. Souto, you may proceed.

Thank you, operator, and thank you, everyone, for joining our earnings call today. With me on today's call are David Velez, our Founder, Chief Executive Officer and Chairman; and Guilherme Lago, our Chief Financial Officer. Starting with this quarter's results, we're introducing a new managerial reporting framework, including managerial indicators and our manager P&L.

All financial metrics discussed and presented today reflect this framework. Lago will provide additional details during his presentation. These managerial measures are important to how we manage the business but are not financial measures as defined under IFRS and may not be comparable to other companies.

A full reconciliation to the most directly comparable IFRS figures is available in our managerial P&L reconciliation report and in the appendix to this presentation. Unless otherwise noted, all growth rates discussed today are presented on a year-over-year FX neutral basis. Today's discussion may include forward-looking statements, which are not guarantees of future performance and involve risks and uncertainties. Actual results may differ materially from those expressed or implied. Please refer to the forward-looking statements disclosure included in this earnings presentation for additional information.

With that, I will now turn the call over to David. Please go ahead, David.

Hello, everyone, and thank you for joining us today. 2025 was a fantastic year for Nubank, and Q4 '25 truly showed the strength of our business model. During the year, effectively, most of our key indicators from customer law to scale, engagement and profitability moved in the right direction, while we continue to invest significantly on long-term growth.

We closed the year with 131 million customers adding 17 million net new customers and maintaining an activity rate of 83%. Scale and engagement remained the foundation of our model. ARPAC reached $15 per active customer, up approximately 9% quarter-over-quarter and 27% year-over-year, driven by deeper monetization across our platform. As a result of strong customer growth in higher ARPAC, revenues in Q4 '25 reached $4.9 billion, up 45% year-over-year.

Gross profit in the same period reached nearly $2 billion, up 38% year-over-year. At the same time, we maintained discipline with an efficiency ratio of 20% under the new methodology even as we continued investing in our core markets and new technologies. Net income reached $895 million, translating into a record 33% return on equity, while maintaining strong capital buffers and scaling our credit portfolio responsibly. These results reflect the priorities we set and the discipline of execution throughout the year.

One way to see this execution is to look at what we put in customers' heads. Across our markets, we launched more than 100 new products and features. More important than the number was the intent. Each launch aimed to deepen engagement to expand access and strengthen unit economics. Individually, these initiatives are incremental. At scale, they compound. In payments, we have both PIX with AI-enabled features, launched instant payments in Colombia and expanded Mexico's cash in and cash out network to more than 30,000 physical points.

In credit, we expanded responsibly launching new payroll loan modalities in Brazil, the subscription-based credit card in Colombia, and rolling out programs like fresh start to help engage customers regain access to credit. We also introduced the under 18 credit card beginning to build financial relationships earlier in customers' lives.

On the affluent segment, Ultravioleta continued to strengthen our value proposition. For SMEs, we scale credit products and launch tools like charging assistant to help small businesses manage cash flow. Behind this execution was a clear set of priorities, cutting our allocation of capital and talent throughout the year. As you may recall, our top priority is to build the largest and most loved retail banking franchise in Latin America.

In 2025, we made measurable progress across all three markets. In Brazil, we became the largest private financial institution by a number of customers, reaching $113 million with an activity rate of 86%. Scale and engagement continue to reinforce each other. In Mexico, we reached 14 million customers, advanced our banking license process and roughly half of our customers received their first credit card through Nu, reinforcing our role in expanding access to credit.

In Colombia, we surpassed 4 million customers and the subscription-based credit card significantly increased approval rates while maintaining healthy unit economics. In our digital ecosystem, we reached over 12 million unique active customers across initiatives such as Nu Cell, Nu Pay and Nu Travel.

Adoption remains early relative to our base, but growth and satisfaction indicators are compelling. On AI and global expansion, our foundation model, Nu Former is now in production for credit decisioning in Brazil and in testing across additional use cases. AI is already improving underwriting, conversion and service quality with PIX with AI surpassing 10 million monthly active users. In January, we also received conditional approval from the OCC for a U.S. national bank charter. Overall, we delivered on our 2025 priorities while strengthening the foundation for what comes next.

Let now turn to how we're thinking about 2026. As we enter 2026, we see this as an inflection year. The year we begin transitioning from a Latin American leader to a global digital banking platform. Our priorities are organized around three pillars. First, winning in our core markets. Brazil and Mexico will continue to absorb the majority of our capital and management attention. In Brazil, we will deepen leadership in the mass market, expansion of wallets and ARPAC, strengthening small businesses and grow or high income presence through Ultravioleta.

In Mexico, finalizing our banking license process is critical as it unlocks the next phase of credit growth and customer debit. In Colombia, we will continue scaling credit and bringing a number of Nu products. Across all three markets, our focus remains on experience, principality and monetization. Second, strengthen foundations for international expansion. During 2026, we will lay the operational groundwork for our U.S. opportunity, building on the conditional bank charter approval. Latin America remains our primary growth engine.

Third, AI as a superpower. We will expand Nu Former to lending in Brazil and credit cards in Mexico and continue putting AI directly into customers' hands, moving closer to our long-term vision of an AI-powered personal banker in every customer's pockets.

With that context, I'll hand it over to Lago to walk through the quarter's financial results.

Thank you, David, and good evening, everyone. Now before moving into this quarter's financials, I will briefly explain an evolution in our disclosures. As Nubank has become a multiproduct, multisegment and multi-country platform. We are introducing a managerial P&L to provide a clear view of value creation and internal performance. This evolution does not change economic reality. It only clarifies it.

The managerial P&L is derived entirely from our IFRS results and represent our structural reorganization of IFRS line items designed to enhance comparability and better reflect economic contribution. The framework preserves net income, cash flow, equity and regulatory capital and is fully reconciled to IFRS.

The key benefit is clear visibility into how margins, operating leverage and value creation evolve as the Nubank platform scales across multiple products, segments and geographies. And to support this new disclosure, we are publishing a detailed managerial P&L reconciliation report on our Investor Relations website, including the full bridge to IFRS and the complete methodology used. We have also updated historical data back to the first quarter of 2021 under this new framework.

With that context, I will now walk you through the quarter's performance already used in the managerial P&L. We ended the quarter with a total portfolio of $32.7 billion, up 40% year-over-year, driven primarily by credit cards and unsecured lending. Credit cards increased 12.2% quarter-over-quarter. This was the strongest quarterly growth since the end of 2023. This reflects continued limit expansion in Brazil supported by our foundational credit models, along with typical fourth quarter seasonality.

Now unsecured lending balance surpassed $8 billion with record high originations of $4 billion in the fourth quarter. Secured lending grew 3.8% quarter-over-quarter. Recent changes to FGTS regulations have reduced Nu originations by more than half. Though the impact on outstanding portfolio remains limited given the longer duration nature of the secure launch. We remain very comfortable with the portfolio's growth trajectory and risk profile underpinned by very disciplined credit underwriting and the evolving nature of our credit models.

I will now turn to deposits where we continue to build a scalable and resilient funding base. We ended the quarter with total deposits of $41.9 billion, up 29% year-over-year, with growth across all three countries. In Brazil, growth reflected typical fourth quarter seasonality, including the 13th salary. In Mexico, following pricing and product adjustments in the third quarter, deposits resumed growth in the fourth quarter. On funding costs, we saw improvements across all geos.

The cost of deposits declined to 87% of the interbank rate on a consolidated basis by the end of the fourth quarter reflecting mixed dynamics, disciplined pricing and seasonality. Now deposits remain a very strategic lever for us. Strengthening balance sheet resilience, supporting earnings and reinforcing customer engagement while we continue to manage pricing with discipline to preserve attractive economics.

Turning to NII, CLA, and risk-adjusted margins. Net interest income increased 13% quarter-over-quarter, driven by portfolio growth and improved funding costs, especially in Mexico. Credit loss allowance increased primarily as a function of growth as we expanded credit card limits and balances, provisions rose mechanically due to front-loaded origination accounting while underlying credit quality remains stable. We also recorded a one-off item related to Mexico.

As background, Prosofipo is a sector-wide deposit insurance fund to which also peoples are required to contribute to. As the largest SOFIPO in the country, new was required to make an extraordinary contribution of approximately $25 million, which is reflected in interest expenses this quarter. This is a onetime nonrecurring regulatory levy not a reflection of the credit quality or the financial health of our operations in Mexico. Risk-adjusted NIM closed at 10.5%, and would have been broadly stable quarter-over-quarter, excluding the Prosofipo contribution.

Moving to asset quality. As our portfolio has diversified across products, segments and geos, we are now presenting consolidated NPL metrics. We believe this provides a more holistic view of credit quality across the Nubank platform. Now given Brazil's relative size, trends remained largely driven by the Brazilian portfolio, where credit performance continues to track our expectations, supported by disciplined underwriting.

As you see in the slide, early-stage delinquencies measured by 15 to 90 NPLs improved for the fourth consecutive quarter, declining 20 basis points to 4.1%, partially reflecting the seasonality of the quarter in Brazil. As a result of prior improvements in early delinquencies, 90-plus NPLs declined 10 basis points, pointing to 6.6% in the quarter. Coverage ratios remained strong, both on total balances basis and on 90-plus NPLs, providing continued comfort across loss absorption.

We typically see a seasonal uptick in the 15- to 90-day NPLs in the first quarter of the year. This pattern is expected for this coming quarter, aligned with historical trends. Overall, we see no signs of deteriorations and remain comfortable with our credit quality indicators.

Turning to gross profit. Gross profit reached a nearly $2 billion in the quarter, up 38% year-over-year. In terms of composition, float contribution increased reflecting strong deposit inflows in Brazil and improved funding economics in Mexico following the pricing adjustments implemented in the prior quarters. Fees also performed well. Driven by very strong purchase volumes supporting the largest quarterly increase in our credit card market shares in Brazil in over 10 quarters. The credit component reflected higher front-loaded credit loss allowances consistent with the strong portfolio growth in the quarter.

Now looking ahead, we will remain credit first. Credit represents the largest profit pool in financial services and is a key driver of engagement and relationship that across our platform. At the same time, fees and float provide diversification and support a more resilient gross profit profile as we continue to scale across products, segments and geos.

Going to the efficiency ratio now. As part of our disclosure evolution, we updated the methodology for calculating this metric to better align with industry practice and enhance comparability. Details of this new methodology are included in the appendix to this presentation and we are also presenting the ratio under the prior methodology for reference. Under the new methodology, the efficiency ratio declined to 19.9% following below 20% for the first time in our history. This reflects operating leverage with net revenues growing faster than operating expenses, even after typical fourth quarter seasonality in marketing and transactional costs.

In the fourth quarter, we also recognized approximately $22 million of transition expenses provisions related to our return to office decision, which becomes effective only in mid-2026. These cost provisions are temporary and not indicative of the ongoing run rate.

Now looking ahead, as David outlined before, 2026 is in fact, an investment year. We are laying the operational foundations for global expansion and accelerating the adoption of AI and other new technologies across the platform. These are deliberate investments in long-term capacity building Nubank and they will likely put upward pressure on the efficiency ratio in the near term.

We are comfortable with this trade-off. The structural drivers of operating leverage, revenue growth, scale and disciplined cost management remain unchanged, and we expect efficiency to continue improving over the medium term as these investments that we are making today begin to generate returns.

To close the P&L review, net income. In the fourth quarter, net income increased 50% year-over-year to $895 million, delivering a record high ROE of 33%, while we continue investing in growth and maintaining quite robust capital buffers. This includes certain nonrecurring items in the quarter, a positive impact of approximately $58 million of net income related to the remeasurement of deferred tax assets following the CSLL rate increase in Brazil and a negative impact of approximately $29 million related to return to office provisions and the Prosofipo levy in Mexico. Now together, these results demonstrates the scalability of our operating model. Growing earnings while sustaining high returns.

Now turning to capital and liquidity. At the holdings level, total capital stands at $8.9 billion. Of that, $3.6 billion covers regulatory requirements across our three geographies. $2.2 billion represents excess capital in our operating entities. And $3 billion see it at the new holdings level as unrestricted cash and equivalents available to fund both continued growth in our core markets, as well as our global ambitions.

Now on the liquidity side, available funding of $38.8 billion represents approximately twice our net credit portfolio of $19 billion, which represents our gross credit portfolio net of credit card accounts payable. Which provides very significant headroom to continue scaling credit responsibly while also seizing the opportunities coming from further balance sheet optimization. Our capital liquidity positions reinforce our ability to invest in growth from a position of strength, and that is exactly what we intend to do.

Taken together, our capital and liquidity positions are not simply a reflection of our past performance. They are, in fact, the foundation of what comes next, and we enter 2026 with the financial strength and to win our core markets, the firepower to accelerate globally and the discipline to do both things responsibly.

Now I'd like to thank you, and we are very happy to take your questions.

[Operator Instructions] I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer.

Thank you, operator. Could you please open the line for Mr. Eduardo Rosman from BTG Pactual.

I have a question for David Velez regarding AI. David, do you see a risk that Nu could be disrupted by AI? Or do you see Nu as a potential winner in this transformation? It would be great if you could elaborate a little bit since I think the stock and then the sector in the U.S. has been suffering lately because of that.

Sure. And the answer is both. It is both a challenge and has potential for disruption as well as significant opportunity. Net-net, we think it's more opportunity than challenge for us. But we have to take it pretty seriously, and we are taking it very seriously.

A couple of ways to think about it. I think there is one specific trend or one common denominator across every technology transformation. And this goes all the way to even the internet era, which is any business model that relies on simply moving bits from point A to point B, where you're effectively a broker tends to be heard the quickest because one of the things that technology does is remove a lot of that friction in those processes.

So I think to -- some of the commentary that has been around in the market about financial services is, I think businesses in financial services that are simply moving money from one point to another point, will have the higher risk of potential disruption. You need to be able to add more value than that. And I think from that angle, we think -- we have always believed that credit, specifically, credit revenue is actually the most sustainable type of revenue in financial services because of the capital intensity, the regulatory nature of it, the balance sheet aspect and the proprietariness of the data where AI plays a role and ultimately allows you to make a better decision on that.

