StoneCo Deletes 2027 — and Points at a 250-Basis-Point Yield Curve
A quarter of 'gradual progress' produced a hard 2026 cut, withdrawn long-term targets, and a cost-of-risk promise that has now slipped three calls in a row.
STNE · Earnings Call · 2026-11-12
The Map Changed, Not Just the Year
The operating quarter was unremarkable-bad. What happened on the call was a different kind of event: StoneCo removed the destination, not merely the year. Adjusted gross profit guidance for 2026 went from BRL 6.6-7.0 billion to BRL 6.1-6.2 billion; adjusted basic EPS from BRL 10.8-11.4 to BRL 9.1-9.5. And then, more tellingly, the 2027 targets — reaffirmed in March and defended again in August — were withdrawn outright. That makes the trajectory of this guidance the story. Back in March, the plan was built on an explicit rate path: “we're assuming Selic at low 12% by the end of 2026 and high 11% for 2027. That's what's behind the guidance number.” — Diego Salgado, CFO · 2026-03-03 Even in August, with the Selic already running closer to 14%, the line held: “None of this changes our guidance ranges, as we've mentioned before” — Diego Salgado, CFO · 2026-08-13 — despite rates being a headwind of north of BRL 300 million for the year. Two quarters of holding the line, then a withdrawal. Mateus Scherer was unusually direct about the cause, and it is a macro variable dressed as a company decision:Do the arithmetic and roughly BRL 500-625 million of pretax earnings swung on the yield curve alone — more than the entire downward revision to 2026 gross profit. The big variable is not churn, not credit mix, not competition. It is the price of money in Brazil after an election that market participants read as fiscally worse. The tape agrees, and not just for Stone. In the global price history, the worst thirty-day cluster of decliners is Brazilian financial exposure: interest on equity down roughly 40% with CIG, BBD and RENT3.SA flagged, alongside a high income segment and private payroll loans bid down across BBD, ITUB and NU. Stone's own stock did the round-trip first: up about 25% after the first electoral round, then down about 30% when the runoff went the other way. This is a country-level repricing wearing a company ticker.the volatility is still very high, and the visibility is still very limited on that front... Post-election, I think we've seen a move in the yield curve of about 250 basis points. As you all know, we have a sensitivity to interest rates of about BRL 200 million-BRL 250 million for every 100 basis points.
The Cost of Risk Promise That Keeps Moving
Here the tension is internal, and it has been building since May. The mid-to-high teens cost of risk has been the fixed star for three consecutive calls. In August it was stated as near-certainty:By November the same CFO, Diego Salgado, has moved the horizon and reopened the model: “this trend towards lower levels will probably take a little bit longer. We are in the process of reassessing our credit models for a materially tougher macro backdrop.” — Diego Salgado, CFO · 2026-11-12 Credit model revision, not portfolio seasoning, is now the swing factor — and it flows straight through the forward-looking provision component. The underlying book still looks respectable on the surface. The portfolio reached BRL 4.1 billion, up 78% year-over-year and 9% sequentially; cost of risk eased to 20.1% from 21.5%; provisions were BRL 197 million; coverage slipped to 196% from above 200% because government-guaranteed lending, which nearly doubled quarter-over-quarter, needs less provisioning against it. That mix shift is deliberate — government programs carry guarantees that let Stone keep serving clients it would otherwise have to decline — but it also means coverage ratios and NPL optics get harder to read. The genuine stress point is the dedicated desk, where judicial recovery filings in Brazil are at record levels and larger tickets have burned the company twice. Management has capped maximum ticket size and redirected growth. The consolation is arithmetic: share of wallet in credit is still mid-single digits, and the newer credit cohorts are performing better even excluding the guaranteed lines.the guidance still stands that cost of risk will trend down to that mid- to high teens in the medium term, probably ending the end of the year already at the high teens level.