The Credit FlywheelNarrow moat

StoneCo (STNE) — moat facet

Data to credit to a deeper relationship to more data — the loop that compounds the franchise.

Put the pieces together and you get a flywheel that is the best argument for Stone's moat widening over time. It runs like this: the payment relationship generates data on the merchant's sales; the data allows well-underwritten, receivables-secured credit; the credit deepens the relationship and raises switching costs; the deeper relationship brings more of the merchant's volume onto Stone's rails, which generates still more data. Each turn makes the next loan safer to make and the customer harder to lose.

Cost of risk (%)20.2%Q2 202521.9%Q1 202621.5%Q2 2026StoneCo Q2 2026 earnings release
The flywheel runs at about a fifth of the book in annual losses, and the rate is edging up.

The power of the loop is that it compounds advantages a standalone competitor cannot easily assemble. A pure lender has the credit appetite but not the data or the collateral; a pure acquirer has the data but not the lending machine; a bank has capital but not the relationship with, or visibility into, the small merchant. Stone is trying to hold all the pieces at once, so that each reinforces the others — and where it works, the merchant ends up embedded in a system that knows him, funds him, and grows with him.

The flywheel's honesty check is that it spins both ways. Turned carelessly, the same loop amplifies mistakes: bad underwriting doesn't just lose money, it poisons the data and the relationship, as 2021 showed. And every part of the loop depends on Stone keeping the payment flow that starts it — lose the merchant's transactions to Pix or a rival, and the whole wheel loses momentum. So this is a genuine, if fragile, engine of durability: real, self-reinforcing, and capable of turning a narrow payments moat into something sturdier, but only in the hands of a company disciplined enough to keep it turning in the right direction — credit revenue nearly tripled in the past year, with the cost of risk along for the ride1.

Moat trajectory: Widening

Widening, carefully. The loop — payments generate data, data enables secured credit, credit deepens the relationship and raises switching costs, the deeper relationship brings more volume and more data — is turning again after 2021, and each turn makes the next loan safer and the customer harder to lose. It compounds advantages a standalone lender, acquirer, or bank cannot each assemble alone. The honesty check is that the wheel spins both ways: careless underwriting poisons the data and the relationship, and the whole loop depends on keeping the payment flow Pix threatens. But turned with discipline, this is the engine most capable of widening Stone's narrow payments moat into something sturdier — and right now it is turning in the right direction.

The number that tests this moat
Reported
Cost of risk
21.5% in Q2 2026, from 20.2%

The flywheel works only if losses stay priced in; cost of risk rising with the book is the governor slipping.

Source: StoneCo second-quarter 2026 earnings release (13 August 2026) ↗
⚠ Threats to the moat
References
  1. ReportedCredit revenue nearly tripled; cost of risk 21.9%.
    StoneCo Q1 2026 earnings release — revenue R$3.58B (+6.5% YoY), adjusted net income R$549M (+3.5%), TPV R$137B (+3%), credit portfolio R$3.22B with cost of risk 21.9%, gross margin 41.6%, retail deposits R$10.1B (+22%); special dividend ~$2.53/share (~R$3.08B) after the Linx sale — Q1 2026 · publ. May 2026 · source ↗
Sources
Generated September 23, 2026