So I think from one angle, there is potential for challenging around the business model, but I think we're very well positioned given the way we are set up in the strength around credit that we have. I think a couple of our opportunities really on the revenue side. And as a reminder, our package $15 a day and our incumbent competitors are something like $40, so we have a significant opportunity to increase ARPAC is around Nu cross-sell and Nu products that we can be delivering to the very significant consumer base that we have.

And I think everything around cross-sell everything about using the data that we already have to offer new products and services, it's a big opportunity and nice [indiscernible] enabler. And here, we've discussed a few times over the past year, the significant lift that we're seeing when we're using our own foundation model on credit, but also cross-sell and a number of other revenue-related opportunities. And then you have the cost side, and I think the cost side is a little bit more clear.

I think every single company really might benefit from that, where every function that you do, especially as a bank from customer service to compliance to regulatory to AML will be significantly enhanced or being significantly enhanced through AI. So net-net, I do think that there are potential resorptive vectors in some of the business models. But I think when you compare -- when you think about the fact that 95% of the world's financial services profits are still concentrated in incumbent banks that still have significantly larger cost structures. Means that we're very well positioned to take advantage of AI as a technology enabler for revenue and cost and ultimately madly be one of the winners in this technology shift.

Operator, could you please open the line for Mr. Jorge Kuri from Morgan Stanley.

I wanted to ask a question about your loan growth for the quarter. And I guess it's a 2-part question. First, can you help us dimension the impact that your clip increases are having on your credit card growth. To what extent -- I know there is evidence seasonality, but if we think of the year-on-year growth at how much do you think came from those clip increases? How much of that acceleration in credit cards, do you think it's still going to roll over into 2026.

And then the second part is on FGTS, is there a way to quantify what was the headwind on your loan book based on FGTS. In other words, excluding FGTS, what would have been the portfolio sequential growth?

Let me try to slice them in those two parts. So your first question was on the clip. Look, this was a year in which we have deployed this new technologies and approach to credit underwriting very successfully so far in allowing our customers to increase kind of their credit limits, especially in Brazil so far.

And the best way for me to kind of illustrate the magnitude of this increase is Jorge, maybe, refer you to explanatory note, #32 of our financial statements in which we are then starting to provide what I call the unused credit limits. And you can see that unused credit limits went from about $18 billion to $29 billion. So an increase of about $11 billion, which accounts for about 60% increase in unused credit limits. It's a big one.

And I think it wouldn't be possible for us to do so if we hadn't be leveraging kind of the entirety of the predictive AI credit underwriting tools that have been kind of developed by us over the past now 18 to 24 months. Have we seen all of those benefits translated into net income? The answer is no, not yet. So usually, I think at least I see kind of credit limits increases playing out in three steps.

First, you have to offer the additional credit limits, then the credit limit translates into purchase volume. And then you have to see whether the purchase volume will then translate into IBB, We are starting to see the first step, Jorge, which is in the fourth quarter of 2025, our market share in purchase volume in Brazil has gone up by about 50 basis points. It was the biggest market share gain that we've seen in Nubank over the past 10 to 11 quarters.

There's two more to come, and then we still have to see kind of all of those purchase volumes reflecting into IBB. Even though 2025 was, I think, a big sign of the magnitude of this ability to increase clip, I don't think it will stop there. You will continue to see this kind of unfolding in new models and new improvements throughout 2027 -- 2026, 2027 and onwards. And I would also say that the advent of the predictive AI technology will not stop at clip Brazil, right? It will be and is being exported to clip Mexico, clip Colombia, and then we're going to go acquisition Brazil, acquisition in Mexico, what you're going to go to fraud. It's going to go to deposits, pricing and designs. So there's a plethora of options that we're going to be leveraging on. So that's my attempt to address your first question, Jorge.

The second question was on FGTS. So the new regulations of FGTS came into effect on November 1, 2025. And we have seen our originations of FGTS loans dropping by about 50% to 60% in the period in which the new regulation has become effective. It was more than offset by the growth in public consignado in public payroll loans, but it has certainly been a headwind to the origination of this very kind of interesting asset class.

And is there a way to quantify that thinking about it on a quarter-to-quarter basis, what would have been the total balance of credit expansion excluding that. So instead of the 11% FX-neutral quarter-on-quarter would have been the number without FGTS?

Yes, it would have been about 13% to 14%.

Operator, could you please open the line for Mr. Pedro Leduc from Itau BBA.

Thank you so much for taking my question. A little more as you look into 2026, and I'm going to use some of the prepared remarks there, especially in terms of efficiency trajectory. You mentioned that there might be some pressures. I'll see if you can maybe go into detail about it.

And of course, it's a ratio also as I'm trying to think about revenues, of course, you're ending at a very high pace of loan book. NII. But as I look forward, can you help us understand a bit on the drivers when we see funding costs go up, I'm sorry if we can see that continuing a little bit on the portfolio. Just help us think a bit about these drivers now that you are already 35% ROE.

Leduc, thanks for the question. Look, I will refer to Slide 16 of our earnings deck, which is -- brings the efficiency ratio evolution. And we have seen kind of over the past quarters and years, the continuation of the operating leverage potential of the organization. We wanted to highlight very clearly that we may see kind of upward pressure on efficiency ratio in the coming quarters, i.e., in the short term, like the next 4 to 6 quarters.

As a result of very deliberate investments, I would bucket them in three categories. Number one is we have recently announced a return to office policy, right, in which starting on July 1, 2026, employees will start going back up to the office 2x per week. That means that we're going to have to kind of prepare the offices, increase the leased area to welcome our employees as they prepare to come back to the office.

We believe that this will bring enormous benefits to the company, including about kind of ingenuity, kind of a innovation, coordination, but it does come with an increase in OpEx in the short term, and we wanted to clarify this. I would say that the return to the office will likely bring kind of our efficiency ratio, all else constant, up by about 80 to 100 basis points.

The second bucket, I would say, Leduc, is the all of the investments that we are making in AI and new technologies. So that brings new talent that we have to hire, eventually new investments in R&D and research in GPUs that will have kind of a short-term cost, which we believe will be way, way, way more offset by the medium-term gains that we're going to have, but we will not shy away to make investments in talent, R&D and GPU to maximize the impact of our efforts in AI.

And I would say that kind of -- we have returned to the office, you have AI. And the third one is the globalization. So there is a lot of investments that we are making in laying down the foundation for us to go beyond Brazil, Mexico and Colombia. And a substantial amount of those expenses are not capitalized and are incurred in 2026 first to collect revenues and margins in the following years. So that's the direction.

I wouldn't be able to provide you Leduc at this point in time, more kind of a precision on the effect of all of the three, but we think that they would put some kind of upward pressure in the coming quarters.

Operator, could you please open the line for Mr. Yuri Fernandes from JPMorgan.

Most metrics, they look very good. But there is one line here that I think investors are a little bit more puzzle this quarter, that is the tax rate, right? And I know there is a managerial adjustments, and we see some incumbents in Brazil also have similar adjustments. So I think it's -- it's easy to understand and explain. But regarding this quarter, and maybe Lago can help me here. I would like to understand what drove the lower accounting tax, if this was the DTA? And you have lower DTAs, but just checking if this was DTA, some kind of tax-exempt bond, IOC. And maybe some kind of color going ahead, what should we expect for the tax rate for Nubank.

Sure. So Yuri, look, I think the lower effective tax rate in the fourth quarter can be explained by, I would say, largely two things. One, completely nonrecurring and on recurring. What's the nonrecurring one. So about beginning of December 2025, the federal government approved an increase in the corporate income tax applicable to fintechs, including those like Nubank that essentially kind of increased progressively the corporate income tax from about 40% to 45% starting in 2026 and then going all the way in the next 2 years.

Even though that in the medium term is a headwind for our effective tax rate in the quarter in which this kind of legislation is passed. We have to remeasure our deferred tax assets. So our DTAs remeasure up. and that increase in the DTA, which was about $58 million unit is recognized in the fourth quarter of 2025, decreasing the effective tax rate in the quarter. So that's the portion that I attribute as a nonrecurring one-off event. The recurring ones is that kind of as we increase the amount of investments that we have been making in technology across the firm in Brazil, but also in the other countries.

We end up also benefiting from kind of a technology investment tax breaks that some of the governments provide. And that may increase a little bit the OpEx, but they are more than offset by lower effective tax rate. Those are the two aspects that have kind of impacted ETR this quarter.

So very clear, Lago. And you also had the nonrecurring on the Prosofipo like the deposit as you mentioned. So not the same magnitude, but also negative versus this tailwind you had in the quarter.

No, you would think that's precisely clear. I think we have basically three one-offs in the quarter, right? What I would say. One is the $58 million DTA reassessment that we just discussed. The other one was the about $25 million one-off expense of the Prosofipo. And the third one was the $22 million provision expense for the return to office program, right? So those are the three moving parts that we have. DTA positive return to the office negative and Prosofipo negative.

Operator, could you please open the line for Mr. Mario Pierry from Bank of America.

Guys. I wanted to focus a little bit more on the provision expenses, right? Because we did see your cost of risk go up this quarter. And last quarter, if I recall, you were talking about your ability to extend credit to existing clients because you're employing AI and then you're seeing a lower cost of risk in this reverse this quarter. So I wanted to understand a little bit better what happened with provisions in the quarter.

Also, if you can talk a little bit -- you showed your NPL relatively stable. But this is a consolidated NPL, correct? And before you were showing us Brazil NPL only like your NPL on a consolidated basis is lower than the previous number. Just trying to understand why the NPLs as you're expanding into Mexico, especially. Are you seeing lower NPLs in Mexico than you had in Brazil?

Let me try to address each of them in order. So the first one is we did have an increase in CLA item this quarter. And I would be very clear, this was entirely attributed to growth not to any type of asset quality deterioration experienced in the quarter. So we didn't see -- we saw asset quality performing very much in line with our expectations, including the seasonality trends. And now we are on like February 25, and we continue to see kind of our asset quality metrics, no trailing our expectations very well in all asset classes in Brazil, in Mexico and in Colombia.

So we watch this kind of quite closely. But as of now, we have not seen any signs of degradation in our asset quality. What we have seen to justify the increase in CLA is not only the increase in the credit book in itself, which you can see kind of in Slide 11 that grew by about 11% quarter-over-quarter.

But also, Mario, in the increase in credit limits unused credit limits which do not show up as credit per folio per se, but our exposures for which we do need to build CLA. So again, CLA growth entirely driven by growth in exposure, not the gradation of assets. The one thing that I would highlight, at least, Mario, that I like to see going on a recurring basis when I look at those numbers is like NPL formation was fairly stable, 3.6 to 3.5x. Stage 3 formation, fairly stable. And one metric that I personally look as a ballpark, Mario, is the CLA divided by average credit portfolio.

So it used to be like 3.9% fourth quarter '24, then 4.3%, then 3.9%. Then in the third quarter of 2025, we went down a little bit from 3.9% to 3.3%, and now it's back to 3.9%. So I think the third quarter, as we updated them all those with higher recovered ratios, it may have come kind of slightly below. Now it's going back to 3.9. I'm sure you're going to ask the questions what's next?

I think what next is something around or below the average between 3.3% and 3.9% on the coming quarters, of course, something that we don't control, but that would be more or less our expectations with the mix that we have today.

So that's your first question. I think your second question was on the NPLs. Would you provide kind of now consolidated NPL trends simply because as we grow the book internationally with Mexico, Colombia and hopefully other countries in the next years, we start to see those metrics kind of better representing the economic reality of the company rather than looking at Brazil only.

However, if we were to post the Brazil only NPL charts, they would equally show kind of a fairly benign trend of asset qualities, moving very much in the direction of seasonality that we expect to see in the fourth quarter.

And then your question about, look, how can you actually aggregate Mexico and Colombia and get to lower NPLs it is justified mostly by the write-off policies that we have in those countries than on the risk of those countries. So for example, in Mexico and Colombia, we can have shorter write-off policies than we have in Brazil, and that kind of affects the overall NPL calculations.

But in general, Mario, no concerns at this point in time with asset quality. It is super point -- super important to highlight, and I know that you've been following this for many years, so I speak more for to the other participants of the call. Fourth quarter of every year, we usually observe a benign movement in NPLs because of seasonality, but equally, we do expect to see kind of an uptick in NPLs in the first quarter of 2026, also following natural seasonality, right?

Operator, could you please open the line for Mr. Gustavo Schroden from Citi.

Hello. Good evening, everybody. My question is regarding credit products and also client mix. We could see a relevant increase in loan book for credit cards and personal loans. But I'd like to explore more of the secured loans. Lago explained about, Lago, you explained about the FGTS change recently, indeed, has impacted the evolution of this portfolio.

But I'd like to understand the appetite for payroll loans, I mean public -- public and private pay loans, how the bank sees these products, we should expect some, let's say, replacement of FGTS by this private pay loans mainly. So any view on that would be great. And also about the client mix, should we -- could you explain us how the bank is evolving in this, let's say, exploring the affluent market, I mean, mid- to high-income customers, especially after this increase in credit limits, that would be great.

Thanks for the question. Let me try to address the first one on the breakdown of originations of our secured loans and then David may address the second one on our performance in both the what we call super core and high income segments.

So I would basically divide our, what we call secure loan portfolio in three, right? So we will have the FGTS. We have the public payroll loans, and we have the private payroll loans. So FGTS is the one that has recently received kind of a negative impact of the new regulations starting on November 1, 2025. It has dropped kind of our originations by about 50%. And we continue to have a very good dialogue with the government to try to influence the agenda for 2026 and 2027.

And we have become market leaders in FGTS. It was a very -- it is and it used to be a very good product, and we believe it will continue to play an important role in the formation of our secured lending book. Even though if regulations don't change, we'll probably play a smaller role than it could have played before. But that's bucket number one.

Bucket number two, public consignado or public payroll, which I put here, including both [indiscernible] and [indiscernible] We are very bullish on this. We think it is still a market that has kind of a lot of opportunity to increase efficiency in the intermediation and in the distributions. We can offer no products at materially lower cost than most of the other market participants and it's now finally entering into time in which we will see interest rates drop in Brazil.

