Brazil Drags the Shares to 1.7x Tangible Book, the Level Where We Said We Would Upgrade

Results vs. Consensus

Santander published nine-month results before the Madrid open on 28 October and presented them on a call at 10:00 Madrid time. This was the first quarter to carry a full three months of TSB and the first to consolidate Webster, which closed on 20 August and contributed forty days. Poland is absent from every line of every period. The figures below are the company's underlying series unless marked otherwise, on the presentation basis that took effect in the first quarter.

Q3 2026 Scorecard

The consensus column is the company-compiled poll of fourteen analysts, published on 19 October, line by line on the company's own underlying basis; the non-recurring line drew estimates from eight of them. Two independent readings of the Street sat either side of it. One wire poll had underlying profit at €3.75bn, net interest income at €12.0bn and provisions at €3.44bn. A broker comparison implied roughly €3.82bn of profit and €12.02bn of net interest income. Depending on the poll, the profit beat is between 0.7% and 2.6%, net interest income is in line to 1% light, and provisions are between 1.3% and 2.6% heavier than expected. The pre-tax miss is only visible in the company's line-by-line set, which is the one we use.

Q3 2026 vs. Q3 2025 (underlying, EUR m)

Santander publishes the third-quarter standalone income statement against the second quarter only. The Q3 2025 column is the nine-month 2025 figure in this report less the first-half 2025 figure in the H1 2026 report, both on the restated Poland-excluded basis; each derived change reconciles to the company's published year-on-year rate. Constant-euro rates are the company's and are not published for the revenue sub-lines. The 2025 non-recurring line is Poland's own earnings, then held for sale.

The year-on-year picture is flattering for an obvious reason: TSB and Webster are in the 2026 column and not in the 2025 one. Management put the acquisitions' effect at 1 to 3 points across the main lines for the nine months, and for the third quarter alone, with a full TSB quarter and forty Webster days, it is larger. That is why the sequential comparison below, which the company also publishes excluding the acquisitions, is the more informative one this quarter.

Q3 2026 vs. Q2 2026 (underlying, EUR m)

The right-hand column is the one to read. Excluding the acquisitions, revenue fell, both of the customer lines fell, provisions rose and pre-tax profit fell 1.1%. What kept underlying profit to a 1.4% decline rather than something worse was a 3.6% sequential fall in costs, which is the transformation doing its job, and a lower minorities line. Management attributed the revenue softness to seasonality: the third quarter is structurally weaker for corporate and investment banking fees and markets income, for Brazilian day counts and for Chile's inflation-linked unit of account. Some of that is true every year. It does not change the observation that the 2% headline growth was purchased.

Nine Months 2026 vs. Nine Months 2025 (underlying, EUR m)

On the nine-month view the business looks as good as it has at any point in our coverage: revenue up 4.5% excluding acquisitions against a mid-single-digit target, costs down 2.0% against a target of down, net operating income up nearly 10% organically and underlying earnings per share up 20.7%. The statutory 25% growth in earnings per share that leads the press release includes the €1,895m Poland gain from January and is not a run-rate. The underlying return on tangible equity was 16.0% for the nine months and 16.8% in the third quarter, against 14.9% a year earlier.

Quality of the Beat

• Revenue: soft, and in the places the market was watching. Total income missed the company poll by 0.8%, almost exactly the €127m trading shortfall. Net interest income and fees were each within 1% of consensus, but both fell sequentially once the acquisitions are removed, and the net interest margin dropped 11 basis points. Exchange rates helped total income by 1.7 points for the nine months.
• Margins: the strongest part of the print again. Costs came in €164m below the poll and fell 3.6% sequentially excluding the acquisitions. The quarterly efficiency ratio improved to 42.1% from 42.8%, and the nine-month 42.6% is the best in seventeen years on management's account. Retail and Openbank, about three quarters of the cost base, cut costs by about 2% even after absorbing both acquisitions, on management's figures.
• EPS: the beat came from tax and minorities rather than from operations. The effective tax rate was 25.2% against 26.6% in the poll, worth €89m, and minorities were €25m lower. Without those two lines, underlying profit would have missed by roughly 1%. The chief financial officer said Brazil's low tax rate should persist "for the coming quarters," which makes it a durable contributor for now, but it is not operating performance.

Revenue assessment. The nine-month revenue line is on target and the third-quarter line is not alarming in isolation, but its composition moved the wrong way. In July the revenue upside came from the rate book in Spain and the United Kingdom; this quarter the rate book held up and the emerging markets gave back. Brazilian net interest income fell 5.2% sequentially in constant euros as the bank shifted toward lower-risk lending and day counts worked against it, Chilean net interest income fell 17.5% on lower inflation indexation, and corporate and investment banking net interest income fell 10.9% after a record second quarter. Spain and the United Kingdom grew. The diversification argument works in both directions, and this quarter it worked against the line the market cares most about.

Margin assessment. A 3.6% sequential cost reduction excluding acquisitions is large for a bank of this size, and it was achieved while integrating TSB and closing Webster. Management said the cost base will keep falling "despite inflationary pressures, without taking into account the synergies from TSB and Webster." That is a stronger statement than July's admission that the 2028 cost path might slip a little, and the numbers support it. The 2028 ambition of an efficiency ratio near 36% is still six and a half points away, but the direction has not wavered in any quarter we have covered.

EPS assessment. This is the third consecutive quarter in which pre-tax profit missed the company's own poll: by 1.7% in the first quarter, 2.2% in the second and 1.1% in the third. Each time, lines below pre-tax profit closed most or all of the gap: the first quarter's miss became a 2.7% profit beat on a low tax rate, the second quarter's became an in-line result, and the third quarter's became a 1.5% beat on tax and minorities. The pattern is consistent and it is not severe: the poll systematically assumes a little less provision expense and a little more tax than the bank delivers. Investors who read the 1.5% headline beat as operational outperformance are reading the wrong line. The nine-month group tax rate of 24.8% against 26.9% a year earlier accounts for most of the gap between pre-tax growth of 12.6% and profit growth of 16.1%.

Segment Performance

Global Businesses (underlying attributable profit, EUR m)

Change columns are the company's published rates, computed on unrounded data. Third-quarter business lines sum to €3,847m against a group total of €3,848m; the difference is rounding. Payments' efficiency ratio is the company's cost-to-income figure; the business is managed on an EBITDA margin, 34.0% for the nine months.

Geographies, Nine Months (underlying attributable profit, EUR m)

Q3 2026 by Geography vs. Consensus (underlying attributable profit, EUR m)

Country consensus is the company-compiled poll, which does not carry Argentina or the rest of the group. Country lines for nine months 2025 sum to €9,625m against a reported group total of €9,624m; the difference is rounding. Spain and Portugal report in euros, and Argentina is always presented in current euros, so their two change columns are identical.

