Nubank Got a Real Bank in Mexico. The Milestone Isn't the Money—the Funding Engine Is

The headline sounds like a beginning: "Nu Holdings commences operations in Mexico." A retail reader could be forgiven for picturing a startup app just now setting foot in a new country. The truth is the opposite. Nu already has 15.8 million customers in Mexico, making it the country's third-largest credit-card issuer. What actually began on August 6, 2026 is something narrower and, for the economics, more important: Nu Mexico's right to operate as a licensed bank.

The distance between "already here" and "just now a bank" is the whole story. For a lender, the constraint that decides how much money it keeps is rarely customer sign-ups. It is the cost of the money it lends out. And for years, that is exactly the constraint Nu Mexico ran into.

The upgrade nobody watches

In 2019 Nu entered Mexico with a lighter regulatory status called a SOFIPO—a kind of savings-and-loan institution that can take deposits but is restricted in what it can offer. Regulatory categories are invisible to customers, but they shape the profit math underneath. A SOFIPO is allowed to gather only so much deposit money and offer only a limited menu of products. That matters, because the single biggest raw material a consumer lender buys is its funding: the deposits and borrowed money it turns into loans.

In July 2026 Mexico's banking regulator, the CNBV, granted the final "operations authorization," and after a technical migration on August 5, Nu Mexico formally began operating on August 6 as a full "multiple bank"—Mexico's largest digital bank, with its 15 million deposits brought under the government's IPAB deposit-insurance umbrella. The full license is the difference between a fintech that lends with constrained, expensive money and a bank that can compete for cheap, sticky deposits and issue products a SOFIPO never could.

That is the part analysts focus on. Becoming a bank gives Nu access to a lower-cost and more stable source of funding through deposits, and it lets Nu chase the market it has identified as the real prize: payroll accounts. Only about 36% of Mexican adults hold a payroll account, and roughly 90% of existing payroll accounts sit with just four incumbent banks. That concentration is the tollbooth Nu wants to drive through—a pool of low-cost, recurring deposits it has long been structurally barred from reaching.

Follow the money one layer down

Here is the mechanism that makes a banking license worth more than a press release. A consumer lender's gross profit is largely its net interest margin—what it earns lending minus what it pays for funding. Cheap deposits widen that margin directly. Nu Mexico's numbers show how far the funding engine has come and how much runway sits unused. As of June 30, 2026, Mexico held about $5.7 billion of Nu's deposits against a loan-to-deposit ratio of just 35%.

Read that ratio carefully. It means Nu Mexico is currently collecting more in deposits than it is putting to work in loans. Cheap money is in the door; the earnings lever is deploying it as the lending book grows. Management has even said Mexico deposits declined modestly in the quarter as part of a deliberate deposit-optimization strategy—trimming costlier funding to improve the mix while keeping liquidity. That is rent-discipline language: pricing the funding side of the ledger before scaling the lending side.

The S-curve supports the optimism. Mexico's revenue per customer is already ahead of Brazil's at the same stage of adoption—an average revenue per active customer of about $12.3, versus $5.6 for Brazil at a comparable point. Management has called Mexico the "next S-curve" and committed $4.2 billion of investment through 2030 to run it.

The theme is large. The earnings contribution is not—yet

Now the part that keeps this from being a simple buy-the-headline trade. A banking license is an enabler, not an earnings event. At the group level, Nu's Mexico operation still contributes a single-digit share of revenue—roughly 7% in 2025—and it has only recently clawed its way to around break-even after years of dragging on group margins. The consolidated numbers the market actually prices come mostly from Brazil, where Nu earned a record $1.1 billion in net income in the second quarter of 2026 on nearly $5.9 billion of revenue, with a group net interest margin of 22.9% and a return on equity of 33%.

In other words, the license is long-dated optionality on a margin story that will unfold over years, not a catalyst that shows up in next quarter's report. This is the purer version of a familiar trap: the theme ("Nu in Mexico") is real, but its contribution to consolidated earnings today is small and still being funded. A diversified winner can have credible exposure and low near-term purity at the same time.

There is also real competition stacking up at the same gate. Banco Plata, Revolut, Mercado Pago and Klar are all pushing into regulated banking in Mexico, and an Argentine group has applied for a multi-bank license. Payroll incumbency is a moat, but it is a contested one, and the same deposit-optimization squeeze that helps Nu's margins could be pressured if rivals bid up the cost of money.

What the clock says

None of this was a surprise when it happened. The banking license was approved back in April 2025, and Nu announced its $4.2 billion Mexico plan in February 2026—months before the August launch. In the stock market's terms, the event was long telegraphed, which is why the commencement itself is not the kind of news that reprices a company. Nu shares trade near $15, roughly 20 times trailing earnings and around five times book value, and are down about 10% on the year despite a strong recent quarter. The market has already paid for a lot of the Mexico story in advance.

That frames the judgment the reader has to make. The license is a genuine structural unlock—it removes a funding constraint and opens a payroll market that is both under-banked and concentrated in a handful of incumbents. But the payoff is measured in the multi-year shape of Mexico's contribution to group profit, not in any single announcement. A holder's time horizon is exactly what this story needs; a seeker of near-term catalysts should look elsewhere.

The confirming metric to watch is not headlines but the funding line: whether Mexico's deposit mix keeps improving, its loan book keeps growing into that 35% loan-to-deposit ratio, and the operation converts break-even into a steadily rising profit contribution. The signal that the hidden winner is becoming ordinary would be Nu forced to pay up for money again—either because the deposit-optimization stalls or because new entrants bid away the cheap funding that today is the quiet edge beneath the app.