And with that, we hope that kind of portability will pick up. And we like to believe that we're going to be one of the biggest beneficiaries of that of the trend. So I think it is one that we think regulation is there, portability is there, interest rate cycle is there. So we are bullish that this will kind of have an even faster growth in 2026.

The third bucket is private consignado. So this is a product with which we are very, very optimistic and bullish on a structural form by which I mean it is a way for fintech such as Nubank to have access to information and to customers who used to be primarily served by incumbent banks which own the payroll service of large corporates in Brazil. So it's a massive opportunity for us. And it's one that we will lean in as soon as we see the mature improvements in credit risk that this product offers.

We are still not seeing that. I think part of that is kind of a counterparty risk of the corporates. Part of that is the collateral is not yet operating at its full potential. We, however, think it it's a matter of when, not a matter of if. You've also been following this quite closely for some time. You may recall that when public consignado was introduced a few years ago, it took kind of a year, 1.5 years for everything to all of the collaterals to be working well. and we are just waiting for this to happen for us to lean in more heavily.

Now let me pause here, see if you have any follow-ups and then pass the floor to David for him to comment on the affluent part of your question.

All clear, Lago.

Perfect.

I think I'll say on the secured lending side is it is -- continues to be a very significant opportunity for us. I think growing within that existing profit pool has been probably more complicated than we expected given the significant operational complexities that the product has.

There is a fair amount of features that need to be built into the product, specifically around portability. Most of the growth of those products are portability and when customers are doing that portability. You need a lot of different integrations. There's also a fair amount of fees. All of that friction is going away. I think the tailwind, if there's one consistent tailwind in Brazilian financial services is that all those -- all that friction and cost that historically have improved had made it harder to move towards the best product, it's going away.

So we're seeing accelerating market share gain, and we are ready to -- we're building a lot of those features, and we're getting significant share on the secured line. So while I wish the traction to date had been significantly higher, I think every single month, we're seeing an acceleration of market share and the tailwinds are helping.

On the high income side, we continue to see a very good growth. Again, this is a competitive environment. It's a competitive segment, a lot of -- a lot of banks in common banks and others are going upmarket. We define a market for us as customers are making above BRL 12,000 per month. So this is not 1% of Brazilian. This is probably closer to 10% of Brazilians. And within this consumer base, we already have two out of -- two out of five, about 40% of those Brazilians in that bracket are customers of Nubank today.

They're just not really using us as their primary card. We are the third car. We have small share of wallet. A lot of the times was because we gave him a low credit limit initially. And if we had opportunities to improve credit limits on mass market, and we're seeing that with AI models, we have even more opportunities to improve credit limits on high income because a customer type that we didn't really understand.

So we have to fix credit limits, which we're doing. We have to improve the value proposition of the product, specifically on credit card, which we are. Over the past couple of quarters, we launched new improvements, different cash back rates. We announced a lot of integration with our Nu Travel platform. So it's a really good product where we guaranteed the price of any ticket or hotel that you book in our app. We're seeing customers getting significant value out of that. So it's very well integrated with the travel value proposition.

We announced our frequent flyer lounge in Guarulhos in Sao Paulo that is getting a lot of acceptance. So there's a long path of opportunities that we have to improve the product on the credit card side. And we see that translating into increasing market share. This segment for us grew something like 40% year-over-year and is gaining share across our portfolio. So we're seeing good traction. A lot of these investments are paid off.

The second part of the value proposition is investments which you might know that obviously, we've discussed it a few times. It's taken a while for us to build a very, very compelling investment value proposition in our app. We're getting very close. We are close to really product parity. We have now all the products that this segment needs in our app. We have fixed income products, equity products, crypto products.

We have all the type of visualizations that this customer is asking. So we're getting very close to have a very good investment platform that it's critical to win this high income segment. So overall, these are -- these two specific opportunities that you mentioned, there are not one, two quarter opportunities where you're significantly gone. These are long journeys of a lot of product improvements, but we feel very good about the progress we've made and the opportunity we have ahead.

And Gustavo, just one additional point. We mentioned about the mass market, which in our definition, our customers will earn up to BRL 5,000 per month. And then you asked about what is called high income, which our customers who earn more than BRL 12,000 per month, which was the answer that the David had provided.

But in the middle, which is what we call super core, i.e., customers who earn from BRL 5,000 to BRL 12,000 per month, it is the segment in which we are growing the fastest, right? So if David mentioned that in the high income, we've been growing at about 40% per year. in what we call super core. We are growing at about 100% in 2025. So I would kind of invite you and others to kind of segment this at least in 3 parts. And I think there's a massive opportunity for us to go into the super core there as well.

Could please open the line for Neha Agarwala for HSBC.

Just wanted to follow up on the private payroll segment. We do understand your concerns regarding operational complexities at this point. But we do see a lot of other lenders being more aggressive in this market. And the market has doubled during 2025.

Do you see the risk of some of your customers who might have personal with you going or have a credit card with you going to other banks and taking private payroll loans. And ultimately, their leverage increases and that could impact the asset quality for those customers for you on the unsecured side?

Very good question. And yes, we are very mindful of those two risks, which I call kind of the cannibalization, i.e., customers borrowing from another bank and kind of us losing the primary banking relationship. That's one. The second one is structural subordination, right? So customers boring and providing the collateral and ourselves becoming stretchy subordinated to someone else. The same can be made when we lean in into this product.

Even though we have been very mindful of this, we have not yet seen any evidence that any of those two risks that you've laid out are materializing within our customer base. In fact, most of the customers who have been applying for private payroll loans have been customers with higher credit risk, at least that has been our experience and most likely customers who would not be entitled to have access to an unsecured personal loans or even sometimes to unsecured credit cards but it -- but we are tracking this very, very closely.

In terms of the growth of the market that you've also pointed out now, I would highlight that there are a few things to adjust in this growth. One is there's just a natural shift from asset classes that were considered private consignado without the collaterals that were instituted by the government and are just now migrating to the new private consignado.

Those are usually loans that have been carried by kind of the more traditional incumbent banks, and they account for a fairly substantial portion of what is seen as the growth of this new asset class, i.e., is just migration from the old to the new.

The second one, we now see kind of players playing in this space with very two kind of different approaches. The incumbent banks who have relationships with the corporates when it comes to payroll loans. They are more focused on the lower risk customers and the digital players are more focused on the higher-risk customers. But when we step back, we are seeing kind of this market operating with first losses of no low double digits, which is not yet conducive to the quality of the collateral that this product can have.

Once we see kind of a credit improving as the product will deliver, we will not shy away to leaning very heavily and the term cannibalization is just not a term that we use. We will be there offering the best product for our customers, irrespective if they will actually use the proceeds to prepay or repay higher yield assets. We are not moving ahead with this as strongly as others, not because of the risk of cannibalization, but more because of conservatism with credit risk.

Understood, Lago. And in terms of cannibalization, yes, NIMs might go down, but risk-adjusted NIMs might not be impacted as much, even if you replace the credit from unsecured to secured with some of your customers, right?

That's correct. The other component of that, Neha, is that you may see at some point in time the amount of capital that you have to allocate to private consignado, possibly being lower than the ones for unsecured. So not only risk-adjusted NIMs may be preserved or even increase in an absolute amount, but the return on equity may be as appealing, if not more appealing because you have to post lower capital to that, yet to be defined.

I just wanted to understand why not offer the private payroll, and I understand that there are complexities and you can price for those complexities and collateral not working smoothly. Why not offer it to some of the customers whom you deem to be riskier and don't want to give them an unsecured loan at this point. Why not start off with the secured private payroll loan with them and price it accordingly.

You I would certainly could. I think what we are saying is that the benefits of the collateral for the higher-risk customers, have not proven to be material enough to justify a substantially different credit underwriting or pricing policy to date. But again, just to be super clear, I think it is a matter of when, not a matter of if this is a good product, it is a good structure. This will benefit kind of consumers, by and large. We just don't think that is yet ready to be kind of the product in which we will lean in that heavily at this point.

Operator, could you open the line for Mr. Tito Labarta from Goldman Sachs, please.

I guess my question is following up a bit more on expenses. First, and you talked about 2026 being an investment year and thinking more about the global expansion. Just help us think a little bit about what investments are needed there? Because I mean you got the initial license pre-approval, I guess, in the U.S. But is there more investments that you need to make in the U.S. already in 2026. Just help us think about what are these investments that you need to lay this global foundation.

And then also just specifically in the quarter, because if I look at the accounting P&L, which I guess is more comparable to the estimates that are out there, there was a big jump in expenses, and I know there was the one-off from the return to office but marketing expenses jumped quite a bit. G&A expenses jumped a bit. If you can just give some more color, what specifically drove those increases in operating expenses in the quarter would also be helpful.

Thanks Tito. Quickly on U.S. We will continue to invest. I mean, kind of we are investing more, mostly on team building and product. It's de minimis. It's not a significant source of investing for launch in the U.S. We did announce a number of bigger marketing partnerships over the past couple of months. And those really are related to both our core markets, as well as U.S. and potentially future markets around the world.

So there is an increased a bit in marketing. Their team increases that we're having for the U.S. launch. But I wouldn't say they're going to -- they expect to be significant in 2026.

And then, Tito, on your questions about the breakdown of our OpEx in the fourth quarter of 2025. I think the marketing one is a traditional seasonal one. It usually spikes a little bit in the fourth quarter of the year. The other one was incorporated in the tax breaks related to technology investments. So many of the increases in [indiscernible] that are recognized as OpEx, but they actually drive quite a bunch of off-tax efficiency. But nothing extraordinary or nonrecurring other than those three moving parts that we mentioned.

Okay. No, super helpful. Thanks, Lago. And maybe just one quick follow-up for David. Any just initial thoughts on what the expansion plan in the U.S. will be like just a high-level footprint on what you're targeting segments go-to-market there? Any color or thoughts that you can provide would be super helpful.

Sure. On a very high level, and we're not really ready yet to disclose specifically what the strategy there is going to be. But at a very, very high level, this is the largest market in the world. And while at a very high level, it seems like a very saturated or competitive market in certain segments. When you dig in into subsegments in certain niches that, by the way, happen to be the size of Brazil.

We actually find opportunity to solve a number of consumer problems that are similar to what we've done in the past. So we're going to have a very targeted strategy. We're going to be very disciplined on investing. There are a lot of focuses on certain potential geographies or subsegments that we are interested about. You're not going to see us kind of shooting in all directions here because it's a bit of a long journey, and we fully acknowledge that this is a very competitive and sophisticated market in certain areas. But we do think that it's -- there are opportunities for us to create a meaningful business in certain sub areas of the United States.

So thank you, everyone. We now have approached 60 minutes of the call. So we are now concluding today's call. On behalf of new holdings, our Investor Relations team, I want to thank you very much for your time and participation on new earnings call today. Over the coming days, we will be following up with questions received tonight, but we are not able to answer. And please do not hesitate to reach out to our team if you have any further questions. Thank you, and have a good night.

The Nu Holdings conference call has now concluded. Thank you for attending today's presentation. You may now disconnect.

Good evening, ladies and gentlemen. Welcome to Nu Holdings conference call to discuss the results for the third quarter of 2025.

A slide presentation is accompanying today's webcast, which is available in Nu's Investor Relations website, www.investor.nu in English and www.investidores.nu in Portuguese. This conference is being recorded, and the replay can also be accessed on the company's IR website. [Operator Instructions] [Foreign Language] [Operator Instructions]

I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer at New Holdings. Mr. Souto, you may proceed.

Thank you, operator, and thank you, everyone, for joining the earnings call today. If you have not seen the earnings release already, a copy is posted in the Investor Relations website.

With me on today's call are David Velez, our Founder, Chief Executive Officer and Chairman; and Guilherme Lago, our Chief Financial Officer.

Throughout this conference call, we'll be presenting non-IFRS financial information, including adjusted net income. These are important financial measures for new holdings, but are not financial measures as defined by IFRS and may not be comparable to similar measures from other companies. Reconciliations of the non-IFRS to the IFRS financial information are available in the earnings press release. Unless noted otherwise, all growth rates are on a year-over-year FX neutral basis.

I would also like to remind everyone that today's discussion might include forward-looking statements which are not guarantees of future performance, and therefore, you should not put undue reliance on them. These statements are subject to numerous risks and uncertainties and could cause actual results to differ materially from our expectations. Please refer to the forward-looking statements disclosure in the earnings release.

I will now turn the call over to David. Please go ahead, David.

Hello, everyone, and thank you for joining us today. In Q3 2025, effectively every single one of our metrics continue to grow, reinforcing our position as the leading digital bank in Latin America and one of the leading fintech platforms globally. Our customer base grew to 127 million customers with more than 4 million net additions in the quarter while maintaining an activity rate above [ 82% ], a clear reflection of the depth of engagement we continue to build with our users. In Mexico, we surpassed 13 million customers now reaching around 14% of the adult populaton. And in Colombia, we're approaching 4 million customers. Both markets continue to demonstrate strong traction, highlighting the scalability of our model.

The solid growth, combined with continued ARPU expansion, which surpassed $13 this quarter has led to record revenues of over $4 billion. These results highlight the compounding effect of our customer expansion, deeper engagement and disciplined monetization.

Our gross profit continues to rise sharply, reflecting strong unit economics and operating leverage. And with a cost-to-income ratio of 28%, we continue to progress on our trajectory of improving efficiency. And finally, we delivered net income of $783 million, another quarter of solid profitability even as we keep investing in growth and innovation across all markets.

This consistent performance is a direct result of our business model, one that attracts millions of new customers every quarter, fosters deeper engagement that expands monetization all while operating on a low-cost and highly efficient platform. This formula continues to drive our earnings growth across markets, but with each component playing a distinct role in every geography.

In Brazil, we now serve over 60% of the ad population and estimate that we're already the largest player in the SME segment by a number of accounts. Having reached scale, revenue per customer has become the main growth driver. Our focus going forward is broadening our product portfolio, deepening engagement across all segments and continuing to execute our credit strategy, increasing exposure among customers with the strongest risk-adjusted returns.

In Mexico, our main focus remains on expanding our customer base, deepening product adoption and advancing financial inclusion, all while laying the groundwork for sustainable long-term monetization. Given the scale-up phase, ARPAC levels are already nearing dosing in Brazil, reflecting the strong unit economics of the credit card business in that market, driven by a higher share of interest-bearing balances and a steadily declining cost to serve supported by our ongoing platformization efforts.

