The MoatNarrow moat

StoneCo is, at bottom, a company that sells small Brazilian merchants the plumbing of modern commerce — the card machine on the counter, the bank account behind it, and increasingly the working-capital loan that gets a shopkeeper through a slow month. It is a fintech, and the fashionable thing to say about fintechs is that they have no moat at all: payments look like a commodity, competitors are legion, and a central bank in Brasília has built a free instant-payment system that threatens the whole card economy. There is truth in every one of those worries, and I will come to them. But there is also a real, if modest, business here — one earning a twenty-six percent return on equity while the market prices it as though it were quietly going out of business.

Where the moat exists, it is built on distribution and service rather than technology. Stone's distinctive weapon is a network of local operating hubs spread across roughly a thousand Brazilian cities — each combining sales, logistics, and support in one place — that lets it reach and serve the small merchant the big banks find too costly to bother with. A shopkeeper in the interior who has a problem with his card machine can get a human being from Stone to fix it quickly, and small business owners, it turns out, will pay for and stay for that kind of service. That is not a glamorous advantage, but it is a genuine one, and it is expensive and slow for a rival to copy city by city.

On top of that distribution, Stone has been assembling something more valuable than a payments processor: a full financial operating system for the small business. The card machine brings the merchant in; then Stone sells him a banking account — there are now some 3.7 million active banking clients¹ and about R$11 billion of deposits — and then, cautiously, a loan secured against the card receivables it can already see flowing through its own rails. Each product makes the next one easier to sell and harder to leave, and the deposits provide cheap funding for the credit. It is the same deposit-and-lend flywheel that has made banks durable for centuries, aimed at a customer the incumbents underserved.

The credit piece deserves special mention, because Stone got it badly wrong once. In 2021 it expanded lending too fast into a broken receivables-registry system, took heavy losses, and had to halt the business — a genuine and humbling mistake. What matters is what came after: a disciplined rebuild, and by the end of 2025 a credit book back to about R$2.8 billion, growing again² but secured against the merchant's own future card sales, which Stone is uniquely positioned to see and to claim. Underwriting off a customer's actual cash flow is a real edge, provided the company remembers the lesson it paid for.

Now the honest catalogue of dangers, because this is not a fortress. Pix — the central bank's instant-payment system — is a marvel³ for consumers and a structural headwind for anyone earning a fee on card transactions, and payment growth among Stone's small merchants has already slowed to the low single digits. Competition is ferocious: the big banks, Cielo, Rede, GetNet, and fellow fintechs like PagBank and Mercado Pago all press on take rates. And the whole model is unusually sensitive to Brazil's interest rate, the Selic, which sets both Stone's funding costs and the macro weather its customers trade in.

Set it all on the scale and you have a narrow moat, not a wide one — a well-run, cash-generative franchise with a defensible niche among small merchants, wrapped in a genuinely hostile competitive and regulatory environment, and priced by a nervous market at around four to five times earnings. The interesting question about StoneCo is not whether the moat is wide (it isn't) but whether it is wider, and more durable, than a low-single-digit multiple assumes. The number that answers it is the banking side: clients and deposits. Payments will keep compressing — that is settled. But while banking clients and the deposits they leave keep compounding, Stone is becoming the operating system the multiple refuses to price; the quarter that growth stalls, the melting-ice-cube camp was right.