Costs Down While Absorbing TSB, But the Revenue Beat Came From the Rate Book, Not the Transformation

Results vs. Consensus

Santander published first-half results before the Madrid open on 22 July and presented them on a call the same morning. This was the first period to consolidate TSB, which closed on 30 April and therefore contributed two months, and the first half in which Poland is absent from every line. Both facts matter more than usual for reading the print, because the company also changed the presentation of its financial information with effect from Q1 2026. The figures below are the company's underlying series unless marked otherwise.

Q2 2026 Scorecard

The consensus column is the company-compiled poll of ten analysts. No Street number existed for the statutory line, because the €250m of TSB restructuring costs booked in the quarter were not sized in advance; the statutory figure is presented here as a level rather than as a variance. A second, independently reported reading of the same poll put underlying profit at €3.75bn and net interest income at €11.47bn, so the net interest income beat is somewhere between 1.9% and 2.1% depending on which rounding you take. Nothing in the report turns on the difference.

H1 2026 vs. H1 2025 (underlying, EUR m)

The non-recurring line is the whole gap between the two profit measures. H1 2026 carries a €1,895m net capital gain from the completion of the Poland disposal in January against €250m of TSB restructuring costs in Q2; H1 2025 carries €456m of Poland's own earnings. Statutory earnings per share of €0.60 is therefore not a run-rate, and the reported return on tangible equity of 17.4% is not either. The clean series is the underlying one: 15.6% return on tangible equity for the half, 16.0% in the second quarter alone.

Q2 2026 vs. Q1 2026 (underlying, EUR m)

Excluding TSB, the sequential picture is materially thinner: total income up 1.0% in constant euros rather than 2.5%, and underlying profit up 3.3% rather than 4.8%. Roughly a third of the sequential profit growth is the acquisition arriving, which is not a criticism so much as a caution against reading the quarter as pure organic acceleration.

Quality of the Beat

• Revenue: good, but sourced from the balance sheet rather than the franchise. Net interest income beat by roughly 2% and carried the quarter; trading income all but vanished, falling from €651m to €81m sequentially. Foreign exchange was a help of 0.7 points on total income for the half and a hindrance of 0.2 points on costs, so the constant-euro and euro readings are close for once.
• Margins: the operating leverage is genuine and it is the strongest part of the print. Total income up 5.8% in constant euros against costs down 0.5%, with TSB inside the cost base. Net operating income up 11.2% in constant euros. Ex-TSB, costs fell 1.5% in constant euros.
• EPS and provisions: this is where the quarter thins out. Pre-tax profit missed by 2.2% because loan-loss provisions rose 8.5% for the half and the "other gains and provisions" line more than doubled to €715m. Underlying earnings per share of €0.48 grew 19.9% against underlying profit growth of 14.9%; the five-point wedge is the averaging effect of buybacks executed over the prior twelve months, not any reduction in the share count during this half.

Revenue assessment. Net interest income of €22,711m for the half is up 6.3% in constant euros, and the chief financial officer explicitly attributed the surprise to the rate path rather than to volume or mix. The company grew its Spanish ALCO portfolio to €60bn to hold interest-rate sensitivity near €500m per 100 basis points, and lifted the UK structural hedge from £103bn to £118bn while extending duration from 2.3 to 2.6 years. Those are deliberate positioning decisions taken against a higher-for-longer view. They are working. They are also the reason a reader should be careful about extrapolating the revenue line: the same disclosure that explains the beat explains what happens if the rate path disappoints.

Margin assessment. The efficiency ratio at 42.8% against 45.7% is the single hardest thing on this page to fake, because the cost base absorbed two months of TSB and still fell in absolute euros. Retail and Openbank together represent roughly three quarters of the group cost base, and both cut. Retail costs fell 1.2% in euros for the half including TSB and 3.1% excluding it, or 2.7% and 4.6% respectively in constant euros. Openbank cut 3%. Management characterises the group cost performance as "down 1% in constant euros, 2% excluding TSB." The published tables read -0.5% and -1.5%. The rounding is generous but the direction is not in question.