Both markets demonstrate the strength and the stability of our model, which is capable of driving rapid growth and scale in earlier stages. -- while expanding profitability as market matures. Diving deeper into Mexico or second score, we see a market now beginning to scale and one that we expect will contribute meaningfully to our results in the years ahead. We're building strong foundations, having reached market leadership position in the Mexican digital banking space, already reaching 13 million customers or around 14% of the adult population compared with about 10% when Brazil entered its inflection point back in 2019.

Even with the product portfolio still largely center on the credit card, ARPAC has already reached $12.5 reflecting strong customer engagement and the favorable unit economics of this product in Mexico. On the cost side, cost of service was already below $1 and recent adjustments to deposit yields are beginning to flow through our cost of funding. Looking ahead, we'll continue stacking U.S. curves with focus and discipline, while Brazil and Mexico remain our core priorities where most of our resources and execution efforts are directed. We also see transformational optionality in the U.S. following our filing for a national bank charter, a step that could unlock new opportunities over time as we remain fully focused on our core markets.

As we continue scaling across markets, we're also building the next generation of our platform, refining how we operate and how customers experience banking. We have heard several investors asking us about our AI strategy, and so we wanted to spend a few minutes on it. Our vision is to become AI first, which means integrating foundation models deeply into our operations to drive an AI-native interface to banking, while creating meaningful benefits for both our customers and our business.

For our customers, AI is enhancing our understanding of each individual and their financial needs, allowing us to deliver personalized recommendations, contextual offers and products and proactive insights at the right amount. It will also transform the way people interact with Nubank, be it through a simpler and seamless app or to a number of additional channels, embedding conversational user interfaces. We think there is a significant opportunity to include Agentic workflows across most products and services, improving customer experiences across the board.

For our business, AI is strengthening how we manage risk and scale efficient. It is helping us to design safer and more precise financial solutions, reducing credit and fraud losses and enabling tailored collection strategies that drive better recoveries. At the same time, it is enhancing productivity across the company from leaner operations to faster development cycles and higher engineering throughput.

When we bring all of this together, becoming AI first means accelerating our flywheel by scaling to offer high-quality products at lower costs. unlocking the full value of open finance, deepening cross-sell and product penetration and opening new revenue streams, all while optimizing pricing and delivering superior value for both customers and shareholders.

But AI is not a buzzword for us. We believe Nubank is uniquely positioned to become AI first and a leader in the use of AI and financial services globally, and we're already starting to see the first breakthroughs. Since our early days, we've known that technology and data will be our strongest competitive advantage, being cloud native and built entirely on modern architecture enables us to simulate, experiment, train and deploy foundation models at scale. Coupled with our proven ability to attract world-class talent, this puts us ahead of incumbent banks and regional finding competitors and places us in a unique position globally.

Over the past 12 to 15 months, we developed new former or proprietary approach for building large generalizable models based on advanced transform architectures and self supervised learning principles similar to those powering world-class and kens. These models provide a deeper understanding of customer behaviors and can be deployed across our critical risk and personalization engines. To reach this level of performance, the first generation of our new former model was built with 330 million parameters and trade on approximately 600 billion tokens, an unprecedented scale of data by financial industry standards. That data represents only a fraction of our full data set, which spends trillions of tokens and reflects the vast scale and diversity of Nubank's platform. Our business model with principality at its core generates a deep repository of high-quality transactional and behavioral data, giving us a distinctive edge by enabling new form to learn from richer context and continuously strengthening its predictive power.

Historically, gains in credit performance have come from our main fronts, incorporating more and better data sources into models, expanding training samples or reducing bias within them, optimizing positive frameworks, including the use of complementary models that evaluate different dimensions of credit risk; and finally, refining modeling techniques from definition of targets to model architecture and feature engineering.

The adoption of Foundation models represents a radical expansion of this last frontier. It brings a research-driven approach that moves the needle through advances in model architecture and training processes, enabling rapid and continuous improvement as AI researchers push the boundaries of what's possible.

When we applied this approach, the models were built to deliver an average improvement about 3x higher than what's typically observed in successful machine learning model upgrades. Translating this into business outcomes, our initial models enable a major upgrade to credit the card limit policies in Brazil, allowing us to meaningfully increase limits for eligible customers while maintaining the same overall risk appetite.

This successful breakthrough within an already robust underwriting model, like credit card Brazil underscores the significant potential of these advanced approaches. We're now focused on scaling this innovation beyond Brazil, already in motion in Mexico and extending them across every part of Nubank from personalization and cross-sell to fraud and collections, further reinforcing both the strength of our model and our ability to execute at scale. That said, we're still just scratching the surface. As always, at Nubank, it's still day 1, but we believe that embedding AI into our business represents a once-in-a-lifetime opportunity to further differentiate Nubank from traditional banks. We're building on years of experience in model governance, privacy and large-scale model deployment to ensure we continue evolving responsibly. This means having robust processes to make sure our tools true from what our customers financial are being with the right guardrails in place to bring these advanced models safely into production within a highly regulated environment. We'll continue to share our progress as this journey evolves.

And with that, I'll hand it over to Lago to walk you through the financial highlights of the quarter. Thanks a lot.

Thank you, David, and good evening, everyone. To begin, I'd like to start with our credit portfolio. Total balances reached $3.4 billion in the third quarter, up 42% year-over-year on an FX-neutral basis, with very solid growth across all products. Credit cards accelerated during the quarter, supported by our ability to continuously enhance the precision of our credit models and increased limits for our customers. all while maintaining very healthy risk metrics as we will see in the following slides.

At the same time, secure lending grew 133% in unsecured loans 63% year-over-year, reflecting the ongoing diversification and maturation of our portfolio. Together, secured and unsecured loans now account for nearly 35% of total balances, up from 27% a year ago. This reinforces our capacity to broaden the credit spectrum and serve a wider range of customers' needs over time.

Moving to loan originations. We reached a record high of $4.2 billion in the quarter, up 40% year-over-year on an FX-neutral basis with growth coming from both unsecured and secured land. In unsecured lending, performance was supported by the strong momentum in our SME portfolio and buy new credit policies introduced for both business and individual customers. These updates are enabling us to safely expand eligibility and increase average loan sizes while keeping new originations more concentrated in lower risk segments.

In secure lending, results were driven by strong originations in public payroll loans or Consignado, which grew nearly 130% year-over-year along with a gradual normalization of INSS loans.

Now turning to deposits. Our balances reached $38.8 billion, up 34% year-over-year on an FX neutral basis, while the cost of funding actually improved from 91% to 89% of interbank rates. This is a clear demonstration of our ability to grow volumes while enhancing efficiency, continue to build a scalable and sustainable funding franchise across Latin America.

In Colombia, deposits continued to grow steadily, even with funding costs below the interbank rate. In Brazil, we saw strong inflows across all segments, reinforcing the depth and the resilience of our deposit franchise. And in Mexico, we had anticipated some outflows following the recent reduction in deposit yields. This was a deliberate move that reduced our consolidated funding cost and this was fully aligned with both our expectations and our long-term strategy for sustainable growth. Recent trends in Mexico reinforce our confidence in our ability to continue expanding and strengthening our deposit franchise.

Moving to net interest income. We reached $2.3 billion in the quarter, up 32% year-over-year on an FX neutral basis, driven again by the continued expansion of our credit portfolio. Net interest margins contracted by about 40 basis points from the prior quarter. This reflects our disciplined approach to optimizing risk-adjusted returns as we continue to expand originations in lower risk segments, including in credit card interest-earning portfolios, unsecured loans are also to lower risk individuals and higher shares of SME and secured loans within our total interest earning portfolio.

While some of these products carry lower nominal yields they strengthened the portfolio's overall quality and resilience over time, as you can see in the next slide. Our credit portfolio continues to outperform our expectations supported by disciplined underwriting and a healthy mix shift towards customers and products with stronger risk-adjusted returns. Combined with better recoveries, these factors drove a 7% decline in credit loss allowance expenses quarter-over-quarter, also on an FX-neutral basis, mainly reflecting lower provisions in our 2 largest products, namely credit cards and unsecured loans.

As a result of this lower cost of credit, our risk-adjusted net interest margins expanded to 9.9% in the quarter, underscoring the resilience and the quality of our portfolio. Next, looking at delinquency metrics for our consumer credit portfolio in Brazil.

The 15- to 90-day NPL ratio remained well within expectations, ending the quarter at 4.2% and slightly below the historical third quarter seasonality. The 90-plus day NPL ratio increased marginally to 6.8% and also very much in line with the expected seasonality and the underlying portfolio dynamics.

Now finally, our coverage ratios remained solid, even though they declined modestly in line with the recent movements in credit loss allowance. We continue to maintain what we believe to be a quite robust provision buffer both over the total portfolio and specifically over the 90-plus day NPL balances.

Moving to gross profit. We delivered another quarter of solid growth, reaching $1.8 billion, up 32% year-over-year, also on an FX-neutral basis. The expansion in gross profit margin now to 43.5% and reflects the consistent top line growth, combined with the continued improvement in the risk-adjusted performance that we saw in the prior slide.

These trends reinforce the sustainability and the scalability of our business model as we continue to balance growth, profitability and risk discipline across the 3 markets in which we operate. In the third quarter, our efficiency ratio decreased slightly to 27.7%, reflecting continued progress in productivity and operating leverage. Yet, we will continue to invest intentionally and strategically to become the largest and the most loved retail financial institution in Latin America. These investments are fully aligned with our long-term value creation strategy, even if they sometimes create short-term pressures on costs. That all said, the structural trend remains clear as we scale revenue growth and disciplined cost management will continue to drive efficiency gains and margin expansion.

Now to wrap up we delivered a record high net income of $783 million and a record ROE of 31%, up 39% year-over-year, also on an FX-neutral basis. We achieved these results while we continue to deliver strong operational growth, always putting our customer at the very center of everything that we do, offering better products, lower fees and an exceptional experience. These results once again highlight the strength and the scalability of our model as well as our ability to combine growth with profitability.

Now with that, we will open the call for questions. Thank you.

[Operator Instructions]. I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer.

Thank you, operator. Could you please open the line for Mr. Yuri Fernandes from JPMorgan.

I think the debate from investors here, I'd love to hear your thoughts, David, are related to your provisions, Lago. Your cost of risk was lower. I think you are doing a risk migration, right, growing more middle income in Brazil. growing more secured lending. When we check your new Stage 3 formation -- new Stage 3 formation improved. But I guess investors, they will try to understand the lower provisions this quarter that helped on EBIT. So if you can like provide some explanation for investors to understand what drove this lower provision. I think this will help with the understanding for the quarter.

Thank you very much, and congrats again.

No, thanks, Yuri, for the question. Yes. I think asset quality has been positive over the past 2 quarters. I think this quarter, we have also seen kind of asset quality performing as per expectation, even it's likely better than expectations. We have also had some effects of the policies that we have intensified over the past now 3 to 4 quarters of reactivating customers in Brazil who had defaulted with us a few years ago and only now after they have cured their debt, we are also kind of offering them additional credit opportunity that has materially improved the recovery levels.

And then finally, we actually have seen through both machine learning, but also the predict AI technology and modeling that David alluded, the ability to actually have greater precision in some of the credit modeling techniques. So what you have seen is kind of a asset performing in line or even better, but it's still, if I would not draw your attention, you do, let me go here, Slide 17, you see that the coverage ratio that we continue to have are at levels that we believe to be fairly robust in both the total balance as well as NPL 90-plus. Now let's see how it goes, but we are also kind of in the mid of the fourth quarter of 2024 now, it's November 13. And we continue to see asset quality performing relatively okay. So that's kind of the main background for the evolution of our CLA this quarter.

Operator, could you please open the line for Mr. Jorge Kuri from Morgan Stanley.

Congrats on the numbers. Great results. My question is around your net interest margin. I heard what Lago said about the mix of credit being responsible for the decline in NIM. I have -- I'm looking at just the nominal numbers and your interest income was up 14% quarter-on-quarter versus a loan book in total that was up 11%. So it doesn't seem that you're growing your income less on the assets, which would be sort of like a signal of mix deterioration. It's actually the other way around.

But on the flip side, your interest expense was up 24% quarter-on-quarter versus your deposits of 6%. And so you talked about the cost of deposits coming down, but it's just in this -- in dollar numbers, it's kind of like doesn't at all. And so I'm just wondering if you can walk us through the dynamics and exactly what explains that NIM contraction.

Sure, it, look, 2 things on this. So first on the revenue and then on the cost. I think on the revenue side, we have seen the growth being kind of more heavily weighted into less risky assets, not only asset classes per se. For example, you can see that if you go to Slide 12, you can see that, for example, secured lending has outpaced the rest of the portfolio. Secured assets has no everything else constant, lower kind of yield levels.

But even within lending and within credit card, we are seeing kind of a faster growth on a balanced basis. towards less risky customers that would have all else equal kind of lower yields. So that is one of the things that would justify, but you correctly pointed out that we have also seen an increase in interest expenses, and that has come entirely from Brazil. So our average funding cost in Brazil has gone up and the average funding cost in Mexico and Colombia have been coming down.

When we look at the average funding cost that we published on Slide 14, you will see kind of the -- what we call the cost of deposits as a percentage of the interbank rate going from 91% to 89% and then may call the question why, how do I kind of connect the dots, right? If you are lowering the cost of funding as a percent expressed as a percentage of interbank rates, how can your cost of funding expressed in dollars been going up? It's because the piece that is going up is the piece denominated in Brazilian reais, which is subject to the nominally higher interest rates of link -- so the weighted average cost of fund the expressed as a percentage of the interbank deposit rate, which is what you see here on Slide 14 has come down. But the overall interest expenses, dollar-wise, has gone up a little bit.

So the combination of lower asset yield because of the mix with a slightly more expensive funding base in Brazil has compressed net interest margins in the quarter which is what you see on Slide 15 that has gone from 17.7% to 17.3%.

