Latin America didn't fix its broken banking system. It bypassed it. Pix processes 79.8 billion transactions annually — 54.7% of all retail payments in Brazil. Nubank has 131 million customers and just became Brazil's most valuable company, surpassing Petrobras at a market cap of $77–85 billion. In April 2026, the White House targeted Brazil over Pix, calling it a barrier to US payment companies. That complaint is the best possible evidence that something extraordinary has been built.
Data sourced from Banco Central do Brasil, Nubank Q4 2025 earnings, QED Investors Latin America 2026 Outlook, IMARC Group, Phoenix Strategy Group Fintech Investment Trends 2025, and MarkWide Research. All figures current as of June 2026.
Latin America entered the fintech era with a specific and powerful disadvantage that turned out to be an advantage in disguise. Its legacy banking system was expensive, exclusionary, and deeply entrenched. Brazilian banks charged among the highest interest rates in the world — consumer credit at 100%+ annually was not unusual. Account maintenance fees, transaction charges, and minimum balance requirements excluded the majority of the population from basic financial services. In Mexico, approximately 50 million adults had no bank account. In Colombia, financial inclusion hovered below 50% of the adult population as recently as 2015.
This was the precondition. A continent where the established financial system had failed most of its population was a continent where the established financial system had no loyal customers to protect. When a better alternative appeared — cheaper, faster, available on a phone, with no branch and no minimum balance — there was nothing holding the customer to the incumbent. The very failure of Latin American banking created the white space in which the most dynamic fintech ecosystem outside Asia could emerge.
The leapfrog pattern — documented in Letter 71 on South Africa's M-Pesa parallel — repeats here with striking precision. Africa skipped landlines for mobile phones. Africa skipped bank branches for mobile money. Latin America is now skipping the credit card for instant digital payments. Skipping the bank branch for the neobank. Skipping the credit bureau for AI-powered alternative credit scoring. The infrastructure that the developed world spent 50 years building is being bypassed in a decade.
Pix is the single most important fintech story of the 2020s — and it was not built by a startup. It was built by the Banco Central do Brasil, launched in November 2020, made available 24 hours a day, 7 days a week, 365 days a year, at zero cost to individuals and minimal cost to businesses. In a country where a bank transfer (TED) previously cost R$15–30 and took until the next business day, Pix arrived like a detonation.
The adoption curve was unlike anything the financial industry had seen. Within 12 months of launch, Pix had more registered keys than the entire Brazilian population. Within 24 months, it was processing more transactions than all credit and debit cards combined. By 2025, Pix processed 79.8 billion transactions — 54.7% of all retail payments — and the number continues to grow. The technology itself is not particularly novel. Instant payment rails exist in India (UPI), the UK (Faster Payments), Singapore (PayNow). What made Pix extraordinary was the mandate — every financial institution with more than 500,000 active accounts was required to participate. The network effect was engineered by regulation, not by market forces.
The geopolitical significance became explicit in April 2026 when the White House Trade Representative listed Pix in its report on foreign barriers to US commerce — arguing that it disadvantaged US payment companies like Visa and Mastercard. This complaint is the most powerful endorsement Pix could receive. The world's most powerful government, acting on behalf of the world's most profitable payment networks, identified a Brazilian central bank payment rail as a threat to their business model. What Pix threatened is the 1.5–3% interchange fee that Visa and Mastercard have extracted from every card transaction for decades. Pix charges zero. In Brazil, that extraction is now largely over.
Pix didn't just create a faster payment system. It created an open infrastructure layer on which every fintech could build for free — dissolving the moat that card networks had spent 60 years constructing and that Brazilian banks had used to extract rents from the entire economy.
Founded 2013 by David Vélez, Cristina Junqueira, and Edward Wible with a single product: a no-fee credit card controlled by an app. The entire banking industry in Brazil laughed. Eleven years later, Nubank has 131 million customers across Brazil, Mexico, and Colombia. In October 2025, it surpassed Petrobras to become Brazil's most valuable company at $77–85 billion market cap.
The structural advantage that built this is not technology. It is cost. Nubank's cost to serve a customer is $0.80 per month. A traditional Brazilian bank's equivalent cost is R$30–35 — roughly 20–25× more expensive. That cost differential is the margin that allowed Nubank to offer better products at lower prices while still generating $2.9 billion in net income. The neobank didn't win on features. It won on structure.
Nubank's expansion from credit card to full banking — investments, insurance, crypto, SME lending — follows the classic platform playbook. Get the customer with one product. Earn trust. Expand the wallet share. Today, Nubank generates more revenue per customer from financial services than many traditional banks — while spending a fraction of what those banks spend to deliver them.
Mercado Pago began as the payment layer for Mercado Libre — Latin America's dominant e-commerce marketplace. It has since become something far larger: an embedded finance platform with 50+ million active users, a credit book that rivals mid-size Brazilian banks, and in recent years more profitable than the marketplace business that spawned it.
