While American investors obsess over AI stocks and oil prices, a Brazilian fintech is quietly rewriting the rules of banking across an entire continent. Nu Holdings (NU) — known simply as “Nu” to its 127 million customers — has built what might be the most efficient financial institution on the planet.
Here’s the number that should make every bank executive lose sleep: it costs Nu roughly $0.90 per month to service a customer. Traditional Brazilian banks? Between $12 and $15. That’s not a marginal improvement — it’s a 93% cost advantage that lets Nu profitably serve millions of people the big banks wouldn’t touch.
The origin story reads like a Silicon Valley fable transplanted to São Paulo. In 2012, Colombian-born David Velez moved to Brazil to run Sequoia Capital’s Latin American office. Opening a local bank account took months. When Sequoia pulled out of Brazil, Velez saw what everyone else had missed: a banking oligopoly so lazy and overpriced that it was practically begging to be disrupted.
He assembled a founding team of three — himself as strategist, Cristina Junqueira (a banker who knew the system’s weaknesses), and Edward Wible (an American engineer who built the entire tech stack from scratch). That last part turned out to be Nu’s secret weapon. While incumbents run on decades-old software that makes innovation agonizingly slow, Nu was built on modern architecture from day one.
Nu’s first product was brilliantly simple: a no-fee credit card. In a market where banking was synonymous with abuse and hidden fees, word spread like wildfire. The company’s Net Promoter Score soared into the 90s — on a scale where anything above 50 is considered excellent. Nu didn’t just acquire customers; it started a movement.
The skeptics said lending to Brazil’s underbanked population was financial suicide. Nu’s answer was the “low and grow” model: issue micro-limits as small as $10, watch real-time payment behavior and smartphone data, then gradually increase credit as borrowers proved themselves. The result? A record 31% return on equity by late 2025 — in an industry where 12% is considered good. Nu is lending to people the big banks rejected and making nearly triple the industry’s average return doing it.
Now comes the expansion play. Mexico and Colombia are next, and the opportunity there is arguably even larger. Mexico remains heavily cash-based, with banking penetration far below Brazil’s. If Nu can replicate even a fraction of its Brazilian playbook, the addressable market is enormous.
For investors watching the usual mega-cap names trade sideways, Nu represents something increasingly rare in 2026: a genuine growth story backed by unit economics that actually work. In a world where most fintech companies burn cash chasing scale, Nu is profitable, growing, and sitting on a competitive moat made of code and customer obsession. The big banks still haven’t figured out how to fight back.