What I would, however, point out is, when you're taking into account the asset quality or the cost of risk, you actually see an expansion in margin. And that is what is shown on the subsequent slide, which is Slide 16. Our risk-adjusted margin has actually gone up from 9.2% to 9.9%. And which goes to show that even though we have kind of increased the growth towards less risky assets that has come at the expense of slightly lower asset yield. This has been more than offset by much lower cost of risk, which has left with the expansion of risk-adjusted NIMs.

All right. That was very clear. And if you remind my follow-up on the previous question, on provisions. You mentioned recoveries stronger than expected. Would you mind quantifying that and what impact it had on the combined provision number?

We don't -- we are not disclosing this one-off impact, Jorge. It's basically the additional of the recoveries, mostly from the customers that we reactivated over the past now 3 quarters. This is a program that we have done by kind of offering a second chance to customers who defaulted was a few years ago have completely kind of paid down their debt and then we have seen that out of those customers. The recovery has been higher than we had booked for but we are not disclosing the breakdown of the additional recovery coming from this pride.

Operator, could you please open the line for Mr. Pedro Leduc from Itau BBA.

Hello, good evening, everybody. Thank you for the call and taking the question. If I may, on credit cards, please. Last quarter, we saw a big increase in newly granted limits this quarter, we may be seeing some of the effects here. There's more cards active, more cards generating revenues, transaction volumes, the cards seem to be going up. Can you talk a little bit more about how you're seeing this rollout perform on the ground? It looks like you did another small increase now in 2Q. If you can talk about that as well. And I think it may tie up also to the -- what we're seeing in the stages and the probabilities. It seems like this growth is coming from slightly better quality mix if you can also include that. I know it's a longer question, but I think you get the spirit.

Sure. Leduc, thanks for the question. So look, we announced a relatively large credit limit increase program in the second quarter of 2025. and the rollout of that credit limit program was spread grossly 1/3 in the second quarter, 1/3 in the third quarter and 1/3 expected to be finalized in the fourth quarter. So we have not yet seen the full effects of that clip program materialized and the financial performance of the company. It's something that we will only see in full most likely towards the mid and end of 2026 because it takes some times for limits to converge into PV and for PVs to converge into IBB. So there's some leeway there as well.

Second point, Leduc, you're right. I think a substantial portion of the credit limit was granted to less risky customers. And so kind of the average unit of risk that we have added has actually lowered over time. However, as we increase the limits to kind of lower-risk customers, we also decreased the flip side, the utilization, right? So I think if you have BRL 1,000 limits, and we increased this by 20% you would experience much higher utilization than if you have 100,000 limits and we increased this by 10%, but both the utilization as well as the credit performance related to this credit limit increase have now both performed largely per our expectations.

So nothing kind of the deviates or forces us to revisit both from the offensive as well as on the defensive side, the pace and tenacity of those movements. Now even though we did disclose in the second quarter that we saw a big clip, a credit limit increase I don't think we should see this as a one-off, right? This is really a continuous enhancement of the programs that will not be kind of a straight line, but we will see kind of a clip programs did introduce from time to time. This is what we've seen over the past years.

And then if you go Leduc what David mentioned at the beginning of his session about the implications of the predictive AI modeling to our credit underwriting. I would say that, first, we have introduced this to credit limit increases that has not yet been introduced to releases by which I mean we have been able to sharpen how we increase credit limits of existing customers. We have not yet applied this to the determination of the new customers to which we granted the initial line, which we call customer acquisition. We have also not introduced this to lending. And we have not introduced this to Mexico and Colombia. So I think there's still quite a lot of runway for us to see further improvements and enhancements in our credit underwriting performance.

Operator, could you please open the line for Mr. Mario Pierry from Bank of America.

Guys. Congrats on the results. Let me double-click a little bit on Mexico and Mexico ARPAC of $12.50 that you're showing -- and which is a quite impressive number, right, especially given that Mexico is fairly new for you and the ARPAC is almost similar to Brazil. Can you give us like a little bit more details on the breakdown of the ARPAC between interest income and fees because -- and I asked this, right, because we saw the regulator in Mexico now proposing our -- card interchange fees. So I was wondering what is your view on that? And how much that could impact your results in Mexico also staying with Mexico, you only give us data, right, the NPL data and coverage data for Brazil operations only. I was wondering if you could share any asset quality metrics from Mexico, that will be helpful.

Let me try to address each of those questions and feel free to follow up if I miss any of them. So I think on Mexico, you mentioned about the evolution of the customer in ARPAC and cost to serve. And I would draw your and the attention of the orders to Slide 7, where you can see the evolution of our customers in Mexico. It's now about 13 million customers accounts for grossly 14%, 15% of the adult population of Mexico, but accounting for now nearly 25% of the bank population in Mexico.

So we can easily say now that about 1 out of every 4 bank Mexicans are customers of new bank, which makes us quite excited. And then as you said, you see kind of the ARPAC evolution in Mexico. Most of them are kind of interest related, both from credit card lending and floating from our deposit base. the fees, the interchange related to both credit cards and debit cards accounts for a smaller portion of the overall ARPAC.

That said, Mario, I think you alluded to the public consultation that the Mexican government has recently issued ended capping the interchanges of both credit cards and debit cards in Mexico. We have, since this kind of came out being in very active dialogue with other industry participants and with the government itself. And even though that accounts for the smaller portion of our revenues, we are concerned with the idea of caps and price control there because they may actually inhibit, the financial inclusion and credit dymphony than we have seen in Brazil and other countries as they make the unit economics of new to credit customers less compelling. So we are kind of in active discussions with all of the industry participants. We are very confident that kind of we will be able to find together as an industry to a good balance that will not put at risk our ability to promote together with other fintechs and the financial inclusion in Mexico over time. Mario, did I forget any of your questions?

No, no, that's helpful. And then on the NPLs. And just to clarify, like when you say, right, that the fees are a smaller percentage, are we talking about like 15%, 20%. Any idea that you could give to us?

No, we don't provide this breakdown. But I think if you take a look at the financial statements that we posted with the regulators in Mexico, you will largely get a good proxy of kind of the weight of each of those components for us. But I think even if the, Mario, for example, let's assume that interchange accounts for a small portion. If you cap this in the magnitude that has been proposed by the government, the existing kind of business plan that we have a significant portion of the new customers of the ones that we would bring from informal related to the bank may be compromised, right? So we do believe that it's our obligation and duty to be able to share this very openly with the stakeholders in Mexico to continue to foster the competition and financial inclusion that we want to do so.

Okay. And the second part of my question was on the NPLs in Mexico, as Mexico becomes more relevant, right? Like are you going to disclose the NPLs for the total group rather than just Brazil? And if you can make any comments on how that is behaving in Mexico and the coverage they're using?

Yes. No, absolutely. We do expect that as Mexico gains relevance in our overall credit portfolio, we will start providing kind of a much more granular disclosure on its asset quality. Today, it still accounts for less than 10%, 15% of our overall book, but we are certainly able and willing to provide those levels of the -- asset quality and asset performance in Mexico overall has been a fairly good story for us. I think we spent the good part of 2023 and 2024 kind of sharpening the data stacks and the models.

And what you have seen over the past 12 to 18 months, it's a relatively strong acceleration of the growth of our credit book in Mexico, growing at a clip of about 50% to 70% on an annualized basis. But more than the top line growth or the size of the book, the asset quality has performed very much in line in some cases, even though better than expected.

Also, we have recently launched kind of the lending product in Mexico. We have been kind of working primarily with credit card, and lending has been doing really, really well in Mexico. I wouldn't be surprised if differently from Brazil. At some point in time, lending becomes an even bigger business for us in Mexico than credit cards.

So I would say, yes, the left side of the balance sheet has expanded nicely in both kind of quant and quality. And then on the right side of the balance sheet, as you may have seen, Mario, we have been kind of sequentially redesigning and repricing deposits it has led to a fairly substantial drop in cost of funding in Mexico. And still preserving what we see very intensively there, which is primary banking relationship, transactionality, activation, so forth, most of the customers. It has actually been going up. We are an all-time high of transactionality there. So very excited with Mexico with what we're seeing. Still early days. But as David mentioned, it's playing out to be as strong, if not even stronger than Brazil.

Thank you. It's very impressive how quickly and how profitable you're growing in Mexico. Thank you.

Operator, could you please open the line for Mr. Marcelo Mizrahi from Bradesco BB.

Thank you very much for the opportunity to participate. So my question is regarding the cost of risk again. So you understand what drives the cost of risk to go down. But as Lago has said, so about the campaign to recovery to bring back clients, so we are already seeing the number of active cards going up. So the question is, looking forward, this level of cost of risk seems that is the new level in the next quarters? So the growth of the NIM will come with this proportionality, so far more from the cost of risk than from the net interest margin.

So thanks for the question. I think it will -- in terms of NIMs, starting with your -- the latter part of your question, it will be a function of both asset mix as well as LDR, right? So I think as we increase kind of -- we continue to increase the ratio and the weight of secured lending in our book. We could eventually see the continuous kind of lowering of the asset yield.

But as LDRs go up, we should expect to see kind of a NIM even expanding potentially. So it will depend on the velocity with which we increase kind of our credit assets versus the velocity with which we continue to increase deposit in both Brazil, Mexico and Colombia.

In terms of cost of risk, we don't provide guidance on cost of risk in the short or in the long term. What we have been doing, as you have followed us for some time, is we have been kind of measuring and managing the business with a paranoid focus on the short-term data that we collect on the margins. So far, the data has proven to be fairly encouraging and reassuring for us to continue to grow the book.

However, as we have done in the past, if and when we see any deterioration in asset quality across any of the segments, any of the products or any of the geos, we will not hesitate to kind of -- to pull the brakes, reassess, revisit whether we will go. So that's one of the reasons why we are so hesitant to provide kind of a guidance on both top line as well as cost of risk.

Can I just a follow-up here on the LDR. So looking for the -- what is happening now in Mexico. So for me, it makes sense to see this -- part of this profitability coming from the leverage of the portfolio. So on the LDR part. Are you guys seeing that already or not?

Yes. Look, I think LDR in Mexico is about 15%, 1-5, right? So certainly, it's in many respects, one of the most liquid financial institutions that we may have in the region. Having said that, over the past 2 quarters, most of the expansion of NIMs in Mexico has come from the lowering of the cost of funding rather than any material changes in LDR. Going forward, however, I think that LDR will play a much bigger role than any material change in cost of funding.

Operator, could you please open the line for Mr. Thiago Batista from UBS.

I have 1 question, actually, 1 question of adding -- about regulation. The first part of the question is about mortgage with there was recent change regulation on saving deposits and mortgage in Brazil, do you believe it is possible to start to operate in this market in the near future? And second, on the FGTS loans, with the change that we saw probably 1 month ago or less than that, do you believe that FGTS loans will be reduced in a material way?

Thank you for your question. So we've looked at the market space in Brazil, and certainly, there are a number of attractive angles, specifically around principality. But it's not -- it's not a priority for us right now. It's not a product that I see yourselves really doing over the next couple of years.

The main reason for that is we think about our balance sheet fundamentally as a small balance sheet that is well capitalized that has very high velocity and very high return on equity. So from that perspective, we're going to be picking products, especially credit products that have short duration, very data-intensive that gives us the opportunity to react very quickly to changing macroeconomic environments and that maintains -- it gives us a lot of agility. And obviously, mortgages is kind of the opposite of that. It does -- it's very long-term duration, removes a lot of agility. It requires a lot of long-term funding. So it doesn't really match with the type of products that we want to be offering directly from the balance sheet and perhaps down the road, there might be an opportunity to partner with somebody to actually do it, but it's not something that we -- that we'll be prioritizing right now.

On FGTS, regulation, yes, I think the regulation would have a decrease of our FGTS originations. But given the size of the portfolio and the rest of the lending products that we offer, it wouldn't really be material. So yes, effect on FGTS, but not really a material effect overall on the portfolio growth.

Operator, could you please open the line for Mr. Gustavo Schroden from Citi.

Guys, congrats on the results. Thanks for the call. Most of my questions were answered. So let me do a follow-up here. The first one is I'd like to understand better this -- the asset quality. Indeed, the bit was on the lower ECL. So despite this some metrics like 90 days NPLs relatively stable and early NPLs improving. When we analyze the transfers to Stage 3 in both credit cards and loans, it is continuing increasing, right? I mean it's rising. So I'd like to understand how we reconcile this increase in transfers to stage with this, let's say, a lower risk credit portfolio you are adopting and this lower provision expenses in the quarter.

Thanks for the question. Look, we've been -- I think later than what we've already mentioned related to the better-than-expected asset quality performance in some of the segments, especially with credit cards Brazil. I think the order, no positive surprise I wouldn't say surprised in the positive outcome that we have had after kind of many months and years of investment is also on the ability to improve our collections, engines and platforms which has had kind of a material improvement in Brazil. And I think it will start to have material improvements in Mexico, most likely starting in the fourth quarter of 2025.

But other than that, it's just kind of a general performance of the portfolio. There's nothing atypical or nothing abnormal that you would have seen over the past 2 to 3 quarters that we wouldn't expect to continue seeing in the next 2 to 3 quarters unless we see kind of material changes in macro.

Okay. Okay. Understood. And the my follow-up would be regarding Mexico. Assuming this, let's say, improving in cost of funding, we follow data from Mexico and we can see that you are improving the cost of funding there, assuming a potential improvement also in loan-to-deposit ratio. And we also followed the NPL ratio, and we see the NPL ratio next when your control. So do you think that assuming these trends you are posting Mexico, we can expect like some positive ROE, our bottom line in Mexico soon.

So Schroden, I wouldn't guide in any way or form as we haven't done in the past on kind of the P&L or net income either for the company or for any of the legal entities. So I would stay away from trying to provide you any high conviction outlook on when we're going to become net income positive for ROE.

That said, I'm much more comfortable providing you with our impressions of the profitability potential of Mexico. If you take a look at our business in Mexico, it is it posts actually unit economics that are as compelling, if not more compelling than Brazil. It has higher ROA. It has higher ROE and allows us to actually provide kind of with material credit asset, no access to a portion of the population that it hasn't yet had no access to credit.