Mercado Pago's competitive moat is unique: it underwrites credit using transactional data from the marketplace that no traditional bank can replicate without a partnership. A small seller on Mercado Libre has years of transaction history, customer ratings, return rates, and seasonal patterns — all of which are far more predictive of creditworthiness than a credit bureau score. The result is credit access for millions of SME sellers who traditional banks could not serve because they had no credit history — only Mercado Libre did.
Mexico's 50 million unbanked adults and 4 million informal small businesses needed a different solution than Nubank's credit card. Clip's answer: a POS terminal that becomes a financial services distribution channel. Start with card acceptance. Layer in working capital loans based on transaction history. Add business accounts, payroll, and inventory management. The terminal is the beachhead. The financial services are the business.
Clip turned the problem of cash-dependent informal commerce into a data-generating asset. Every transaction through a Clip terminal is a data point that improves credit underwriting for the next loan. Mexico's Fintech Law, enacted 2018, provided the regulatory framework that allowed Clip and its peers to scale. Mexico now has over 700 active fintech companies — the second-largest ecosystem in Latin America after Brazil.
Three stories that together represent the next wave. Rappi Pay — the financial layer of Colombia's super-app — demonstrates embedded finance at scale, offering digital accounts and credit to gig workers and informal economy participants previously excluded from formal financial systems. Félix Pago is transforming the $160 billion annual US-to-Latin America remittance corridor with real-time WhatsApp-based transfers that cost a fraction of Western Union. And stablecoins — USDC and USD-denominated digital assets — are becoming the de facto savings instrument for Argentinians and Venezuelans trying to preserve purchasing power against hyperinflation.
The stablecoin story is particularly significant. In a region where several currencies have lost 50%+ of their value in a decade, a digital dollar accessible via a smartphone is not a crypto speculation — it is a survival tool. Nubank and Mercado Pago both now offer crypto and stablecoin access directly in their apps, blurring the line between fintech and digital assets in ways that regulators in developed markets are still trying to understand.
The first wave of Latin American fintech was about infrastructure — building the payment rails, the digital accounts, the open banking APIs that made financial inclusion possible at scale. The second wave, which is already well underway, is about intelligence — using AI to do what the first wave's infrastructure makes possible: underwrite credit for people who traditional banks cannot serve because they have no credit history, only behavioural data.
In Colombia, 66% of fintechs already use AI, and another 27% plan to adopt it within a year. Companies leveraging AI report 44% lower costs, 56% faster processing, and 50% better customer service response times. Nubank uses AI-driven algorithms to assess credit risk in real-time, and reports that its AI models have reduced default rates by 12% over two years. Mercado Pago underwrites credit using marketplace transaction data. Clip underwrites SME loans using POS terminal data. These are not conventional credit scores. They are behavioural models that can assess the creditworthiness of a street vendor in Mexico City or a gig worker in Bogotá using data points that no traditional bank's underwriting model was ever designed to process.
The broader implication is significant. AI credit scoring is the mechanism through which the 50 million Mexican adults without bank accounts — and the hundreds of millions across Latin America in similar situations — can access credit for the first time. Not because someone changed the rules. Because someone built a better model. The combination of Pix-style payment rails (generating transaction data), open banking (sharing that data across institutions), and AI (extracting creditworthiness signals from that data) is the complete architecture of the financial inclusion revolution. Brazil has all three. Mexico and Colombia are building the third layer on top of the first two.
The geopolitical risk is real and has already materialised. The White House's April 2026 targeting of Pix is the opening salvo of a broader battle. US payment networks — Visa, Mastercard, American Express — generate significant revenue from Latin American transaction flows. Pix has already compressed interchange fee income in Brazil. If Pix-style systems spread across the region, the revenue impact on US payment networks is substantial. US trade policy pressure on LatAm fintech regulation is not a hypothetical risk. It has begun.
Currency volatility remains the structural constraint on regional fintech ambition. The Argentine peso crisis, periodic Colombian peso depreciation, and the chronic Brazilian real weakness all create friction for cross-border financial products. A fintech that works brilliantly within a single country faces significant complexity when it tries to operate across six currencies with different inflation rates, different regulatory frameworks, and different exchange rate regimes. The regional integration story — one LatAm financial ecosystem — is still more aspiration than reality.
Political risk is unevenly distributed but consistently present. Colombia's fiscal situation, Brazil's electoral cycle, Mexico's OXXO-Pay regulatory scrutiny — political environments shift, and financial regulation follows political environments. The fintechs that have built durable businesses have done so partly by operating across multiple countries, so that political risk in one market is offset by stability in another. Single-country dependence remains a structural vulnerability in a region where political conditions change faster than business models can adapt.
Long-horizon thinking on capital, technology, and the forces shaping the next decade of wealth creation. Written from first principles. Not consensus. Not noise.