The country table is the quarter in one picture. The developed markets beat: Spain by 3.8%, the United Kingdom by 13.4%, the United States by 9.6%, Openbank Europe by 4.8%. The two Latin American markets the poll carried separately and that missed, Brazil and Chile, missed by 17.6% and 16.0%. Group profit beat anyway. That is the diversification argument working exactly as advertised, and the share price fell 5.6% on the day.

Retail & Commercial Banking

The engine, and a stronger quarter than the headline suggests. Underlying profit of €2,321m rose 9.5% sequentially in constant euros, and 6.6% excluding the acquisitions, because costs fell 4.7% organically while revenue held flat. For the nine months, profit of €6,445m rose 12.5% in constant euros and 9.5% excluding acquisitions; costs fell 4.6% organically and 8% in real terms. The efficiency ratio is 39.9%, down 2.9 points, and the third-quarter ratio was 38.8%. Excluding acquisitions, loans grew 3%, deposits 6% and mutual funds 14% year on year. The cost of risk rose 6bp to 1.20%, entirely Argentina; excluding it, it was 1.08%.

"In commercial, our new model continues to generate strong results in Spain and the U.K., connecting high-growth companies with more value-added solutions, delivering 17 percentage points of operating jaws in each market with strong revenue growth and lower cost year-on-year."
— Héctor Grisi, Chief Executive Officer

Within Retail, Spain earned a pre-tax profit 29.1% higher for the nine months and 7.2% lower in the quarter, the latter a provision normalisation after a second quarter flattered by balance-sheet optimisation. The Spanish loan yield fell 15bp year on year to 3.47% while the cost of deposits held at 0.47%. The United Kingdom excluding TSB grew pre-tax profit 44.5% sequentially, with the efficiency ratio down 8.6 points year on year to 51.0%. Brazil's retail pre-tax profit rose 16.2% sequentially on cost cuts and fees, but net interest income fell 6.0% as the mix moved toward lower-risk lending, and the yield on loans fell 36bp year on year against a cost of deposits that rose 39bp.

Assessment: Retail is doing what the transformation case promised, and doing it while absorbing two banks. The concern is the revenue line: organic loan growth of 3% and net interest income growth of 3.5% excluding acquisitions for the nine months are thin, and the third quarter added nothing organically. The profit growth is a cost story at a time when the cost story is the only part of the print the market did not question.

Openbank

The problem child improved. Underlying profit of €580m rose 7.7% sequentially, and the nine-month figure of €1,407m is down 9.3% in constant euros only because of the €245m gross United Kingdom motor-finance provisions booked in the first half and the end of United States electric-vehicle tax incentives, each of which management sizes at roughly €200m of lost profit. Excluding motor finance, pre-tax profit rose 13% to €2,544m, revenue rose 3% excluding Webster and costs fell 5%. The quarterly return on tangible equity was 12.0% against 9.0% for the nine months.

Credit stabilised rather than recovered. The non-performing ratio was 5.59%, two basis points lower than in June and thirty higher than a year ago; coverage was 69%, which the company shows as unchanged on the quarter and three points lower on the year; the cost of risk was 2.10%, up 4bp year on year and broadly flat on the quarter. Openbank Europe had corporate single names in Germany, and the quarter carried a €69m charge, reclassified into other results, for a Court of Justice of the European Union ruling on charging interest on costs associated with consumer credit. Openbank United States cut its cost of risk 25bp year on year.

"As anticipated, profit continues to reflect the end of the electric vehicle tax incentives in the U.S. The tax rate is now expected to remain broadly stable. We expect profitability to improve as we continue to scale the business and deliver further efficiencies."
— Héctor Grisi, Chief Executive Officer

Assessment: the motor-finance charge has stopped recurring, the non-performing ratio has stopped rising, and the quarterly return is recovering toward the "c.16%" 2028 target. That is better than we feared in July. But a fresh consumer-credit legal charge arrived in Europe in the same quarter the old one stopped, which is precisely the pattern our consumer-volatility bear point describes. We leave it where it is.

Corporate & Investment Banking

Seasonally weak and credit-mixed. Third-quarter profit of €691m fell 19.4% after a record second quarter, with revenue down 10.4% and costs flat; the quarterly return on tangible equity fell to 16.9%. For the nine months, profit of €2,433m rose 15.0% in constant euros on revenue up 13.8%, the efficiency ratio improved 3.8 points to 42.6% and the return was 19.2%. United States fees grew more than 30%, and new business continues to be originated at a return of around 22% on management's account.

Credit deserves the same flag we have raised for two quarters. Loans grew 28% in constant euros to €178bn. The non-performing ratio rose 9bp sequentially and 37bp year on year to 1.11%, coverage was unchanged on the quarter at 37% and eight points lower on the year, and the twelve-month cost of risk reached 0.32%, up 24bp. Third-quarter provisions fell 18.5% sequentially, which the company attributes to lower single-name charges. Asked about the composition of a book growing this fast, management put data-centre exposure at less than 1% of the group balance sheet and described it as energy-backed project finance with strong offtakers.

Assessment: the third quarter was a normal seasonal give-back and the nine-month record is excellent. The credit metrics are moving the wrong way slowly, on a book that is growing quickly, in a business where single names have now been cited as a provision driver for three consecutive quarters and sized in none of them. The data-centre answer is reassuring on the specific risk the market is worried about this month; it is not an answer to the broader question.

Wealth Management & Insurance

The cleanest line again. Nine-month profit of €1,651m rose 18.8% in constant euros, fees rose 11.4%, the efficiency ratio improved 2.1 points to 34.6% and the return on tangible equity was 58.2%. Assets under management reached €588bn, up 11%, private-banking customer assets and liabilities rose 13% to €388bn, and insurance gross written premiums rose 20% to €10.0bn, with Spanish life-savings sales up 43%. Third-quarter profit fell 4.1% on provision normalisation and transformation charges, with operating income slightly higher.

Assessment: insurance is where management sees the largest fee opportunity in the group, and the 20% premium growth says the opportunity is being taken. At a 58% return on tangible equity, this is the business that does the most for the 2028 target per euro of capital. Nothing in the quarter argues against it.

Payment Solutions

Nine-month profit of €105m against €35m a year earlier, on revenue up 15% in constant euros and an EBITDA margin up 2.4 points to 34.0%. Getnet's total payment volume rose 9% to €198bn, Getnet Platforms processed 14 billion account-to-account and card transactions, three times the prior year, and Ebury's active customers rose 25% to about 29,500. The cost per payment transaction fell 39%. Third-quarter profit fell to €27m on Brazilian tax normalisation, while operating income rose 22% sequentially.