If you take a look at the more than 13 million almost 14 million customers in Mexico, about 20% of those did not have access to kind of a banking or credit before joining new bank. And we think that in Mexico, we enjoy a very favorable cost structure compared with many of the other players in the region that allows Nubank to play at segments that incumbent banks are enabled and -- will it play to price it lower and still have compelling kind of unit economics. And we are super excited with what lies ahead in Mexico. We are still very, very, very early.

But as we continue to gain scale, we will see kind of economies of scale and operational leverage playing out there. In fact, today, Mexico already has a cost to serve that is about $1, which is much better than what Brazil had when it was at the same point in time of development of Mexico, and it already has kind of very encouraging ROEs and ROAs trends.

The question in Mexico become how fast the economy will truly digitalized and how much kind of a banking penetration will grow. We are now excited not only to witness this but also to be a very active agent in promoting this together with other players in the industry and with the Mexican government.

The other point I would just add here is that if we wanted to be profitable in Mexico, we would be profitable already. It's a decision. We literally touch about on mechanization and we're profitable immediately. We have already had the scale to generate that profitability. But that would actually be a really bad decision. It would be sacrificing the future for a short-term decision. We've always told investors we're optimizing for the long run. We're really optimizing to try to make investments that will pay for long as possible. And this is a very attractive market.

Another data point that I do think Lago mentioned previously is -- on the ARPAC question, Visa country has a 40% higher income per capita than Brazil and where credit cards are majority, about 80% of revolvers versus only in Brazil, about 10% to 15% for our portfolio revolvers.

So anyway, it's a big market, low penetration, a lot of the advantages that we have, like our capabilities on credit underwriting, the efficiency ratio. Good unit economics provide a really compelling investment opportunity. And so we'll continue investing the excess capital that we have in trying to maintain a leadership position in the country.

Okay. Okay. Okay. That's a great answer. And as I said, we have followed your data in Mexico in the -- and we can see these trends improving. This is why I was asking about the potential profitability maybe sooner than we were expecting. But thanks again, and congrats on the results.

Rate, could you please open the line for Mr. Tito Labarta from Goldman Sachs.

First, I have a follow-up question, Lago, your comment on the higher interest expenses just driven by higher funding costs. in Brazil and SILIQ being a little bit higher, I think. But just to understand, because average SILIQ only increased modestly in the quarter. Was there any impact perhaps from just more working days in the quarter. You had also launched the Turbo money boxes. I was wondering if that had any impact. I was a bit surprised by how much the interest expenses jumped.

So no, Tito, you're right, there were a few additional working days in the quarter, but it would have been equally offset by the revenues as well. But what you will see is that as we kind of took a more aggressive stance on the segmented portion of our deposits in Brazil, by which I mean for a selected profile of customers that we think that our primary bank relationship customers or are prone to become primary banking relationship customers. We have been more aggressive with the money boxes with the Turbo Caritas and that has all else constant increase our cost of funding in Brazil. So that is unequivocal observation.

What I was trying to allude only, Tito, is that how would you reconcile what I've just said. We have Slide 14 which is where we show kind of the cost expressed as a percentage of interbank rate coming down. And I was just trying to say that the reason why it comes down is because we do a weighted average of percentage of CDI, a percentage of IBR and percentage of tea in Mexico. And as both Colombia and Mexico went down. Then line here on Slide 14 goes down, notwithstanding the fact that overall cost of funding denominated in dollar has gone up because of our deliberate intention to play more aggressiveness on the segmented roles of Caritas in Brazil.

Okay. No, that's very clear. That helps clarify a lot, yes, I mean we expected funding costs in Mexico and Colombia to come down. It was just a little bit surprised by how much that had gone up specifically in Brazil, and as you mentioned, more than offset by the revenues.

So my second question is somewhat on the revenue. Just thinking on the loan growth, very good loan growth overall. But first on the secured lending, right? And David, you mentioned your FGTS could be a headwind, but it shouldn't have an impact. Do you expect that to be potentially offset by private payroll loans? I don't think you're necessarily growing significantly there. Just think about what could offset that potential headwind would be helpful.

Yes. No, absolutely, Tito. Look, let me put it this way. the secure lending class or segment that we define here is largely composed by FGTS, public payroll loans and private payroll loans. So grossly, those are the 3 components. You do still have a smaller portion that we call IPL, investment backlogs, but that's a much minor portion.

We do expect to see a material headwind in terms of FGTS in the new regulation kind of prevails. But we do believe that this will be more than offset by an increase in public payroll loans at this point in time, more so than on private payroll loans. On public payroll loans, we have seen a fairly material uptick in our ability to originate public payroll loans in Brazil. In the third quarter, we expect to see this in the fourth quarter as well. And as we see in nominal interest rates in Brazil finally coming down, we do expect to see portability going up as we have seen in all of the prior cycles, and that will give us the opening to actually get a disproportionately higher shares of the public payroll loan market in the country.

Now going to your third and final piece, which is private payroll loans. We are very, very bullish on this product in the medium and long term. We are still more cautious than some of the other players in the industry with respect to its cost of risk. Mostly related to what we call employee-related collateral, but we are seeing this kind of improving quarter-after-quarter, and we are reflecting and watching this very carefully on when and how we will lean in more aggressively in the future. It's not something that we are taking as a base case now, but we are certainly paying very close attention to that. For now, public payroll loans is the one that will offset the slowdown in FGTS Q2.

That's very clear. Thanks a lot, a congrats on the results.

So thank you, everyone. We now have approached 60 minutes of the call. So we are now concluding today's call. On behalf of new Holdings, our Investor Relations team, I want to thank you very much for your time and participation on Nu earnings call today.

Over the coming days, we'll be following up with questions received tonight, but we are not able to answer. And please do not hesitate to reach out to our team if you have any further questions. Thank you, and have a good night.

The Nu Holdings conference call has now concluded. Thank you for attending today's presentation. You may now disconnect.

Good evening, ladies and gentlemen. Welcome to Nu Holdings conference call to discuss the results for the second quarter of 2025. A slide presentation accompanies today's webcast, which is available on Nu's Investor Relations website, www.investors.nu in English and www.investidores.nu in Portuguese. This conference is being recorded, and the replay can also be accessed on the company's IR website.

This call is also available in Portuguese. [Operator Instructions] [Foreign Language] [Operator Instructions]

I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer at Nu Holdings. Mr. Souto, you may proceed.

Thank you, operator, and thank you, everyone, for joining the earnings call today. If you have not seen the earnings release already, a copy is posted in the Investor Relations website. With me on today's call are David Velez, our Founder, Chief Executive Officer and Chairman; and Guilherme Lago, our Chief Financial Officer.

Throughout this conference call, we'll be presenting non-IFRS financial information, including adjusted net income. These are important financial measures for Nu Holdings but are not financial measures as defined by IFRS and may not be comparable to similar measures from other companies. Reconciliations of the non-IFRS to the IFRS financial information are available in the earnings press release. Unless noted otherwise, all growth rates are on a year-over-year FX neutral basis.

I would also like to remind everyone that today's discussion might include forward-looking statements, which are not guarantees of future performance, and therefore, you should not put undue reliance on them. These statements are subject to numerous risks and uncertainties and could cause actual results to differ materially from our expectations. Please refer to the forward-looking statements disclosure in the earnings release.

I will now turn the call over to David. Please go ahead, David.

Hello, everyone, and thank you for joining us today. In Q2 2025, we delivered another quarter of strong growth as we continue to strengthen our position as the leading digital bank in Latin America and one of the leading financial technology platforms in the world, our customer base expanding to nearly 123 million customers with over 4.1 million net additions, all while maintaining an activity rate above 83%, underscoring the depth of engagement across our platform.

In Mexico, we surpassed 12 million customers now serving approximately 13% of the adult population. And in Colombia, nearly 10% of the population is already choosing Nu as their financial partner. The combination of sustained customer growth and a 34% ARPAC CAGR since 2021 has created a powerful compounding effect, driving revenues to $3.7 billion in Q2, representing an 85% annualized growth rate since 2021.

Gross profit has risen 78% annually, reaching $1.5 billion as we capture benefits of scale, cutting our efficiency ratio by more than half to 28.3% in Q2 2025. Quarterly net income has almost tripled in the past 2 years to $637 million. These results come despite our ongoing investments in growth and most importantly, in keeping our customers loving us fanatically.

And we will continue to invest with focus and intention. This performance reinforces a key message. Growth isn't coming at the expense of sustainable results. Quite the opposite. We're proving that it's possible to scale efficiently with discipline and still generate strong earnings.

Taken together, these elements have broadened our platform into a powerful multiproduct, multisegment and multi-geo growth engine. Today, 104.7 million mass market customers, 3 million high-income clients and 5.2 million small businesses engage with Nubank through a diverse suite of products ranging from credit and insurance to investments in crypto. This spread is no accident. It is the result of a deliberate cross-sell strategy that expands a single product relationship into a broad ecosystem.

By meeting customers' needs at every stage of their financial journey, we don't only deepen loyalty but also multiply the ways we can create value. This broad-based momentum is reflected in Q2. The active unsecured loans customer base expanded 56% year-over-year, while the secured customer base more than doubled and crypto customers increased 41% year-over-year.

All segments continue to post solid growth. And in our less mature countries, our core credit card franchise is scaling quickly. Card customers rose 52% in Mexico and 34% in Colombia.

We're not only scaling. We're locking new markets, pioneering adoption in underpenetrated segments and building the foundation for the long term. And as we continue to grow and deepen customer relationships, we're doing so with a business model that delivers results and adjusting growth but also in profitability.

As we look ahead to this next chapter, having the right leadership in place is more important than ever. We've recently made significant additions to our management team that elevate our ability to execute on our long-term strategy and deepen our leadership bench. Over the past few months, we welcomed 3 truly exceptional leaders to Nubank. Roberto Campos Neto joins us as Vice Chairman and Head of Public Policy. As the former Governor of the Central Bank of Brazil, Roberto brings not only unmatched regulatory insight but also a strategic vision for how technology and policy can shape more inclusive financial systems.

Eric Young, our new Chief Technology Officer, brings deep expertise in scaling complex tech platforms and leading high-performing engineering teams at a global scale, having run and led products that reach over 900 million customers around the world. Ethan Eismann, our new Chief Design Officer, is a world-class design leader with a track record of building intuitive human-centered digital experiences that delight hundreds of millions of users around the globe.

All 3 are joining Nubank's management team and will report directly to me. They're world-class experts in their craft and just as importantly, seasoned business leaders with experience and judgment to help guide Nubank through our next chapter of growth.

If there's one thing that has to defined Nubank since day 1, it's our people. We've always had the right team for each stage of our journey, leaders who are not only exceptional in their domains but who elevate the company around them. That remains true today. These additions reflect our ongoing commitment to having the best possible team in place for the next cycle, a cycle that will require even greater scale, complexity and ambition. Together, Roberto, Ethan and Eric represent kind of talent advantage we believe is one of Nubank's greatest strengths, a dream team for where we're heading next. We're thrilled to have them onboard, and I want to offer a very warm welcome to all 3.

With that, I'd like to pass the floor to our CFO, Guilherme Lago, who will walk us through the details of our financial results. Over to you, Lago.

Thank you, David, and good evening, everyone. Let me start by reinforcing how our business model creates value. We acquire customers at scale, increase engagement over time, monetize as cohorts mature and we do all of this on a low-cost, highly efficient platform.

On the left side of this slide, you see monthly ARPAC consistently increasing across all cohorts, reaching $27.3 for customers who have been with us for longer. And even among these more mature customer cohorts, monetization keeps expanding. In the second quarter of 2025, our monthly ARPAC crossed the $12 mark for the first time, reaching $12.2, up 18% year-over-year. Meanwhile, as you can see on the right side of the slide, cost to serve remains stable at $0.80 per active customer, reflecting the efficiency of our platform. This operating leverage is one of the most important and competitive advantages of Nubank. It is what allows us to offer better pricing to customers while consistently increasing our earnings power.

Moving to our credit portfolio. Total balances reached $27.3 billion in the second quarter, up 40% year-over-year on an FX neutral basis. All segments contributed to this growth. Secured lending grew 200% on an FX neutral basis, unsecured loans 70% and credit cards 24%. The continued diversification has been a mark of our credit portfolio quarter after quarter. Secured and unsecured loans now represent more than 1/3 of our total portfolio, up from 25% just a year ago. This shift in mix is intentional, and it speaks to our ability to expand our credit portfolio spectrum over time and better serve customers in every single market where we operate.

Moving to loan originations. We are operating a retail credit business at scale across Brazil, Mexico and Colombia. In the second quarter, we maintained a strong pace from the previous quarter, originating $3.6 billion in loans. That marks a 43% year-over-year increase on an FX neutral basis, and it is the highest origination volume we have ever reached. This consistent origination growth reflects both the sheer size of our consumer platform and the maturity of our credit underwriting engine in Latin America.

Turning to our credit card portfolio in Brazil. Installment balances remain the primary component of our interest earning portfolio. This reflects our strategy of promoting more structured and predictable forms of credit helping our customers finance purchase and transfers in a responsible way.

Our mix is fundamentally different from that of the industry. While many players rely heavily on revolving balances, we have been building a more sustainable model centered on lower risk, lower cost interest earnings installments, and this translates into better products for our customers and healthier unit economics for Nu.

Now turning to the other side of the balance sheet, funding. We continue to execute our strategy to build a scalable and sustainable deposit franchise across Latin America. Total deposits reached $36.6 billion in the second quarter, up 41% year-over-year on an FX neutral basis. Brazil remains the anchor of our deposit base, but we are also seeing strong progress in Mexico and Colombia, where we have expanded both volumes and attach rates.

This deposit growth is a core pillar of our long-term strategy. It is what enables us to become the leading and most competitive retail financial institution in the region. We have been lowering deposit yields in Mexico and Colombia in the recent months, with some significant changes implemented only now in early July 2025. As a result, our second quarter cost of funding did not yet fully reflect these adjustments. We expect the full impact to materialize only gradually and over the coming quarters.

Now turning to net interest income. We delivered strong growth again this quarter, up 33% year-over-year on an FX neutral basis, reaching a record high of $2.1 billion in the quarter. NIM improved 80 basis points quarter-over-quarter on an FX neutral basis even with a slight reduction in our loan-to-deposit ratio, which went from 44% to 43%. In our most scaled market, Brazil, NIM continued to expand, supported by healthy spreads and growing volumes.

