Nu holdings will continue to disrupt the South American banking market

Two years ago i bought Nu at $11. Now there were two main things which gave me conviction.

1. The product was genuinely better than anything the incumbent banks were offering
2. Warren Buffett's Berkshire Hathaway had quietly taken a large position.

Now of course following a person's investment is always the right call. People have different amounts of capital, entry positions and also risk aversion. However i believed that he saw something in the uni economic which the market hadn't. My initial goal was to understand what he saw.

Upon initial research i made two observations. Product quality was top-notch and the cost was almost zero. This sent me down a research rabbit hole that convinced me this was a generational disruption story.

Latin American banking has historically been one of the most extractive industries on Earth. This is because Brazil's 5 largest banks have operated as a comfortable oligopoly for decades on end. Moreover they charge some of the highest fees in the world. Not only that but they offered poor service and faced almost zero new competition. Something you might find fascinating is that annual credit card interest rates in Brazil often exceed 300%. Additionally for the average person to open a bank account you would need extensive documents, in-branch visits and waiting time.

All of this was a systematic issue in Brazil. For most Latin American's especially low-income or rural this was an issue. Not only was the formal banking system unfriendly but it was very inaccessible. The funny part? This wasn't a market niche. It was 600 million people.

Nu Holdings was founded in 2013 with a deceptively simple premise. It was build a bank people actually like, charge nothing to get started, and let the product do the selling. The first product was a no-annual-fee credit card managed entirely via a phone app. There were no branches, no physical infrastructure, and no legacy IT systems to slow things down. Customers joined through referrals. The waiting list became a status symbol.

Now the part which Berkshire probably saw. Nu's main competitive advantage wasn't just the low cost structure or its marketing, but the product itself. Nu consistently ranks as Brazil's most loved financial brand. Its NPS (Net Promoter Score) is among the highest of any financial institution globally. Having personally worked at a bank i was best placed to know that a metric like that matters enormously in a business built on referrals. In a sector where the incumbents generated loyalty through captivity rather than satisfaction, Nu generated it through genuine preference. That is an extraordinarily difficult thing to replicate.

By the end of 2024, Nu served over 114 million customers across Brazil, Mexico, and Colombia, adding 20.4 million net new customers during the year alone. They have become the third biggest FI in Brazil by customer count. The cool part? They got there without a single physical branch.

The 2024 results answered the one lingering question sceptics had. Can this business model actually make money?

Revenue reached $11.5 billion for the full year, up 58% on an FX-neutral basis, while net income nearly doubled to close to $2.0 billion. Return on equity hit 30% while Nu was still heavily investing in Mexico and Colombia, both of which are pre-profitability markets. This value is comparable to some of the biggest banks in the world. Additionally, the cost structure is the most important number in the entire model. Nu's average monthly cost to serve each active customer is $0.90. A traditional Brazilian bank spends roughly 15–25x that amount. That gap doesn't close. Legacy banks cannot shed their branch networks, their headcount, or their decades-old IT infrastructure. Nu was built without any of it.

Now looking in the future i see mainly positives. Nu has 92 million Brazilian customers but remains underpenetrated on revenue per customer. The journey from $10.70 ARPAC toward the $25+ seen in mature cohorts is already underway. Moreover this is a function of time and trust, not new customer acquisition. As Brazilians age with Nu and turn to it for mortgages, payroll, and wealth products, that revenue compounds without meaningful incremental cost. Furthermore, Mexico surpassed 10 million customers in 2024. It is three to five years behind Brazil on the adoption curve, which means investors today can see the destination while still being early in the journey. Mexico's banking penetration is even lower than Brazil's with less than 40% of Mexicans hold a formal bank account. The prize is enormous, and Nu's early metrics in Mexico are, by management's own account, tracking ahead of where Brazil was at a comparable stage.

Now in today's day and age it is a given that for any company to succeed they must be innovating despite the foundational strength of the business model. New products like NuCel (mobile telephony) and NuTravel hint at a longer-term ambition: to become the financial operating system for Latin American daily life. This is speculative but directionally important. The customer relationships Nu is building today are precisely the foundation that platform businesses are built on.

Rather than relying on a discounted cash flow model (which distorts for digital banks due to how lending portfolio growth flows through cash), I anchor on earnings power.

The picture is clearer now than it was when I bought at $11. Through three quarters of 2025, Nu has generated approximately $11B in revenue on an annualised run-rate, with Q3 alone delivering $783M in net income. This implies full-year 2025 net income of roughly $2.8–3.0B. Analyst consensus projects full-year 2025 revenue of ~$14.7B.

Looking to 2026, the consensus estimates $0.92 EPS which is a ~77% jump from the 2024 EPS of $0.52 on continued operating leverage as Brazil matures and Mexico approaches profitability. At 20–22x forward 2026 earnings, that implies a fair value range of $18-20, broadly in line with where the stock trades today (~$17–18).

The market has largely caught up to the 2025 thesis. The remaining upside is a 2026–2027 story: if Mexico's profitability inflection arrives ahead of schedule, which early signs suggest it might and ARPAC continues its cohort-driven climb toward $25+, a $22–25 price target over a 2-year horizon is credible. Downside is cushioned by the quality of the franchise and I'd expect the stock to find support around $13–14 on any meaningful selloff.

Finally the most important headwind in my opinion. The Central Bank of Brazil (Banco Central do Brasil) has watched Nu's rise with a mixture of admiration and caution. In late 2025, it issued new requirements that will compel Nubank to obtain a full banking license in Brazil by 2026. This is a significant operational and capital requirement that incumbents have long held.

This matters for several reasons. A full banking license subjects Nu to higher capital adequacy requirements, more intensive supervision, and potentially stricter limits on lending and deposit-taking activity. It also means compliance costs will rise structurally.

My view is that this is a manageable but genuine headwind. Nu has already announced its intention to obtain the banking license, framing it as a natural evolution rather than a burden. The company is well-capitalised as it held $2.4 billion in excess cash at the holding company level throughout 2024. Moreover the transition, while costly, doesn't threaten the core business model.

MY FINAL TAKE

I bought Nu because I believed a best-in-class digital product, backed by Berkshire's validation of the unit economics, was being priced as a risky emerging market bet rather than the durable compounder it was becoming. Two years later, the profitability inflection has happened, Mexico is working, and the customer flywheel is intact. The regulatory overhang is real and worth watching, but it doesn't break the thesis. Instead it just raises the bar for execution. I'm holding because the story is playing out, and the Mexico chapter is still early.