Assessment: the operating leverage is arriving on schedule, and the business remains 1% of group profit. It is the option with the highest terminal value in the portfolio and the one with the least bearing on the next twelve months.

Key Operating Metrics

Employee, branch and customer figures for 2025 exclude Poland. Webster brought approximately 4,500 employees, 190 branches and 4.2 million customers; the "excluding Webster" trends are our arithmetic on those approximations. Net interest margin is the company's quarterly figure on average earning assets in current euros. Price to tangible book at each period end is the company's disclosed ratio; the 1.71x is the 30 October close of €10.12 on September tangible book.

Three lines in that table did more work than the rest. The leverage ratio fell 43bp in one quarter to 4.38%, which is what consolidating a US bank of roughly €62bn of deposits does to a balance-sheet measure that does not risk-weight. Tangible book fell 6.3% for the same reason. And the share count fell for the first time this year, by 133m, in the same quarter that 330m new shares were issued; the €5.0bn of buybacks executed between February and August, 463m shares, were cancelled on 1 September and outweighed the issuance. All three are consequences of one transaction, and two of them run against the shareholder in the short term.

Key Topics & Management Commentary

Overall Management Tone: confident on everything inside the developed markets and visibly less certain on Brazil, which came up in five of the thirteen analysts' questions. The posture on costs, capital and the 2026 targets was firmer than in July: management confirmed that profit will exceed €15bn including acquisitions and said year-end CET1 will most likely land above target. On Brazil it moved to an open concession that the rate path assumed at Investor Day will not arrive, and it declined to put a number on what replaces it.

1. A Beat Built Below the Pre-Tax Line, for the Third Quarter Running

Against Santander's own fourteen-analyst poll, the third quarter missed on total income by €127m, beat on costs by €164m, missed on provisions by €89m and on other results by €6m, which nets to a pre-tax shortfall of €59m, or 1.1%. Tax came in €89m below the poll and minorities €25m below. Those two lines turned a pre-tax miss into a €55m, 1.5% beat in underlying profit. The same shape appeared in the first quarter, when a 1.7% pre-tax miss became a 2.7% profit beat on a 24.9% tax rate, and in the second, when a 2.2% pre-tax miss became an in-line result.

"We expect to have a relatively low tax rate in Brazil for the coming quarters. This is also benefiting the group. We do not expect to see significant changes in the coming quarters."
— José García Cantera, Chief Financial Officer

Brazil explains the mechanism. Its nine-month pre-tax profit fell 17.8% in constant euros while its attributable profit fell only 5.0%; most of the difference is a lower tax charge, which management tied in April to interest on capital and which falls as the policy rate rises. The same high Selic rate that is holding back Brazilian loan growth and credit is lowering Brazil's tax rate. The group effective rate for the nine months was 24.8% against 26.9% a year earlier.

Assessment: nothing here is improper and the tax benefit is real cash. But three quarters of pre-tax misses against a company-compiled poll is a pattern, and it means the Street is still underestimating provisions while overestimating tax. We would rather own a bank whose beats come from costs, and the third quarter's cost beat was genuine. We do not count the tax line as evidence for the thesis.

2. Strip Out the Acquisitions and the Quarter Went Backwards

The company's own ex-acquisition bridge shows revenue down 1.5%, net interest income down 1.9%, fees down 1.3%, costs down 3.6%, provisions up 2.8% and underlying profit down 1.4%, all sequentially in constant euros. Webster's forty days added about €89m of profit and a full quarter of TSB added more, against a reported sequential increase of €80m.

"Underlying profit increased 2% quarter-on-quarter. In constant euros, it also rose 2%, supported by a good cost performance and favoured by the incorporation of Webster from 20 August 2026 and a full quarter’s contribution from TSB. Excluding M&A, underlying profit was impacted by the seasonal effects typically associated with the quarter."
— Q3 2026 Financial Report

Seasonality is a fair part of the explanation. The third quarter is historically softer for corporate and investment banking after strong first halves, and the company's constant-euro series shows revenue also slipping between the second and third quarters of 2025, by 0.4%. The difference this year is that the cost line, not revenue, carried the quarter.

Assessment: the nine-month organic average is about €3.67bn of underlying profit a quarter, with the acquisitions on top. That is consistent with the guidance and with our model. It is not consistent with a reading of the headline as acceleration, and it explains why the market discounted the beat.

3. The Net Interest Margin Fell Eleven Basis Points

The group net interest margin was 2.60% in the third quarter, against 2.71% in the second and 2.61% a year earlier. Management attributed the fall to lower inflation indexation in Chile, seasonality and the shift to lower-risk lending in Brazil, and normalisation in corporate and investment banking markets income. The chief financial officer added, of net interest income, that the Brazilian day-count and Chilean indexation effects alone "explain the drop at group level," and that a deposit of almost €5bn from a Spanish public institution distorted the Spanish margin and should reverse in the fourth quarter.

"Overall, NII is proving stronger than anticipated at Investor Day. A solid activity and a more favorable rate environment in Europe more than offset higher rates in Brazil."
— José García Cantera, Chief Financial Officer

Spain and the United Kingdom both grew net interest income in the quarter: Spanish retail by 2% on loan pricing and balance-sheet management, the United Kingdom by 6% excluding TSB on yields and volumes. Spanish deposit betas rose from 26.4% to 28% including corporate and investment banking; they were flat in the United Kingdom and fell everywhere else. The United Kingdom structural hedge stands at £119bn with a 2.6-year duration and a third-quarter yield of 3.34%, against a nine-month yield of 3.2% and 2.6% in the first nine months of 2025.

Assessment: the European rate book continues to do what we described in July, and it continues to be a rate position rather than a transformation outcome. The margin fall is concentrated in Latin America, where it reflects deliberate de-risking in Brazil as much as rates. One quarter of margin compression with identified causes is not a trend; a second quarter with the Spanish deposit reversed would be.

4. Brazil: The Election, the Rate Path, and a Plan That No Longer Holds

Brazil's presidential runoff on 25 October returned the incumbent, against the outcome markets had begun to price after a late swing in the polls. Management told the call that Brazilian market rates had moved by as much as 300 basis points and had retraced to about 200. Into that, Brazil reported third-quarter profit of €504m against a poll of €612m. Its non-performing ratio rose to 8.02%, up 13bp on the quarter and 148bp on the year; coverage fell to 71%, twelve points lower than a year ago; and the cost of risk reached 4.25%. The return on tangible equity was 12.8% in the quarter and 13.9% for the nine months.