In Mexico and Colombia, we continue investing to become the leading and most loved retail financial institution in these countries. While these investments, however, naturally weighed on the short-term margins, we believe they are critical to unlocking long-term value. Looking ahead, we see further room for margin expansions as we optimize the balance sheet, gradually reallocating liquidity from cash into credit and lower our cost of funding in Mexico and Colombia.

Now on to credit loss allowances and risk-adjusted NIM. CLA expenses remained relatively stable in the quarter. In early Q2, we began rolling out our major upgrades to our credit models. This will significantly increase credit card limits in Brazil throughout the remainder of 2025. As a result, we recognized provisions this quarter, front-loading expected credit losses, which have not yet been fully offset by the corresponding growth in the interest earning portfolio and related revenues, naturally creating a temporary timing mismatch.

Now excluding this effect, credit loss allowance would have declined quarter-over-quarter on an FX neutral basis, reflecting the normalization of seasonal dynamics that had impacted Q1. Now despite these dynamics, strong NII more than offset the small increase in CLA expenses, driving our risk-adjusted NIM up to 9.2% in the second quarter of 2025.

Next, delinquency metrics for our consumer credit portfolio in Brazil. The 15- to 90-day NPL ratio declined to 4.4% in the second quarter, a 30 basis point improvement versus the previous quarter. This was in line with our expectations and slightly better than the typical second quarter seasonality, which usually shows a 20 basis points drop. Now the 90-plus day NPL ratio increased by 10 basis points to 6.6%, reflecting the rise in early delinquency observed in Q1 and following the usual seasonal pattern. Finally, coverage ratios remained solid and stable. We continue to carry a fairly robust provision buffer, both across the total portfolio and specifically across the 90-plus NPL balances.

Shifting to gross profit. In Q2, gross profit reached a record high of $1.5 billion, up 24% year-over-year on an FX neutral basis, a clear reflection of the strong momentum of our business. This performance was driven by strong NII expansion and stable credit loss allowances. Gross profit margin also improved sequentially, climbing now to 42.2%, up from 40.6% in the past quarter.

Looking at the composition of our gross profit, we continue to see the benefits of our business model not only in terms of growth and profitability as we have seen in the prior slides but also in terms of diversification and resilience. By leading with credit, we drive stronger engagement and deepen customer relationships over time, which unlocks cross-sell and increases shares of wallet. But fees and float have also become meaningful contributors to our gross profit and have added resilience and consistency to our revenues across cycles. Ultimately, being a credit-first fintech has helped us ignite what we call the principality flywheel. And with that, we have earned the right to cross sell and diversify our gross profit base.

Now turning to efficiency. In the second quarter, our efficiency ratio rose slightly to 28.3%, driven by 2 main factors: number one, RSU expenses from the initial vesting of our 2025 annual grant, which typically happens around March of every year; and number two, higher marketing investments during the quarter. Now as David mentioned earlier today, we are investing with intention to become the largest and the most loved financial institution in Latin America. While these investments may temporarily increase our efficiency ratio in the coming quarters, they are fully aligned with our long-term value creation strategy.

The long-term trajectory remains intact. Our model continues to benefit from operating leverage with significant room to unlock additional efficiencies as we scale. Supported by strong revenue growth and disciplined cost management, we expect the efficiency ratio to further decline over the coming years, driving, number one, continued margin expansion; number two, sustainable profitability; and number three, deeper competitive moats.

Before we wrap up, it's important to highlight how our business model consistently deliver bottom line performance and does so at scale. Net income reached $637 million in the second quarter, up 42% year-over-year on an FX neutral basis. Return on equity reached 28%, continuing to track well above industry peers. Now what makes this performance especially notable is how we got here, by charging lower prices and offering better experience to our customers while still delivering strong bottom line results. And we are just getting started, which brings us to Mexico, where we are seeing encouraging momentum and a clear path to scale.

Customer growth is accelerating, and our core product, credit cards, is scaling. We reached 6.6 million credit card customers this quarter, up from 4.3 million a year ago. Over the last 12 months, we accounted for more than 1/4 of all Nu credit cards issued in Mexico. This is a clear sign of our early success in expanding access to credit in the country. At this stage in Mexico, our most important KPIs are: number one, growing a solid and engaged customer base; number two, building a large and resilient local currency liability franchise; and number three, continue to improve our credit underwriting models to approve more customers and drive sustainable portfolio growth.

On the funding side, our liability franchise continues to show signs of strength. Even after adjusting down our deeper rates, deposits continue to exceed $6 billion, underscoring the value of our brand and the appeal of our products. Our interest-earning portfolio has gained strong traction recently, growing over 70% year-over-year on an FX neutral basis. We will continue to scale credit but at the right pace, accelerating when the signals are clear and consistent with our long-term strategy in Mexico. And we will never hesitate to pull back if and when the situation requires. We are very confident in our opportunity to win in Mexico, and our focus remains on disciplined execution and long-term value creation.

With that, we will now open the call for questions. Thank you.

[Operator Instructions] I would now like to turn the call over to Mr. Guilherme Souto, Investor Relations Officer.

Thank you, operator. Could you please open the line for Eduardo Rosman from BTG?

Congrats on the numbers. I have a question for David. In recent months, we have seen important changes, right, in the management team, including the announcement of a new CTO this week. So could you please help us understand the significance of these changes for Nubank in this next phase? And please, if you could also connect this topic to the company's kind of international expansion, right, that would be great as well. Specifically, do these new additions suggest also a possible acceleration of the growth outside Brazil, including entering new markets beyond Mexico and Colombia?

Rosman, thank you. Thanks a lot for the question. We -- as we've said in the past, we have made this number of changes over the past couple of months really thinking about the next 5 to 10 years. We think we are -- we have ahead one of the most interesting opportunities in technology in the world. Financial services is still the largest market in the world that hasn't really been disrupted by technology. Over 95% of the market capital financial services globally, over $8 trillion, is still very much dominated by all traditional banks. And that is very different from what has happened in all our different segments.

So as we think about the next 5, 10 years, we are preparing to play in the world -- in the top leagues, in the world class. And as we prepare to play in the world leagues, we are bringing a world-class team. And this likely is going to mean adding talent sometimes that come from Latin America globally but also some talent that comes from some of the top world-class technology companies.

And so that is a bit what we are preparing here. I think the addition of Roberto is very strategic in helping us strengthen our positioning in Latin America. We have regulated entities in the 3 markets that we operate. We will have many more regulated entities later on as we internationalize.

So regulators are a key counterparty of us. We've always been ahead in terms of regulatory compliance, and we treat that very, very seriously. Public policy is a key aspect of also of what we do as a regulated financial institution. And in Roberto, we also were able to find all of that knowledge but also a lot of technology and strategy knowledge. So it was a very key strategic addition to the team, and it has been truly phenomenal to being able to work with him here and the team over the past month already.

And with the addition of Eric and Ethan, I think we are saying we are on the way to build one of the world-class products in financial services. We already have one of the most sophisticated technology stacks of any company in Latin America. We're in the middle of an AI transformation that we're taking extremely seriously, and we want to take advantage of all these opportunities that open ahead. And so I think as we bring somebody like Ethan with his knowledge of having run products for hundreds of millions of customers and the same also, Eric, we are just getting prepared for the next stage.

So to summarize, I do think these additions are -- help us both strengthening the market-leading position we have in Brazil and Latin America by upping up our game and also prepare us to really go play in the big leagues as we think about internationalization over the next few years.

Operator, could you open the line for Jorge Kuri for Morgan Stanley?

Congrats on the numbers. Great quarter. I wanted to maybe double click on your Slide 11, your loan origination. You have 1% FX neutral growth, which is a very different number from what we've seen over the last year, where the quarter-on-quarter growth were in the double digit. And I appreciate that the year-on-year number is really strong, 43% FX neutral, and that's certainly a better way to look at it. But just given some of the things that happened in the quarter, specifically on your credit line increases, your clips on credit cards that you started to implement and reach record levels for the company, given that you are extracting more value out of your Hyperplane acquisition, given that we saw a big acceleration of Pix at the end of the first quarter that we assumed it was going to continue or has continued during the quarter, so if you can help us understand if some of these things are just not reflected in the numbers, going to get reflected going forward. And any other dynamics for us to understand this different half of originations in the second quarter versus what we've seen over the last year?

Jorge, this is Lago. Thanks for the question. So let me try to unpack this in -- by asset classes. So let me talk about kind of the Slide 11 to which you allude brings kind of the evolution of originations only for loans. I'm going to try to address unsecured, then secured and then credit cards, which are not here.

So starting with unsecured, we have had now a fairly robust set of kind of growth figures over the past quarters. We had an exceptionally strong first quarter of 2025, especially because it was a quarter in which we launched a few new kind of models and policies that allow us to embrace customers that were not eligible for unsecured credit lines at that point in time. And usually, when you do so, you have kind of the first-time effect of early kind of adopters of the new policy, which kind of increases the origination volumes.

And also first quarter for unsecured credit is typically a seasonally strong quarter. We do expect, Jorge, that we will continue to grow unsecured lending originations fairly strongly throughout the remainder of 2025 and 2026 as long as we continue to see the asset quality numbers that we are seeing in our book. Not only until the end of the second quarter but until now today, August 14, everything seems to be super on track. We believe that we now account for over 20% of the origination market share of new unsecured loans in Brazil. And this should not only continue to gain speed, but it should be complemented by lending products in Mexico that have recently been launched. So feeling fairly good about the evolution of unsecured loans.

Now going to the secured loan story, Jorge. I think, here, I would have to split the story here in 2 sub-asset classes. So you have the INSS and the public payroll loans excluding INSS. So for those who are not aware, INSS is the public payroll loans directed to pensioners and retirees. In the second quarter, there was kind of a major disruption in the INSS system. So the overall volume of the origination of the industry dropped by more than 50% to 55%. And our origination dropped by about 50% as well. We even gained market share there but in a declining kind of origination quarter. We do expect that this will be fixed and resolved very promptly. We are assuming that by end of August, early September, origination of INSS, not only for us but for the entire industry, will resume their historical growth. Most likely, in fact, we may see actually a spike in originations in the next months to offset the lower originations.

Now if you exclude INSS for all of the other public payroll loans, Jorge, especially SIAPE, our originations grew by more than 50% in the quarter. So we are making very good strides in our view in the ramp-up of the public payroll loans there not only by kind of adding more customers to existing contracts but also by adding more contracts kind of to our portfolio of collateral agreements, a few of which will come into force in the second half of this year.

And then if you look at the overall evolution of our portfolio, then I would draw your attention to Slide 27, Jorge. You will see that even with kind of this one-off headwind from INSS, we continue to see all of the asset classes expanding at what we believe to be a fairly healthy pace. So in the quarter, portfolio grew by 8% FX neutral, and that growth was followed by loans, credit cards IEP and credit cards, non-IEP.

And my last attempt to address your question. You also mentioned credit cards. So look, credit cards, yes, we have been seeing kind of fairly material improvements in our ability to do credit underwriting and to continue to expand the credit card portfolio. It has to do with the adoption of new models and technologies to how we do credit underwriting, going all the way to better kind of traditional machine learning models but also neural networks and predictive AI technologies but more and more, Jorge, by the adoption of new data that we acquire, right?

So the more customers stay with us, the more data we accumulate. We are now the leaders in open finance consent. The combination of better modeling technique with more data has allowed us to consistently increase kind of credit underwriting, credit limits and utilizations. Based on our latest reading, our market share in Brazil in credit card receivables may have grown by more than 100 basis points in this quarter specifically. So we are fairly encouraged by what is ahead of us, not only in the existing segments but also as we expand into new segments.

Lago, that was super clear. And congrats on the quarter and all of the great new hires.

Thanks, Jorge. Operator, could you please open the line for Yuri Fernandes from JPMorgan.

Also congrats on the margin, the risk-adjusted margin expansion and the good quarter. I have a follow-up also to Lago just on asset quality. Most metrics, they look good, right, stable overage 15 to 90 days improving slightly better than seasonality. The only thing that caught my attention, Lago, was a higher Stage 3 formation, up quarter-over-quarter. I would like to get your view on this because when I go to 2024 and 2023, I also saw some seasonality in the second Q. So just checking if this is basically seasonal. From your answer to Kuri, I get an impression that you feel comfortable with asset quality but even we have many investors concerned with the macro situation in Brazil? It would be good to get your feeling on the formation and also how you see asset fund.

No, thanks, Yuri, for the question. So the short answer is yes. I think the increase in NPL formation as well as in Stage 3 formation that you can see on Slide 26 of our presentation is almost entirely explained by the seasonality of basically the spike in seasonal delinquency in the first quarter kind of flowing through the second quarter.

Now broadly on asset quality, we are fairly mindful of the macroeconomic kind of situations in the markets where we operate, Brazil, Mexico and Colombia, also how it may impact credit cycles. And this is a concern that has lingered not only with us but also with many investors and other stakeholders since late last year and early this year.

So far, and I say so far until now, August 14, we haven't seen kind of that deterioration playing out materially in our asset quality figures. All of our asset quality figures are performing largely as expected. That doesn't, of course, mean that we have to assume that this will stay as it is going forward. We continue to underwrite with kind of largely 2 kind of pillars in our mind. Pillar #1, we always assume that the future will be worse than the past. So irrespective of where any of us here the company may think we are in the credit cycle, when it comes to credit underwriting decisions, we always assume that there's going to be a deterioration in the credit cycle over the next 12, 24 and 36 months.

And then above and beyond that, which is Pillar #2, every cohort of unsecured credit that we underwrite has to abide by the following kind of a stress test, which is losses have to go up -- can go up by up to 2x, and that cohort still has to be NPV positive. So with that, we built enough credit buffer resilience that will allow us to continue to grow conservatively and with conviction that we can withstand kind of unfavorable economic cycles over time.

Operator, could you please open the line to Geoffrey Elliott from Autonomous.

Could we talk about the mix of credit card balances? The last 5 quarters, interest earning installments have been between 27% and 29% of total balances. Are we now in a range which is normalized and where you'd expect to stay? Or is there scope for that to move higher with increased originations of Pix credit?