EPS assessment. Underlying earnings per share of €0.48 is the number to hold on to, and it grew 19.9%. The statutory €0.60 is inflated by a capital gain that will not repeat. What deserves attention is the composition: the share count on the register was 14,689m at 30 June, identical to 31 March and to 31 December. Every euro of per-share growth in this half came from profit and from the trailing average effect of prior buybacks. That is not a criticism of execution, since the €5.0bn programme is genuinely running, but it does mean the per-share compounding investors are underwriting has not yet shown up in the denominator.

Segment Performance

Global Businesses (underlying attributable profit, EUR m)

Segment lines sum to €7,329m against a reported group total of €7,328m in H1 2026 and to €6,376m against €6,377m in H1 2025; the €1m differences are rounding within the published tables. Both change columns are the company's published figures, computed on unrounded data, so they will not always reproduce exactly from the ratio of the rounded €m levels shown. The gap is immaterial everywhere except Payments, where the base of €16m makes the rounding visible.

Geographies (underlying attributable profit, EUR m)

All three change columns are the company's published figures, computed on unrounded data, so they will not always reproduce exactly from the ratio of the rounded €m levels shown. The Q2 column carries the company's own sequential change because the quarterly by-country levels for Q1 are not separately published. Spain and Portugal report in euros, so their current and constant-euro changes are identical; Argentina is always presented in current euros. Country lines sum to €7,329m against a group total of €7,328m in H1 2026; the difference is rounding.

Retail & Commercial Banking

The engine, and this half the well-behaved one. Underlying profit of €4,124m grew 11.7% in constant euros, and 10.2% excluding TSB, on total income up 3.9% in constant euros and costs down 2.7%, or +2.5% and -4.6% respectively excluding TSB. Loans grew 9% in constant euros, but only 2% excluding the acquisition, which is the honest read on organic lending: this is a mortgage-led balance sheet growing at low single digits and monetising better, not a franchise taking share. Deposits told a better story, up 6% ex-TSB. Digital sales rose 21% year on year and the share of products available digitally rose seven points to 73%; the cost per active customer fell 6% and non-commercial full-time equivalents per million customers fell 19%.

"In Spain, NII, as you have seen, the first half profit is up 12% year-on-year, is mainly driven by 4% revenue growth with an increase in active customers. It's very important that you see that we're growing 200,000 customers per quarter on a net basis, okay? It's very important to acknowledge the benefits of ONE Transformation. Cost is down 3%, okay? Efficiency ratio improved 240 basis points year-on-year to 33.6%."
— Héctor Grisi, Chief Executive Officer

Spain is the exemplar and it is worth separating what happened there. Spanish net interest income rose 7.7% for the half and 7.5% sequentially, total income rose only 4.3%, and the gap is a 23.0% fall in trading gains. Spanish loans grew 11.9% year on year and debt instruments grew 40.7%, which is the ALCO build showing up on the balance sheet rather than in the lending book. Spanish costs fell 2.7%. Sequentially, Spanish profit fell 2.9%, because a €424m first-quarter trading result became €16m in the second.

Assessment: the transformation is doing exactly what it was sold to do on the cost line, and Retail is the reason the group can absorb an acquisition without the efficiency ratio moving the wrong way. The caution is that Retail's revenue growth is increasingly a balance-sheet outcome rather than a customer outcome, and 2% organic loan growth is a thin foundation for the 2028 targets if the rate tailwind reverses.

Openbank

The problem child, and the one place where the numbers are genuinely getting worse rather than optically worse. Underlying profit of €827m for the half fell 18.5% in constant euros. Recurrent pre-tax profit excluding the UK motor finance provisions was €1,651m, up 15%, so management's framing that the underlying franchise is improving is defensible. But the credit metrics are not a framing question: the non-performing loan ratio ended the half at 5.61%, up 64 basis points year on year, with coverage down seven points to 70%, and return on tangible equity at 7.7%, down 2.7 points and by a wide margin the weakest of the five global businesses. Cost of risk was stable at 2.09% year on year but up 2 basis points sequentially. The €12.3bn leasing portfolio contracted 18% in constant euros.