"I agree with you that rates are not going to come down as much as we thought in the plan, but nonetheless, we have controlled sensibility."
— Héctor Grisi, Chief Executive Officer

Management's defence had three parts. Brazil is about 9% of group loans, and about 80% of the group's loan book is now in developed markets. Rate sensitivity in Brazil has been halved in six months to €30–40m per 100 basis points. And the de-risking is working on the leading indicators: 90-day arrears stabilised at 4.1% against a system rate approaching 5%, and early delinquencies improved to 3.9% while the market's deteriorated to 4.7%. The chief executive said the upside built into the Brazil plan was "due to ONE Transformation," not rates, but conceded that it is "too early" to say when the bank will return to growing its loan book.

Assessment: the Investor Day bridge named Brazil moving from a 15% to a 20% return as one of the components of the group's 2028 target. That component now depends on transformation alone, in a market where the bank is shrinking its consumer book and the policy rate sits around 14%. We treat it as at risk and open a new bear point on it below. Against that, Brazil is 14% of nine-month group profit and roughly 9% of loans, and the shares have already absorbed a large share of the damage.

5. Doubling Down: The Santander Brasil Minority Buy-In

On 30 July, a week after our last note, Santander announced a voluntary exchange offer for the roughly 10% of Santander Brasil it does not own, paid in new Santander shares at 0.4056 Santander shares per unit, with a maximum value of about €1,908m at the 30 July share price. Asked for an update after the election, the chief executive said the offer would go ahead "with the same conditions" already announced.

The arithmetic is accretive. Brazil's €2.1bn run-rate is Santander's roughly 90% share, so the 10% minority earns about €230m a year; at roughly €1.9bn of announced consideration, the group is paying about 8x those earnings with its own shares, which trade at about 10x 2026 profit at today's price. It also removes a listed subsidiary and simplifies capital allocation, and because it is paid in stock it does not consume CET1 the way a cash deal would. The maximum issuance is about 156m shares, roughly the number the €1.8bn first-half buyback just retired.

Assessment: financially sensible and strategically consistent with running Brazil as a fully owned franchise. But the timing deserves a flag. It increases the group's exposure to its weakest market by a tenth, it was announced a week after a call on which no analyst could ask about it, and it will undo most of one half's buyback in share-count terms. It is not the kind of bolt-on the February moratorium excluded, but it is a capital-allocation decision in stock, and we track it under the integration bear point.

6. Webster Closed at 147 Basis Points, and the Costs Are Now Visible

Webster completed on 20 August for $12,509m (€10,724m), 65% in cash and 35% in 329.8m new shares, at a 147bp CET1 cost against the roughly 150bp guided in July. It brought about €50bn of loans, €62bn of deposits, 4.2 million customers, 4,500 employees and 190 branches. In forty days it contributed about €89m of underlying profit. Santander US, including Webster, earned a 12.6% return on tangible equity for the nine months, and management reiterated a path to about 18% by 2028, of which 2 to 3 points comes from Webster.

"I must tell you that the franchise by itself, before Webster, was on track to deliver the 15% RoTE. Webster is basically putting on the table between 2%-3% to take the franchise by the end of 2028 to 18%."
— Héctor Grisi, Chief Executive Officer

The non-recurring charge in the quarter was €416m after tax and minorities, made up of €94m of integration costs and €322m of intangible-asset impairments, mainly in the United States and the United Kingdom, split roughly 40% United Kingdom and 60% United States on the chief financial officer's account. That brings acquisition-related charges to €666m for the year, with further "integration impacts from TSB and Webster in the fourth quarter" referred to but not sized. On synergies, the chief executive said it is "too early to tell" whether Webster's can be accelerated. Funding synergies in the United States could be "in the region of 50 million, maybe more."

Assessment: the closing was clean, the capital cost came in slightly under guidance and the first forty days of profit were material. The impairment line, more than three times the integration costs, is the less comfortable detail: it is a write-down of intangible assets, mainly in the United States and the United Kingdom, taken as part of the integration process, and a reminder that integration costs arrive in more than one line. It is a one-off, but it argues for patience on any claim that the integrations are de-risked.

7. The Share Count Finally Moved

Our test for this quarter was simple: if shares outstanding were still 14,689m at 30 September, the capital-return pillar would break. They were 14,556m. The €5,030m of buybacks executed between February and August, 462.7m shares, were cancelled on 1 September; Webster's 329.8m new shares were issued on 20 August; the net is a reduction of 133m, or 0.9% in the quarter and 2.2% on the year. The €1,825m first-half 2026 programme completed on 28 October with 150.6m shares bought, and their cancellation will take the count to 14,406m, 3.2% below a year ago.

"Underlying earnings per share grew 21%, reflecting strong profit generations and buybacks, while TNAV plus cash dividend per share increased 11%, even after the initial impact from Webster."
— Héctor Grisi, Chief Executive Officer

Two things cap the pace from here. The Santander Brasil offer could add up to about 156m shares. And the remuneration policy changes from 2027 results: the 50% payout shifts from an even split to about 35% in cash dividends and 15% in buybacks, so ordinary buybacks fall from roughly a quarter of profit to roughly a seventh. Cumulative buybacks against 2025 and 2026 reach about €9bn with the programme just completed, against the at-least-€10bn commitment.

Assessment: the register did what it was supposed to do this quarter, and the pillar is not broken. It is not repaired either. Pro forma for the cancellation and a fully accepted Brazil offer, the count is about 14,562m, essentially the 30 September level, and the policy shift toward cash dividends in 2027 slows the shrinkage structurally. We keep the pillar at risk, with a better trajectory than in July.

8. Costs: The Part of the Story That Keeps Delivering

Nine-month costs fell 2.0% in constant euros excluding acquisitions and 5% in real terms; third-quarter costs fell 3.6% sequentially on the same basis. Retail costs fell 4.6% organically for the nine months, Openbank's fell 5% excluding Webster, and United Kingdom costs fell 11% year on year excluding TSB. The quarterly efficiency ratio was 42.1%. Management attributes almost 2 points of the efficiency improvement to simplification and automation, and has put the 2028 cost base below €27bn at 2025 exchange rates.

"As you can see, year-to-date cost fell around 2% year-on-year, excluding M&A. That basically put us on track to deliver the lower cost base in 2026. Q3 costs fell alone 4% in the quarter-on-quarter, excluding M&A."
— Héctor Grisi, Chief Executive Officer

Data and AI initiatives generated €194m of what the bank calls business value in the nine months, against a target of more than €1bn over 2025 to 2028; it is the bank's own measure, combining revenue, cost and provision effects, so we note it rather than model it. The global app is live in five countries, and the Gravity core platform migration in Brazil is scheduled for the turn of the year, which management says is on plan despite slower equipment deliveries from suppliers.