Thanks, Geoff. Look, I think I would say they should stay more or less where it is, maybe kind of small variations up or down, maybe a little bit upside risk here depending on how pronounced Pix financing and other transaction financing products may unfold. But I wouldn't suggest that there is a lot of room for us to go materially beyond the 29% that you alluded.

Operator, could you please open the line for Neha Agarwala from HSBC.

Congratulations on the numbers. Quickly on the deposit side of the franchise, 2 notable trends. First, on the Brazil, there was a big pickup sequentially on the deposits. What was the driver for that? Are you trying -- are you being more active consciously in trying to gather more deposits in Brazil? So any explanation on that?

And on the Mexico side, you brought down the rates quite significantly, lowered the gap versus the [ TA ]. What have been the early reactions from the customers? Are you seeing outflow of deposits in July, early August? Or has that been fairly stable?

Neha, thanks for the question. So let me try to break it down. In Brazil, we did see or we continue to see an increase in deposits there. I think that has to do primarily with the increase in engagement and share of wallet that we have had with our customers. I would not ascribe a lot of value to that to any kind of initiatives to pay up for deposits in Brazil, even though we have launched a few new features that kind of a reward customer engagement and loyalty over time.

If you were to compare, for example, the cost of funding of Brazil in isolation, it would have been practically unchanged over the past 2 quarters at kind of low 80s. So I wouldn't justify the increase in deposits based on increase in cost of funding but largely on increasing customer engagement and sequential gains in shares of wallet. Also, kind of progressively as we make some strides into more affluent segments, it's natural that we should also see increases in deposits over time. So that's the story about Brazil.

The story about Mexico, just to maybe put everyone in perspective, Neha, if you allow me, so we did announce some material shifts to the design and the pricing of our deposits in Mexico in early July, and that is expected to lead to kind of the lowering of our cost of funding in Mexico. None of that, however, is reflected in the numbers that we see here in the second quarter. So those are things that we will see throughout the remainder of 2025.

Now back to your question, Neha, look, we have been watching this super carefully since we've made the movements. Everything has been kind of evolving as expected. As we continue to offer the -- what we call the money box capped now at MXN 25,000, it basically allows us to even better serve and offer an even stronger value proposition for the vast majority of our customers, nearly 90% of the customers. And so we believe that we will be able to maintain customer engagement NPS at the segments that we care the most.

We did run kind of some risk of having what people call the yield seekers eventually moving their money out. That outflow has not been material so far, even though we watch this carefully. So so far, so good. Now if things continue to play out as we have seen, we do expect to be able to continue to have a fairly robust local currency, low-cost retail deposits in Mexico that has already materially derisked our funding strategy in the country but progressively at lower funding costs.

I would just add one more point here to Lago, which is the following. When we launched Mexico, our savings account product was fairly basic. It was an online saving account without a lot of the functionalities. We didn't even have ability to allow customers to deposit off-line or to withdraw, which is very key in a market like Mexico. So in a way, it was a product -- it was truly an MVP as we launched. That meant that we had to pay higher yield.

As we launch additional products and the product gets much more robust, we've added OXXO as a distribution channel. Customers are now able to withdraw cash. We've added a number of, what we call, self-driving bank functionalities inside our app. Then the value proposition increases. That means we have to compensate less on the yield. And that's why we've been able to decrease yield without seeing significant changes in the flows of deposits we're seeing into Mexico. And the same strategy has really been applied to Brazil and Colombia.

Super clear. If I can have my follow-up on a separate topic but a brief one. The Hyperplane expansion and the credit limit that you talked about, is there any particular segment of customer base where it is more targeted towards higher income or mass market or your super core segments because that will eventually have an impact on probably stronger loan growth in the second half of -- or in 2026 for your loan book?

So far has been mostly focused on mass market, but we expect that a lot of these new AI-enabled architecture will be now applied to a number of different models. The amazing opportunity of Hyperplane is that it's not only a modeling -- the team did not only bring a number of modeling capabilities but also a true new platform that allows us to put into production and develop a number of different models at the same time.

And so this model was the first one. We expect a number of new models coming in for a number of different segments for the different countries and for different applications, such as collections, fraud, cross-sell. So we're very excited about this, and it's early days of applying this new technology to a lot of the decisioning that we have across Nubank. But we expect to see meaningful changes across the board.

Operator, could you please open the line for Pedro Leduc from Itau.

Both on cards, please. First, the number that you give us, and this has to do with Pix financing, that the number of clients using it transactionally fell a bit, 17.3 to 17.1 this quarter. In the last call, you had mentioned that you had slightly become more comfortable to gradually resume the product to those certain clusters you had withdrawn from after tests have worked well. So can you give us an update on the Pix financing process? When -- how you see it roll out, when you can see it get more traction?

And then the second question, just -- it's sort of related, has to do with the number of active credit cards and that has also fallen a bit. And we can see that you're rolling out more limits. I'd sort of expect the opposite, no more active cards. So if you can help us square this out a little bit.

So Leduc, thanks for the question. Let me try to touch on each of them separately. So Pix financing. Pix financing or better said, the whole transactional financing kind of a family of products, of which today Pix financing is by far the biggest one, but that continues to grow. We did show in the last quarter that we had resumed growth there that we were at the end of the first quarter already at a -- with a bigger kind of Pix financing and transactional financing portfolio than we had in the second quarter of 2024 when we decided to pull back. That only continue to increase. So today, we are even ahead of what -- where we were in the first quarter of 2025.

The performance of the portfolio continues to be fairly robust, and it has been widely adopted by our customer base. So you may see, Leduc, here a few kind of noise in seasonality when you go from 1 quarter to another. But by all measures, it has been kind of a remarkable success for the customers. It has been adopted by -- so I think as of the end of the second quarter, over 40% of our credit card customers were also active with some type of transactional financing, primarily Pix financing functionality. So the attach rate there is very high, very healthy, not only in the first order but also in the second order impact.

So we do expect this to continue to grow. It certainly has a high correlation with the overall usage of Pix in the economy, but we don't have any concerns as we discussed when we presented the results of the third quarter of 2024. So far, so good.

Second question that I also wanted to address, which is credit cards. So we -- the number of active credit card customers in Brazil, depending if you measure this in terms of purchase volume or in terms of revenues, has remained largely flat, right? So it's like in 1 measure, it goes by plus 100,000. The other one -- the other measure goes less than 100,000, but overall, it has remained flat.

What we do expect to see going forward is that the main lever of earnings growth for our credit card business in Brazil will mostly come from the increase of utilization in ARPAC per product rather than the increase of the number of credit cards. It doesn't mean that the number of credit cards will not grow. Yes, it will continue to grow, but I think the biggest leverage will be in ARPAC and utilization.

And then you mentioned, look, if you are increasing eventually credit limits, shouldn't you see necessarily an increase in a number of active customers. Not yet especially because the credit limit policies that we have implemented has been directed primarily at existing customers, not at initial lines. It is only natural that as we collect more data as we continue to improve the models, what you suggested will likely happen as well.

Operator, could you please open the line for Mario Pierry from Bank of America.

Congrats on the quarter. Lago, I wanted to discuss a little bit the private payroll product. You -- it doesn't seem like Nubank is too excited about the product or at least I haven't heard you guys talk too much about the opportunity. When we talk to other industry players, right, they think this is one of the best products to come to Brazil in the last, I don't know, 20 years. So -- and again, they talk about the potential size of this market being significant. Can you discuss a little bit about your strategy in the private payroll product? Like looking at origination data, right, we haven't seen Nubank being very active yet. When do you think you're going to be more active? Why are you holding back? Are you seeing this product as an opportunity? Or do you think this is a threat to your business?

Mario, thanks for the question. Look, we are very excited about private payroll loan product. We think that has been a fairly important and thoughtful product innovation that has been added to Brazil. And I think new bank has much more to gain than to lose with this product by a wide margin.

Let me share a few thoughts on this, right? So we -- differently from kind of the more established kind of incumbent banks, we don't have kind of a B2B2C distribution channel to kind of -- to compete for corporate payroll loan business in Brazil, which is a fairly important one. And basically, the private payroll loan product allow us to basically break into this segment in a very profitable manner, right?

I don't need necessarily now to have a B2B contract with company X to be able to access all of the employees of this company and also enjoy the benefits of the payroll flow that goes through the bank account. I can have access to tons of data that, so far, we have been kind of precluded from having access to. So we expect that this product will improve kind of collaterals across the board. We'll lower data, symmetry and therefore, we'll help the overall economy lower spreads and materially increase the size of the pie. So far, so good.

Why are -- why have we not been kind of as aggressive at inception of this product? So the product was announced in late March. We launched the product right after this in early April. But we have not yet been kind of fully -- kind of we haven't been able to raise our comfort levels to adequate levels with respect to the quality of the collateral, right?

So I think some of the collateral that are structural to this product have not yet been fully tested and implemented. And the first data points that we have actually seen in the industry has suggested to us that the risk has not yet been fully addressed. So first payment defaults are set to be in the 10% to 18%, which I think it is higher than what at least we would expect so far. So we are not yet fully comfortable with the quality of the collateral, number one.

Number two, we don't see necessarily material first-mover advantages there. I think if we are the lowest cost manufacturer of this product, we will be able to secure a very meaningful market share position when the collateral system is more tested and solidified. But Mario, you've been doing this for a long time like many of us, and you may recall that, at the beginning of the public payroll loan systems, the consignado publico, the collateral was not working super well in the first months or even in the first quarters. It took some time, but the product actually end up being a remarkable success. We do expect that private payroll loan will follow suit, and we believe that as the lowest cost manufacturer of the industry, with no more than 50% of the target market for this product within our customer base, we will have a fairly relevant ability to win there when the product is more mature.

Yes. Okay. No, that's clear. Some of the players that we're seeing, right, already more active in this segment, they are talking about, yes, we're seeing higher -- elevated provisions and delinquencies, however, they still think that this is a 30% ROE product right now. So it feels like it could get even better. So I was just wondering, right, like I understand your concerns about the quality of the collateral. It's a new product. However, right, Nubank is always moving ahead of everyone else and trying to innovate. So that's why I was a little bit surprised that you're not more active right now, but I fully understand your concerns.

Operator, could you please open the line for Tito Labarta from Goldman Sachs?

Congratulations on the strong results. A little bit of a follow-up, but just how do you think about your loan growth along with your deposit growth, right? Because loan growth seems to be accelerating. You're doing well there, but deposit growth remains very strong. On the one hand, it's a headwind to earnings. But on the other hand, it's good for client engagement and client addition. Are you comfortable just continuing to grow that deposit book and get these clients even if it is a bit of a headwind to earnings? Or at some point, would you want to try to slow down the deposit growth to match the loan growth? Just how do you think about, I guess, the assets and the liabilities growing in conjunction?

Tito, thanks so much for the question. Look, a few thoughts there on how we are thinking about kind of deposits from both a financial standpoint but also from a strategic standpoint. I would say that from a financial standpoint, we are very comfortable with the loan-to-deposit ratio that we see in Brazil, Mexico and Colombia not only with respect to the quantum but also with respect to the duration and with the resilience of the retail deposits. If anything, we have buffer to either grow credit more rapidly, but we don't think that growing kind of credit more rapidly just because you have more funding is the most kind of wise approach to this, or we would have the ability to eventually lower prices and bring deposits down. That's from a purely financial standpoint.

However, from a strategic standpoint, especially in markets where kind of information asymmetry is lowering very fast, including in Brazil with open finance, we do believe that we need to be the best place for our customers across Latin America to receive payments, make payments and store value. And to that extent, having a very competitive and compelling kind of a deposit design, which includes, but is not limited, price is paramount to our primary banking relationship customers.

So we don't expect that we will play down with no deposit rates in Brazil anytime soon. We do expect that in Mexico and Colombia, we will continue to actively kind of reshape the size and the price of the deposits to optimize the value proposition for the customers, loan to deposit and liquidity resilience there.

So that's our thought process there. I think Mexico and Colombia, you should see increases in NIM as a result of that optimization. In Brazil, I think you should see kind of a relatively stable NIM with respect to the deposits only.

Okay. No, that's great. Very helpful. And if I can, just a quick follow-up, I guess. When you think of client monetization, I'm looking at Slide 18, [indiscernible] gross profit breakdown, right? I mean credit is still a big component. Fees have been around this 30% level for some time and then the rest is float. Do you think that's an optimal level? Is there an optimal level that you would like to get to? Just how do you think about that breakdown between, I guess, lending, fees and other sources of monetization?

So Tito, the one thing that I would point out is I would kind of respond to this by sending you to another slide, which is Slide 9, right? If you take a look at this slide, which is a very clear comparison between the revenues per active customers that we have and the cost per active customers that we have. Today, we have a weighted average ARPAC at about $12.2. The more mature cohorts are already at $27, $28. The ARPAC of incumbent banks are largely at $45. So we do expect over time that we will take the ARPAC from 12 to 15 to 20 and onwards towards the levels of incumbent banks, while our cost to serve will remain at or below $1. So that is kind of the power of the operating leverage of our digital banking model.

Now what are the main levers for us to bring kind of the ARPAC from $12 to $30 to $40? If you take a look at the profit pool of retail banking in Latin America as a proxy, about 70% -- 65% to 70% of that is credit. So credit is expected to be kind of the book of that growth going forward. That does not mean, however, that all credit are created equally. You will have more secure credit, more unsecured credit, so the mix of credit will kind of shift, but I wouldn't be surprised if credit accounts for a substantial part of the ARPAC expansion.

And it's one of the reasons why we are so excited with kind of digital banking models that are able to provide competitive solutions and resilient solutions for credit for -- at scale because that is really where kind of the book of the profit pool is not only in Latin America but across many other markets in the globe.

Congrats again.

So thank you, everyone. We now have approached 60 minutes of the call. So we are now concluding today's call. On behalf of Nu Holdings, our investor relations team, I want to thank you very much for your time and participation on Nu earnings call today. Over the coming days, we'll be following up with questions received tonight but we are not able to answer. And please do not hesitate to reach out to our team if you have any further questions. Thank you, and have a good night.

The Nu Holdings conference call has now concluded. Thank you for attending today's presentation. You may now disconnect.