The purple card at 100 million customers: Nubank and Brazil's fintech decade

David Vélez, a Colombian engineer working in Brazilian venture capital, used to tell the founding story with a prop: the antiseptic bulletproof booths through which Brazilian bank branches once served customers, and the fee schedule that charged them for the privilege. In 2013 he and his co-founders — Cristina Junqueira, a credit executive from a traditional bank, and American engineer Edward Wible — launched Nubank with a no-fee purple credit card managed entirely from a phone. The company has since passed 100 million customers across Brazil, Mexico and Colombia, listed on the New York Stock Exchange in one of the era's largest fintech IPOs, and forced every incumbent in the region to rebuild (company disclosures).

Why Brazil was the perfect launchpad

The opportunity was structural. Brazil's banking system was among the world's most concentrated and most expensive: five banks held the overwhelming majority of assets, spreads were wide enough to fund branch cathedrals, and tens of millions of working adults sat outside credit altogether. Regulation was the quiet enabler — Brazil's central bank spent the 2010s modernizing the rails (interoperability, open-finance rules, licensing pathways) precisely so challengers could plug in. Nubank attacked the most abused product, the credit card, with underwriting built on data rather than pay slips, and grew through the channel incumbents ignored: the smartphone of a customer the branches never wanted.

What the incumbents did next

Competition arrived faster than the incumbents' cost structures could handle. Every major bank launched a digital brand; digital-only banks multiplied into the dozens; and the central bank's own infrastructure — Pix, the instant-payment system launched in 2020, now processing transactions in the hundreds of millions daily — made money movement free and immediate for every player (BCB payment statistics). The result is a two-sided verdict: Brazilian finance is now among the cheapest and most real-time systems in the world for consumers, while bank spreads, fee income and branch counts have compressed painfully for the providers. The customer won the decade; the industry is still digesting it.

The economics under the app

Nubank's model matured from a card company into a full banking relationship: deposits funding credit, interchange and interest income replacing the fee stack it abolished, and a cost-to-serve a fraction of a branch network's. The frontier questions are the classic ones of scaled lending — credit quality through the cycle, funding costs as the account base matures, and the temptation of becoming the very institution it was founded against. Brazil's macro cycle stress-tests the model annually: when the central bank lifted rates to multi-year highs to fight inflation, fintech lending books felt it faster than incumbents', and the sector consolidated accordingly (BCB financial-stability reports).

What it exported

Two things. First, the template: no-fee, app-first, data-underwritten banking is now the default pitch from São Paulo to Mexico City to Bogotá — Nubank's own expansion into Mexico and Colombia is the region's most-watched case of homegrown LatAm scale, a direction the older generation of startups rarely attempted. Second, the talent diaspora: fintech executives trained at Nubank and its peers seeded the next cohort of companies, making São Paulo the region's financial-technology labor market the way Buenos Aires became its software one.

The sober assessment

A decade in, the honest scorecard: financial inclusion measurably widened — central-bank data show the share of adults with accounts rising sharply through the fintech years — consumer costs fell, and the region produced a genuinely world-class financial-technology company. Bank profitability, cyber-fraud management and credit-education gaps remain the unfinished files, and the next decade's test is whether the upstarts keep their cost discipline at 100-million-customer scale. The bulletproof booths, mercifully, are mostly gone; the purple card is what replaced them.

How the incumbents answered

The competitive response arrived in two waves. First came the digital brands: every major Brazilian bank launched or acquired a mobile-only proposition — stripped of the branch cost structure the challengers attacked, priced with the fee-free accounts the market now demanded — and the old guard's scale advantages (funding costs, regulatory capital, brand trust with older depositors) made the copies commercially durable. The second wave was institutional: the incumbents adopted the challengers' infrastructure, riding the central bank's Pix and open-finance rails with engineering budgets the fintechs could not match, and quietly re-priced their legacy account bases before the fintechs could migrate them. The result is a market that looks less like disruption than acceleration: the consumer outcome is nearly universal free, instant, app-based banking, and the corporate outcome is a handful of scaled digital franchises — some new, some rebuilt — competing on credit quality and product depth rather than distribution.

The competition's next frontier is credit, where the stakes concentrate. Pix solved payments, but Brazil's real arbitrage was always the spread between what money costs and what borrowing charges; the digital banks' underwriting data — transaction histories the incumbents never saw — is the asset that decides who captures it through the rate cycle. The 2022-2025 tightening was the first full-cycle test: delinquency spiked at the young portfolios, the survivors tightened and re-segmented, and the crisis-era question — does fintech credit behave differently, or just arrive faster? — received its first empirical answers. The easing cycle now beginning will write the second half.

For how interest rates are shaping the credit cycle behind fintech's balance sheets, read our report on the Copom's first cut, and explore the Brazil section.