Assessment: this is the pillar that has never wobbled in our coverage, and July's hedge on the 2028 cost path has been replaced by a firmer statement backed by a quarter of falling costs. Costs beat the poll by 2.4%, and this quarter they were the only source of operating leverage.

9. CET1 at 12.7%, Below the Target Range

The CET1 ratio fell from 14.0% to 12.7% (12.69% unrounded), entirely because of Webster's 147bp. Excluding it, the ratio rose 14bp, on 59bp of profit generation, 33bp of distributions, 7bp of net risk-weighted asset growth, 4bp of regulatory and model effects and 1bp of other items. Nine-month net organic generation was 75bp. The year-end target is 12.8–13%.

"For the fourth quarter, we expect supervisory charges roughly around 5 basis points. Also, remember that in the fourth quarter, we have the operational risk charge that happens once a year that is around 8 basis points. However, also in the fourth quarter is the most active in terms of asset sales and mobilization ... We would expect to do better than 12.8% at year-end."
— José García Cantera, Chief Financial Officer

The arithmetic is tight but reachable. From 12.69%, a quarter of organic generation at the third-quarter pace of 19bp less the 13bp of named charges reaches about 12.75%, and the fourth quarter's asset sales and risk transfers have to supply the rest, along with any unsized integration effects. Requirements also move: the Bank of Spain's countercyclical buffer rose from 0.5% to 1.0% on 1 October, and the European Central Bank's supervisory review cut the Pillar 2 requirement from 1.74% to 1.65% for 2027; together, minimum CET1 rises 7bp to 9.92%. The distance to the maximum distributable amount is 254bp.

Assessment: a 12.7% ratio is not a capital problem, and management has met its capital guidance through two acquisitions this year. It is a constraint: for the first time in our coverage the ratio sits below its range, and the margin for anything unexpected in the fourth quarter is close to zero. A year-end print below 12.8% would undermine the excess-capital part of the shareholder-return story and would be a downgrade trigger for us.

10. Credit: A Stable Headline with Pockets Moving

The twelve-month cost of risk rose 1bp to 1.16% and the three-month figure fell 1bp to 1.14%. The non-performing ratio improved 3bp to 2.90% and coverage held at 64%. Excluding Argentina, the cost of risk was about 1.09% and provisions grew about 6%, in line with organic lending. Argentina's cost of risk is 10.80%, up 455bp on the year, but its non-performing ratio fell 138bp in the quarter. Stage 2 loans were €95bn against €93bn in June.

"...towards the end of the year, we expect the cost of risk at around 1.15% for 2026. Slightly, given the cost of risk in Brazil and the longer interest rates that we see there, but we see it improving towards 1.1% in 2027 and 2028 as we see the benefit from TSB Webster, because that mix helped us quite a lot."
— Héctor Grisi, Chief Executive Officer

That guide is a notch softer than July's, when management pointed to a 1.0–1.1% range across 2027 and 2028. "Improving towards 1.1%" is the top of the old range, and the reason given is Brazil. Underneath the stable group figure, Brazil's coverage is down twelve points on the year, corporate and investment banking's down eight, and Openbank's down three. TSB and Webster, about 8% of the credit book, mechanically lower the group ratios because they are mortgage and commercial books.

Assessment: nothing here is a credit event, and the acquisitions will flatter the group ratios for several quarters. That is exactly why the coverage trends in the three businesses that carry the risk matter more than the headline. Coverage falling while non-performing loans rise, in Brazil and in corporate and investment banking, is the combination we would watch first.

11. The Rate Book: Spain's Sensitivity, the UK Hedge and 2027

The chief financial officer gave a constructive outlook for Spanish net interest income in 2027, "up mid-single digits or maybe a bit better," on the argument that 70% of Spanish assets reprice at variable rates and that rate increases in Spain lift asset yields faster than deposit costs. He put Spanish sensitivity to a 100 basis point rate change at about €350m. In July the same speaker put it at roughly €450m, against an aim of holding it near €500m. The €60bn ALCO book disclosed in July was not updated. Deposits in Spain are expected to grow at a low-to-mid single-digit rate in 2027.

"Our sensitivity in Spain to 100 basis points interest rate change is EUR 350 million, basically. It increases very quickly as we grow deposits. That is why when we look at NII in 2027, we see NII in Spain up mid-single digits or maybe a bit better than that."
— José García Cantera, Chief Financial Officer

Elsewhere the group has moved toward neutrality: the United States is now "completely neutral" to rates after Webster, and Brazil's sensitivity has been halved. The United Kingdom remains positively geared through a £119bn structural hedge whose yield is still rising as it rolls.

Assessment: the fee-led mix shift we underwrote at initiation was meant to reduce rate dependence. The group is reducing it in the Americas and keeping it in Europe, where management expects rates to rise. That is a sound position in this cycle, and Spain's 2027 guide is the most concrete forward number on the call. It still means the European revenue upside is a rate call.

12. The Developed-Market Engine

The United Kingdom earned €557m in the quarter, 13.4% above consensus and up 34% sequentially excluding TSB, with a quarterly return on tangible equity of 17.4% against 12.8% for the nine months. Costs are down 11% year on year excluding TSB and the cost of risk is 12bp. The United States earned €525m, 9.6% above consensus, and 24% more than a year earlier for the nine months excluding Webster. Spain earned €1,161m, 3.8% above consensus, with a 25.6% return on tangible equity. United Kingdom loans, however, were flat on the quarter.

"Year-to-date profit in the U.K. grew around 22%, excluding M&A year-on-year. That is basically by 4% year-on-year core revenue growth, okay, is important. The important thing is that you see the effects of ONE Transformation. This is probably the best franchise in terms of ONE Transformation is impacting, because you see an 11% year-on-year cost decline in the franchise."
— Héctor Grisi, Chief Executive Officer

Management reiterated about 16% for the United Kingdom's return by 2028 with TSB's 400 million cost-synergy target, gated by the Part VII transfer expected to complete in the first quarter of 2027 and a platform migration toward the third quarter of 2027. The chief executive also flagged press speculation about higher United Kingdom bank taxes, noting the bank already pays an effective rate of 45 to 46% there.

Assessment: the three largest developed-market franchises all beat, and two of them are integrating acquisitions while doing it. This is the core of our upgrade: the businesses that will carry the 2028 target are performing ahead of the Street, and the business the market sold the shares on is 14% of profit.

Guidance & Outlook

Every 2026 target was reiterated and none was formally raised. Management said the group is "comfortably on track," that profit will be "well ahead" of the €14.1bn bar excluding acquisitions, and that it will be "above EUR 15 billion of underlying net profit, including M&A for the whole year."