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

The sequential recovery is real: Q2 profit of €537m was up 85% on Q1, because motor finance provisions fell from €207m to €39m and transformation charges eased. Openbank Europe swung from €45m to €208m on the same dynamic. Openbank deposits grew 3% and loans 3%, and the funding-optimisation story in the United States is working, with wholesale funding replaced by group deposits.

Assessment: we established consumer earnings volatility as a contained risk in February on the argument that the underlying franchise was improving even as the charges recurred. Six months later the charges are receding on schedule and the credit metrics have deteriorated instead. A 64 basis point rise in the non-performing ratio alongside a seven-point fall in coverage is the kind of pairing that deserves escalation, not reassurance, and we escalate it below.

Corporate & Investment Banking

The best half in the division's history and the largest single contributor to revenue growth. Underlying profit of €1,742m rose 17.2% in constant euros on revenue up 16%, with the efficiency ratio improving 3.9 points to 40.8% and return on tangible equity at 20.3%. Loans grew 22.0% in constant euros. Fee income accelerated sharply, driven by Global Banking, with United States fees up 40% off an already high base. Management said new business is being originated at a return on tangible equity of around 23%.

Two qualifications. First, the shape of the revenue is unusual and management said so: the year-on-year line-by-line comparison is misleading because transactions move between net interest income and trading income depending on how they are distributed and booked, and the honest read is total revenue. Second, CIB's cost of risk rose 22 basis points year on year to 0.29% and 6 basis points sequentially, on single names in Europe, the United States and Brazil, with the non-performing ratio up 25 basis points to 1.02% and coverage down six points to 37%. That is off a base where essentially nothing was provided in late 2024 and early 2025, so the percentage change overstates the alarm. It is still a credit line moving the wrong way while the loan book grows 22% in constant euros.

"If you look at the different lines and the movements between the different lines, I think that might not give you the right way of looking at that, because depending on when the transaction is distributed, how it's accounted, the transactions, there might be some movements between NII and income from financial transactions. The best way to look at CIB's performance is to look at total revenue, because you've seen other operating income, very good performance, trading income, financial revenue weak, NII strong."
— José García Cantera, Chief Financial Officer

Assessment: CIB is doing what a network business should do at this point in the cycle, and the 23% origination return is the most encouraging capital-allocation datapoint in the release. But a division growing loans 22% while its coverage ratio falls to 37% is one where the credit outcome is a forward question, not a settled one, and the sequential fall in profit of 4.0% shows how quickly the trading component can turn.

Wealth Management & Insurance

The cleanest line in the group. Underlying profit of €1,083m rose 19.3% in constant euros with fee income up 11%, assets under management at a record €581bn (+13%), private banking customer assets and liabilities up 15%, and insurance gross written premiums up 11%. The efficiency ratio improved 1.3 points to 35.0% and return on tangible equity was 56.9%. Sequentially, profit rose 19.7%. Management framed insurance as the largest single fee opportunity in the group on the grounds that Santander is under-penetrated relative to its natural market share in every market it operates in.

Assessment: this is the part of the fee thesis that is unambiguously working, and at a 57% return on tangible equity it is also the part that consumes the least capital. If the group is going to earn above 20% by 2028 without the rate cycle, a disproportionate share of it needs to come from here and from Payments. Nothing in this half argues against that.

Payment Solutions

Underlying profit of €78m against €16m a year ago, on an EBITDA margin up 3.8 points to 32.6%. Getnet total payment volume rose 10%, Getnet Platforms processed 9 billion account-to-account and card transactions in the year, five times more than a year earlier, and Ebury grew active customers 28% to more than 28,000. The absolute contribution remains 1% of group profit.

Assessment: the operating leverage is arriving on schedule and the margin improvement is the number that matters, not the profit multiple off a €16m base. Payments is still too small to move the group and will remain so through 2027, but it is the option with the highest terminal value in the portfolio and it is now self-funding.

Key Operating Metrics

Employee, branch and customer figures for 2025 exclude Poland. Including Poland, the group had 204,330 employees and 7,683 branches at June 2025 and 198,403 employees and 7,124 branches at December 2025. Price to tangible book is the company's disclosed ratio at each period end.

