Santander Brasil's 'New Cycle' Is the Same De-Risking Story, Four Years In

A new CEO promises continuity without inertia — but the quarter's real news is a write-off calendar that distorts NPLs into 2027, and an analyst question that simply went unanswered.

SANB11.SA · Earnings Call · 2026-10-28

The new cycle that isn't

Gilson Finkelsztain's first earnings call as CEO opened with a line built for the occasion: “There is continuity in our key pillars, but continuity does not mean inertia.” — Gilson Finkelsztain, CEO · 2026-10-28 The quarter itself was unremarkable: BRL 2.9bn of recurring net income, an 11.9% ROE, expenses growing well below inflation, an efficiency ratio of 38.8%. That ROE sits a long way from the group's stated ambition of 20%-plus, and everyone on the call knew it. What made the framing interesting is how little of it was actually new. Everything Gilson then described — higher-quality clients, more secured loans, a deliberate tilt toward high income clients and large corporates, and a willingness to shrink the low-income mass market — has been the operating script since at least 2022. The company's own keyword history bears that out: Client NII, Consumer Finance, Cost of risk and Mass Retail recur quarter after quarter; the genuinely fresh name on the board this period is the CEO's own. Back in early 2023, then-CFO Angel Santodomingo described the identical trade-off with almost identical words: “we have moved the balance sheet not only to a more solid one, but also to a more collateralized one, which also means, obviously, that has lower cost of risk and also means having lower cost of risk that the spread is lower.” — Angel Santodomingo, Executive (likely CFO or senior management) · 2023-02-02 Four years of the same trade, still waiting for the offset.

What the write-off calendar is hiding

The single most revealing exchange of the day was not about profitability at all. It was about accounting plumbing. Day NPL indicators — the short-dated ones — improved for a second consecutive quarter, yet the over-90-day ratio rose for individuals and SMEs, and the two have been diverging for three straight reports. The explanation is now a schedule rather than a number, and it is worth spelling out: the bank has moved to a uniform 360-day write-off for most of the portfolio, extended consumer-finance terms from 18 to 21 months for secured goods, and is writing off clean personal, card and revolving lines in eight to nine months. Camila Toledo's summary of the consequence was the most candid sentence on the call: “This tends to normalize throughout 2027. That is why we are focusing on talking about short-term NPL, because they do not have this effect.” — Camila Toledo, Investor Relations · 2026-10-28 In other words, the headline delinquency gauge will be a moving target for another five quarters, and management is asking investors to look at a different one. That is a legitimate request and also a convenient one. It is also the third consecutive quarter this topic has surfaced — in July, an analyst pressed on the same 4,966 model review and got "work in progress" in reply; now there is a timetable, but the moving parts have grown. Above that sits agri business, ranked the company's top theme this quarter and unique to it among reporters. Without the CPR-linked farm exposure, over-90 NPLs for individuals would have been below 5%; the agribusiness segment alone was worth roughly 60 basis points. Management calls it a one-off and simultaneously warns it could worsen if high interest rates persist.

The provisioning question, again

The wholesale surprises have become a quarterly feature rather than an exception: BRL 500m in this quarter on top of a prior quarter already adjusted for similar cases. Carlos Muñiz, whose Portuguese has improved markedly since his first call in July, declined to close the door.

I cannot tell you whether we will have any additional names, especially in the wholesale universe, that could impact our numbers in the fourth quarter. But we talked about NPL indicators, and now we see that the behavior is much better than we had in the past in these portfolios.
Carlos Muñiz, Executive (likely CFO or similar) · 2026-10-28

The read-through is that the bank is running a de-risking programme whose losses arrive with a lag, and each quarter brings a fresh reminder of the old book. The new cohorts do look better — management is credible on that — but the old ones keep sending invoices. That is precisely why Daniel Vaz's July question about when risk-adjusted NII inflects got the answer “maybe the state would be closer to 2027” — Carlos Muniz, Chief Financial Officer (CFO) · 2026-07-29 — a full year later, and now the language has shifted only to "gradual and sustainable recovery."

Client NII and the hedged hole

Client NII fell again, with a 33-basis-point drop in the client spread: roughly 30bps from the consumer-finance FIDC sale and lower CDI, and the balance from deferred banking-correspondent expenses. Gilson's rebuttal is that this is by design — “since the end of 2024, we have been saying that we are reducing our sensitivity to interest rates.” — Gilson Finkelsztain, CEO · 2026-10-28 True, but the phrase "sensitivity to interest rates" was already a headline keyword in the 2024 second quarter, and the hedged legacy book nonetheless still delivers a negative market NII result in 2027, just a smaller one. The hedge protected the capital; it did not create earnings. Worth noting the global weather. Cost of deposits is among the market's hottest banking keywords this quarter, echoed by regional lenders from Customers Bancorp to HomeTrust to Germany's Deutsche Bank. Santander Brasil is on the other side of that coin: its squeeze is mix-driven, not funding-driven, because it is deliberately lending to borrowers who pay roughly half the spread. Management is explicit that transactional credit — cash, receivables, acquiring — is the intended substitute for clean credit in the B1/B2 segment.

The question that went unanswered

The opening analyst question of the session was simple and specific: the bank's physical footprint and headcount are down roughly 12% year on year — is that process finished? “Back to Portuguese. Question from Daniel Vaz with Safra.” — Camila Toledo, Investor Relations · 2026-10-28 The question was never answered. In July, an almost identical query produced a long, thoughtful Carlos Muñiz answer on cost-to-serve and branch formats. Three months later it was dropped. Similarly, the capital question has been quietly reframed: in July, Carlos reaffirmed “we'll maintain our payout policy of 50%” — Carlos Muniz, Chief Financial Officer (CFO) · 2026-07-29; this quarter, when asked where the capital efficiency levers are, the answer was entirely about asset rotation and market share, with no CET1 ratio and no payout figure offered. What management stops talking about matters. The bear case writes itself: a bank at 11.9% ROE whose revenue line is being deliberately compressed, whose provisioning is subject to one-off wholesale events it won't forecast, and whose headline NPL will be distorted by accounting for another year. The bull case is equally coherent: the new cohorts are clean, the cost line is genuinely controlled, and Gilson's own words — “our teams are being more demanding than we were in the past, the conditions to accept and the conditions that we require to renegotiate” — Gilson Finkelsztain, CEO · 2026-10-28 — describe a credit culture that will look better in hindsight than in real time. The missing evidence is a price signal: no usable tape for this name in the period, so the market has not yet voted. Until it does, this is a call about discipline versus duration — and duration keeps getting extended.