Implied fourth-quarter requirement. Excluding acquisitions, the group needs more than €3.1bn of underlying profit in the fourth quarter to clear €14.1bn, against a nine-month average of about €3.67bn; including them, it needs more than €3.82bn to clear €15bn, against €3,848m in the third quarter. The first bar is easy. The second requires the fourth quarter to at least match the third with a full quarter of Webster, while absorbing the annual operational-risk charge and unsized integration costs, and it is the one management chose to volunteer.

Where consensus sits. One estimate aired on the call put 2026 at €15bn or slightly above, which management endorsed. The more consequential consensus figures are Brazil's: €2.8bn in 2027 and €3.3bn in 2028 against a 2026 run-rate of €2.1bn, which management declined to endorse. If those estimates come down, as we expect, 2027 group consensus comes down with them by several hundred million euros.

Guidance style. Conservative on the formal targets and increasingly specific in the spoken ones: €15bn including acquisitions, Spain's 2027 net interest income, the fourth-quarter capital charges. The one place the guidance moved the wrong way, the 2027–28 cost of risk, was delivered inside an answer rather than in the presentation, which is consistent with how this team handled the cost-path hedge in July.

Analyst Q&A Highlights

Brazil after the election: margins, credit and what the 2028 plan assumed

The opening question went straight to Brazil: deposit costs up, the margin under pressure, provisions still high, and an election that had changed the rate outlook. It asked how much of the 2028 return target rested on lower Brazilian rates and whether the Gravity migration there was behind plan. The answer put Brazil in group context, conceded the rate point and moved the plan's Brazilian upside onto the transformation.

Q: "Looking more into the medium term, in Brazil, the election has changed the outlook for interest rates. I wonder how much of the uplift to the 20% RoTE target was based on this, on external tailwinds such as lower Selic rates, and how can you offset for that, if you could consider further restructuring."
— Francisco Riquel, Alantra

A: "On the other side, all the upside that we have projected in Brazil for the future is due to ONE Transformation. So we're doing a huge amount of simplification in our bank. ... So the plan does not change from what we had. It's exactly the same one. ... I agree with you that rates are not going to come down as much as we thought in the plan, but nonetheless, we have controlled sensibility."
— Héctor Grisi, Chief Executive Officer

Assessment: the answer is internally inconsistent in a revealing way. If the Brazil upside was entirely transformation, the rate path would not matter to the plan; the concession that rates will not fall as assumed says it did. The useful hard data came alongside: Brazil's rate sensitivity has been halved, and its early-delinquency ratio is now better than the market's. Management says the Gravity migration in Brazil remains on plan for the turn of the year despite slower equipment deliveries.

Whether the Street's Brazil numbers for 2027 and 2028 are achievable

A follow-up framed the gap directly: Brazil running at about €2.1bn this year, with consensus at €2.8bn for 2027 and €3.3bn for 2028. The question was whether management was comfortable with that path. The answer was the shortest of the call.

Q: "My first question is on Brazil. It is at a run rate of EUR 2.1 billion this year. Consensus has EUR 2.8 billion in 2027 and EUR 3.3 billion in 2028, which is a remarkable improvement. Are you comfortable with that?"
— Marta Sanchez Romero, JPMorgan

A: "Thank you. In terms of the EUR 2.1 billion, I think it is one of the possible outcomes that we have designed for that. So, that could be it, as you have said."
— Héctor Grisi, Chief Executive Officer

Assessment: a management team comfortable with a 57% profit increase over two years would have said so. Declining to endorse it, and describing the current run-rate as "one of the possible outcomes," is as close as this team comes to telling the Street its numbers are too high. We read it as a soft cut to Brazil and model it that way.

The tax rate that carried the beat, and the Brazil minority offer

A question bundled the Santander Brasil exchange offer, the timing of the promised cost-of-risk benefit, a tax rate that "was very low this quarter," and the Danish Compromise application. The answers confirmed the offer, guided cost of risk, said the low Brazilian tax rate will persist and put the Danish Compromise two or three quarters away.

Q: "Also tax rate was very low this quarter. Can you give us an indication for Q4 and 2027 in terms of tax rate in Brazil?"
— Andrea Filtri, Mediobanca

A: "No. We expect to have a relatively low tax rate in Brazil for the coming quarters. This is also benefiting the group. We do not expect to see significant changes in the coming quarters."
— José García Cantera, Chief Financial Officer

Assessment: this confirms that the line which turned a pre-tax miss into a profit beat is expected to keep helping. That is positive for near-term earnings and neutral for the thesis, because it is the mirror image of the high rates that are hurting Brazilian loan growth and credit. On the Danish Compromise, management repeated that approval will not release capital immediately; its value is in the treatment of future insurance growth.

Capital: getting from 12.7% back above 12.8%

A question asked how the remaining regulatory headwinds, recalled as about 10bp, would affect the 12.8% target the bank had just confirmed. The answer sized the fourth-quarter charges and leaned on asset sales.

Q: "You flagged that, José, during the presentation, but just wanted to get a bit of a sense of the regulatory headwinds that are pending. From memory, there should be around 10 basis points still to come in regulatory headwinds, and how that will impact the 12.8% that you confirmed during the call."
— Ignacio Ulargui, BNP Paribas

A: "For the fourth quarter, we expect supervisory charges roughly around 5 basis points. Also, remember that in the fourth quarter, we have the operational risk charge that happens once a year that is around 8 basis points. However, also in the fourth quarter is the most active in terms of asset sales and mobilization, where we expect to see a pretty strong quarter despite market conditions. We have closed already several transactions. Some are about to close. We would expect to do better than 12.8% at year-end."
— José García Cantera, Chief Financial Officer

Assessment: a quantified answer and a confident one, and the 13bp of named charges is close to the questioner's recollection. The reliance on fourth-quarter asset mobilisation "despite market conditions" is the same market dependency we flagged in July, now carrying a year-end target rather than a buffer.

A profit bridge to 2027

A question asked what 2027 profit should look like on the way from roughly €15bn this year to more than €20bn in 2028, and how much would be organic rather than acquired. The answer confirmed 2026 and declined 2027.