Two lines in that table are doing more work than the rest. Employees at 185,279 are essentially unchanged from March despite the arrival of more than four million TSB customers, which means the transformation is offsetting an acquisition rather than compounding on top of one. And coverage at 64%, down three points in a year while the cost of risk drifted up two basis points, is the quiet counterweight to a release whose headline is credit quality "remained solid."

Key Topics & Management Commentary

Overall Management Tone: confident and specific, and noticeably more relaxed than the February call, which had been reorganised around an acquisition announcement and ran defensively. This one ran to the usual structure, ended on time, and management volunteered numbers rather than deferring them: the ALCO portfolio size and yield, the structural hedge duration, the remaining TSB restructuring charge, the composition of balance-sheet mobilisation. Where the answers thinned was on anything requiring a forward commitment beyond the current year, with the 2028 bridge handled by reaffirmation rather than by arithmetic, and one carefully hedged admission that the cost targets "maybe slide a little bit."

1. The Revenue Beat Was a Rate Beat, Not a Mix Beat

The thesis Santander has sold since 2023 is that a fee-led mix shift makes group profit less dependent on the rate cycle. In 2025 that thesis was visibly true: fees grew 9% in constant euros while net interest income grew 0.6%. In the first half of 2026 fees grew 7.4% in constant euros and net interest income grew 6.3%, a gap of just over one point. Fees still grew faster, so the guidance was technically met. But the mechanism that produced the revenue upside was not the mix shift.

"All in all, this reflects a stronger and more resilient NII profile than anticipated in Investor Day guidance, as the benefits from higher for longer rate environment in most markets more than offset the more moderate contribution from Brazil."
— José García Cantera, Chief Financial Officer

That is an unusually direct statement. The plan presented in February assumed a rate path; rates came in higher; the difference showed up in net interest income. Management also disclosed that the group's net interest margin in the second quarter was the highest of the last twelve months, and that group rate sensitivity is structurally positive, driven by Spain and now by TSB in the United Kingdom, with negative sensitivity confined to Brazil.

Assessment: the beat is real and the positioning was skilful, but it is not the beat the equity story is priced for. A market paying 1.9x tangible book for a self-help compounder received, this half, a rate-levered bank that happened also to cut costs. Both things are good. Only one of them is durable independent of the European Central Bank.

2. The €60 Billion ALCO Book and the £118 Billion Structural Hedge

Pressed on how Spanish net interest income could run 7.7% against low-to-mid single digit guidance, the chief financial officer gave the mechanics without being asked twice.

"Second, the ALCO portfolio, we have EUR 60 billion of ALCO at an average yield of 3.3% six-year duration. This is slightly more than we had anticipated because, as I said, we are growing in current accounts more than expected, or more than planned, because of the success of our commercial strategy. Because we want to keep the interest rate sensitivity stable around EUR 500 million per 100 basis points, we increased the ALCO portfolio to this EUR 60 billion. We plan to keep it at this level. We don't plan to increase the ALCO portfolio above this level. Right now, the interest rate sensitivity is below EUR 500 million. It's EUR 450 million, more or less."
— José García Cantera, Chief Financial Officer

The United Kingdom position was disclosed with the same candour, and it moved further.

"In December, we had GBP 103 billion of structural hedge. In June, the amount was GBP 118 billion. Duration was 2.3 last year, is currently 2.6, and the yield was 3% and is 3.2%. We have now a structural hedge with a slightly longer duration and higher yields."
— José García Cantera, Chief Financial Officer

Spanish debt instruments on the balance sheet rose 40.7% year on year, which is the €60bn book showing through. The commitment not to grow the ALCO portfolio further is the useful part of that answer, because it caps the strategy: from here, Spanish net interest income growth has to come from repricing and volumes rather than from further duration.

Assessment: this is competent asset and liability management, disclosed properly, and it is why the quarter beat. It is also a directional position with a six-year duration in Spain and a lengthening one in the United Kingdom. Investors should hold the €450m per 100 basis points figure alongside the growth guidance, because it is the sensitivity that converts a faster easing cycle into a revenue miss.