Q: "Firstly, I was wondering if you could help us with the profit bridge to 2027. I guess this year you're going to be at EUR 15 billion or slightly above. You guided for more than EUR 20 billion in 2028. But what is it reasonable to expect in terms of profit in 2027, and how much of that would you think is underlying growth from 2026 versus the impact from M&A?"
— Miruna Chirea, Jefferies

A: "We are on track to be above EUR 15 billion of underlying net profit, including M&A for the whole year. What is important to say is that for the numbers that you can run for 2027, 2028, and onwards is exactly what you have in, is the track that we're going to get by 2028 to what we guided the market towards above 20% RoTE. Then you can make your numbers, and you're going to see that we're on the right track to get there."
— Héctor Grisi, Chief Executive Officer

Assessment: another reaffirmation of the 2028 target without a bridge. Getting from about €15bn to more than €20bn requires roughly 15% compound growth over two years, at a moment when the Brazil component has been conceded and the organic third quarter shrank. The new disclosure is the €15bn floor, which is useful; the absence of a 2027 waypoint, with a Brazil reset in hand, is less so.

Deposit competition in Spain and the 2027 funding mix

A question noted that time deposits rose sequentially in every core country and asked about the 2027 funding strategy. The answer from the chief financial officer gave the most specific rate disclosure of the call.

Q: "Then, just going back to deposits, I know you said Spain was a little bit a one-off, but if we look across your market, term deposits or time deposits increased quarter-on-quarter in all core countries, quarter-on-quarter. Maybe if you could just talk a little bit about your funding strategy for 2027."
— Sofie Peterzens, Goldman Sachs

A: "Sofie, let me give you some details about betas. The only country where we had a slight increase in betas in the quarter was Spain, from 26.4% to 28%, including CIB. ... The U.K. was flat, but everywhere else, betas were down. In Chile, in Argentina, in Mexico, in the U.S. ... Our sensitivity in Spain to 100 basis points interest rate change is EUR 350 million, basically."
— José García Cantera, Chief Financial Officer

Assessment: the beta data is reassuring and specific. The sensitivity number is the detail to hold, because it is about €100m lower than the figure given in July for the same exposure and was presented without reconciliation. The Spain 2027 guide of mid-single-digit net interest income growth rests on rates rising, which leaves the European revenue line exposed if that reverses.

Corporate lending growth and data-centre exposure

A question asked for the composition of a €178bn corporate and investment banking book growing about 28% a year, specifically exposure to technology and data centres. The answer gave a single, reassuring number.

Q: "The second one is on the CIB book, EUR 178 billion growing around 28% per annum, with big growth actually in Spain and the U.S. So if you can give us some color on breakdown of that book between different segments, industries, exposure to tech data centers."
— Ignacio Cerezo, UBS

A: "In terms of the concentration that we have, I must tell you, in data centers, we have less of 1% of the total balance sheet of the bank in data centers. There is nothing to worry about. Data centers, it is important to understand, is project finance. That is something that the bank knows very well how to do."
— Héctor Grisi, Chief Executive Officer

Assessment: less than 1% of a €2.1 trillion balance sheet is still up to about €20bn, which would be more than a tenth of the corporate book, so the reassurance depends on how close to 1% the figure is. The broader breakdown by segment and industry, which was the actual question, was not given. With coverage in the division down eight points on the year, it remains the disclosure we most want.

What They're NOT Saying

1. A Brazil profit path to replace the one the Street is using. Asked directly whether €2.8bn in 2027 and €3.3bn in 2028 were achievable, the chief executive called the €2.1bn run-rate "one of the possible outcomes." No alternative trajectory, no restated Brazil return target and no timing for a return to loan growth were given, although the chief executive conceded the rate assumptions behind the plan will not be met.
2. How much of the 2028 group target rested on Brazilian rate cuts. The question was put in those words and answered by assigning the Brazil upside to the transformation. A plan whose Brazilian component did not depend on rates would not need the concession that rates "are not going to come down as much as we thought in the plan."
3. Why Spanish rate sensitivity fell by about €100m in a quarter. The figure given on this call, about €350m per 100 basis points, compares with roughly €450m in July and a stated aim of holding it near €500m. The financial report gives only a qualitative statement of positive group sensitivity, and the €60bn Spanish ALCO book was not updated.
4. The size of the corporate and investment banking single names. Cited as a provision driver in the first, second and third quarters, across Europe, Brazil and the United States, and sized in none of them. The division's coverage ratio is 37%, eight points lower than a year ago, on a loan book growing 28%. The data-centre figure answered a narrower question.
5. The leverage ratio. It fell from 4.81% to 4.38% in the quarter, after 5.04% in March. In July an analyst asked whether a 4% floor would be acceptable and got no answer; this quarter nobody asked, and the presentation did not mention it.
6. The size of the fourth-quarter integration charges. The chief financial officer referred to absorbing "integration impacts from TSB and Webster in the fourth quarter" on the way to the CET1 target. Year-to-date acquisition charges are €666m after tax; the remainder was not sized, and neither was the remaining TSB restructuring charge we were told in July would be about €250m.
7. A formal guidance raise. Profit will be "well ahead" of the €14.1bn bar and above €15bn including acquisitions, on management's words, but no published 2026 target moved for the third consecutive quarter.
8. A 2027 waypoint. Asked for 2027 profit, management pointed to the 2028 target and invited analysts to "make your numbers." With a Brazil reset in progress, the year between €15bn and more than €20bn is now the least specified part of the plan.

Market Reaction

• Pre-print setup (Madrid, SAN.MC): closed at €11.27 on 27 October, up 11.9% year to date against an IBEX 35 up 10.7%, up 30.9% over twelve months and down 9.8% over the trailing thirty days. The 52-week closing range entering the print was €8.75 to €12.95.
• The week before: the shares fell 4.1% on 22 October in a sector-wide Spanish bank sell-off, and 3.1% on 26 October, the first session after Brazil's presidential runoff, on 44.0m shares, 1.8 times the 30-day average, while BBVA, CaixaBank and Sabadell each rose more than 2%.
• Reaction session (Madrid, 28 October): opened at €11.01, 2.3% lower, traded a €10.60 to €11.08 range and closed at €10.64, down 5.6% or €0.63, on 49.5m shares, 2.0 times the 30-day average of 24.5m. The IBEX fell 1.4%; BBVA fell 1.1%, CaixaBank 1.4%, Sabadell 0.9% and Bankinter 0.2%; the Euro Stoxx 50 rose 0.1%.
• Reaction session (ADR, NYSE): from $12.47, the ADR opened at $12.00 (-3.8%), traded $11.63 to $12.03 and closed at $11.69, down 6.3%, on 14.9m shares against an 8.1m average. The S&P 500 rose 0.3%.
• Follow-through: on 29 October the shares went ex-dividend for the €0.127 interim payment and closed at €10.51, unchanged on a dividend-adjusted basis. On 30 October they fell 3.7% to €10.12, with BBVA also down 3.7% and the IBEX down 1.1%. Across the three sessions the price fell 10.2% and the total return was -9.1%, against an IBEX down 2.6%.

The reaction was not to the headline, which beat, but to three things in the detail that the market had been primed to look for. Brazil came in 17.6% below consensus two sessions after an election result that had already cost the shares 3.1% on a day when their domestic peers rose 2%. The beat was built on tax and minorities rather than on pre-tax profit, and it arrived alongside an 11 basis point fall in the net interest margin and a sequential decline in organic revenue. And CET1 printed below its target range for the first time. A shareholder base that had paid 1.9x tangible book for a self-help compounder found, in one release, the emerging-market exposure it had been looking through, an organic quarter that went backwards, and a capital ratio with no slack.

The volume says this was repositioning rather than panic. Turnover of 2.0 times the average on results day is heavy but below the 2.4 times recorded on the July print, which moved the price 1.4% the other way. The shares opened 2.3% lower, never traded above €11.08 and closed within four cents of the session low, which is the signature of steady selling through the day rather than a gap that was bought.

Measured from the 21 October close, before the sector sell-off and the election, the shares have fallen 15.5% against an IBEX down 3.5%. That relative move of roughly 12 points is worth about €21bn of market value, around ten times the €2.1bn that Brazil earns for the group in a year. Some of that is the wider bank-sector move, and some is a reasonable re-pricing of the 2028 target. But the gap between what Brazil contributes and what the market has removed is the core of our change of view below.

Street Perspective

Debate: Is Brazil a contained drag or a broken leg of the 2028 plan?

Bull view: the bull case on the Street is that Brazil is 9% of loans and about 14% of profit, that its rate sensitivity has been halved, that its early delinquencies are already better than the market's, and that the low tax rate is a cushion for as long as rates stay high. The developed markets, 80% of the book, beat by double digits in the United Kingdom and the United States. On this view a re-rating driven by one market is a gift.

Bear view: the bear camp contends that a higher-for-longer Brazilian rate path delays credit normalisation and loan growth together, that the consensus Brazil path of 57% profit growth over two years has been visibly disowned, and that the bank chose this moment to buy the remaining 10% of its Brazilian subsidiary. If Brazil's 2028 return stays near 14% rather than reaching 20%, the group target slips by roughly a point.

Our take: the bears are right that Brazil's contribution to 2028 should come down and that the plan is now one component short. The bulls are right about the magnitude: holding Brazil at its 2026 run-rate costs the group about a point of 2028 return, and the shares have already lost a multiple of Brazil's annual earnings in market value.

Debate: Was the beat real?

Bull view: some sell-side desks argue that the quality of the quarter was better than the reaction implied. Costs beat by 2.4%, the efficiency ratio reached 42.1%, the developed markets beat, the underlying return reached 16.8% and management put a €15bn floor under 2026 profit. The tax benefit is real cash and is guided to persist.

Bear view: the skeptics point to a third consecutive pre-tax miss against the company's own poll, provisions running hot, revenue missing, the margin down 11bp and organic profit falling sequentially. Remove the tax rate and minorities and the quarter missed by about 1%.

Our take: both are correct on their facts. The cost line is the genuine beat and it is the most durable thing in the print; the tax line is a beat in the accounting sense and a symptom of Brazilian rates in the economic one. The quarter was in line operationally and better than the poll on costs, which is roughly where a 1.7x book multiple already sits.

Debate: Is 1.7x tangible book an entry point or a value trap?

Bull view: the argument gaining traction is that the shares now trade at roughly the multiple that the third quarter's 16.8% underlying return warrants, with a 2.5% cash yield, around 2% of the share count retiring each year through ordinary buybacks, and a credible path to at least 18–19% returns from the United Kingdom, the United States and the cost base alone. On that view the 2028 target is free.

Bear view: the bear argument is that book value just fell 6.3% on Webster goodwill, the capital ratio has no slack, the 2028 bridge has never been quantified and one of its named components has just been conceded, and that a de-rated European bank with emerging-market exposure can stay de-rated for a long time.

Our take: we side with the bulls on valuation and with the bears on the bridge. At 1.71x the shares no longer require the 2028 target to be met in full; they require the current return to hold and the developed-market franchises to keep compounding, both of which the third quarter supports. That is the change from July.

Model Update & Valuation Framework

The table revises the assumptions we set out in July against the third quarter. The largest changes are to Brazil and to the capital and book-value lines that Webster moved.

Valuation framework. We apply the warranted multiple we have used since initiation, (return on tangible equity less growth) divided by (cost of equity less growth), at an 11% cost of equity and 3% terminal growth, to tangible book of €5.92 at 30 September. The reference price is the 30 October close of €10.12, the latest at publication.

How that compares with July. At the July reaction close of €12.08 the shares stood at 1.91x the June book of €6.32, against a warranted 1.625x on the second quarter's 16.0% return: an 18% premium, with 11% of upside to the 2028-target value. Today, on the same method and the same kind of book (period-end tangible book, underlying return), they stand at 1.71x against a warranted 1.725x on the third quarter's return: no premium, and 17% to 24% of upside depending on whether Brazil is haircut. The multiple fell from 1.91x to 1.71x; the return it is set against rose from 16.0% to 16.8%. The shares are 16.2% lower than on 22 July against an IBEX down 4.7%.

Valuation impact. From €10.12, the downside anchor on the nine-month return is about 5% and the base-case upside is about 17%, before a 2.5% cash dividend yield and roughly 2% a year of buyback-driven share count reduction. That is no longer a balanced distribution. In July it was roughly 11% of upside against 15% of downside; today it is roughly 17% against 5%.

Thesis Scorecard Post-Earnings

The scorecard grades the eight pillars carried since initiation and adds one. Two pillars changed status; one new bear point is opened on Brazil.

Overall: the operating thesis is unchanged to slightly stronger, and the risk distribution has moved. The costs, the developed markets and the integrations all did what we asked of them; the share count finally fell. Brazil is worse, and we have opened a bear point to say so and cut it in our numbers. What changed most is the price: the valuation bear point that kept us at Hold through three quarters of good operating news has been removed by a 16% fall since July, most of which traces to a market that contributes 14% of profit.

Action: upgrade to Outperform. In July we said we would move on a de-rating toward 1.7x tangible book, which was €10.74 on the June book. The shares closed at €10.64 on results day and at €10.12, 1.71x the September book, two sessions later. That is the trigger on either reading, and we honour it. We would return to Hold on any of three outcomes outside our base case: a year-end CET1 ratio below 12.8%; Brazilian coverage falling below 65% with the non-performing ratio still rising; or a re-rating above 2.0x tangible book, about €11.84 on today's book, before the 2028 bridge is quantified. The three are independent, and any one of them returns us to Hold.