3. Trading Income Collapsed and Nobody Asked Why

Gains on financial transactions fell from €651m in the first quarter to €81m in the second, a decline of €570m, and were down 28.7% for the half. Other operating income moved the other way, from €113m to €440m, covering roughly 57% of the shortfall. Group total income still rose 3.7% sequentially because net interest income and fees together added €810m.

The concentration is Spanish. Spanish trading gains fell from €424m to €16m, roughly 72% of the group decline, while Spanish other operating income rose from €104m to €387m. Management attributed the first-quarter Spanish strength to a one-off state valuation and the year-on-year weakness in CIB trading to a foreign exchange hedge that was a tailwind last year and a headwind this year, plus lower volatility in global markets and a business mix with no equities franchise.

Assessment: the trading line is small enough at group level that the collapse did not matter this quarter, which is the point of a diversified revenue base. What was not explained is the offsetting €283m sequential swing in Spanish other operating income, and no analyst pressed for it. Two volatile lines moving in opposite directions inside one country by roughly €400m each is worth a footnote that was not given.

4. Costs Fell While the Balance Sheet Absorbed TSB

Total costs of €13,211m were down 0.3% in euros and 0.5% in constant euros, and down 1.3% and 1.5% respectively excluding TSB. Reported in real terms, the group says costs fell 5%. Retail and Openbank, roughly three quarters of the cost base, both cut. CIB costs rose 7% and Wealth 6%, but in both cases well below revenue.

Management framed the achievement as structural rather than cyclical, pointing to a product catalogue reduced from around 10,000 items three and a half years ago to almost 4,000, the Gravity back-end fully implemented in Spain, Mexico and Chile and technically ready for Brazil, and the global app live in four markets.

"What I see is that, yes, we're going to get to the cost targets, and we maybe slide a little bit. Let's see. Because some of these situations, you get high inflation in some of the countries, et cetera. I don't want to overpromise. I want to tell you that we will deliver exactly as we have said for 2028. That would be my best guess. I'm slightly positive, I would say."
— Héctor Grisi, Chief Executive Officer

Assessment: that answer is the most candid thing on the call and it is worth more than a reaffirmation. "We maybe slide a little bit" from a chief executive who has beaten every cost target for three years is a hedge, not a warning, but it is the first time the 2028 cost path has been described as anything other than certain. The half itself is unambiguous: cutting the absolute cost base while consolidating an acquisition is difficult, and Santander did it.

5. The Efficiency Ratio Changed Basis in January

Effective with the first quarter of 2026, Poland is excluded line by line from all comparatives and certain recurring operating charges tied to labour and legal processes were reclassified out of "other results and provisions" into total costs. The reclassified amount was €676m in the first half of 2026 and €865m in the first half of 2025. The result is that the FY2025 efficiency ratio, presented three months ago as 41.2%, is now presented as 45.3%.

Neither change is aggressive; if anything, moving recurring labour and legal charges into costs is the more conservative treatment, and it makes the ratio harder to beat. But it does mean the improvement to 42.8% is measured against a restated 45.7%, and it arrives less than six months after the group described its FY2025 efficiency ratio of 41.2% as its best in more than fifteen years.

Assessment: the underlying improvement is genuine and it is 2.9 points. The presentational risk is that a reader tracking one number across three quarterly releases will conclude that efficiency deteriorated by 1.6 points when it improved by 2.9. Santander disclosed the change properly, but it disclosed it in a February filing to the Spanish regulator and a footnote, not in the release headline.

6. TSB Closed on Schedule and at the Guided Capital Cost

TSB completed on 30 April for cash consideration of approximately £2.9bn (about €3.3bn) and cost 55 basis points of CET1, against the timetable and the capital envelope guided in February. It contributed two months, added more than four million customers, and took group United Kingdom customers to 27 million with 16 million active. Cost synergies of at least £400m are targeted, supporting a Santander UK return on tangible equity of around 16% by 2028. Management referred to the same figure in euros on the call. The Part VII transfer scheme was submitted in the week before the call. Of an expected restructuring charge, €250m has been booked and management expects roughly another €250m in coming quarters.

The United Kingdom result was the strongest in the group in percentage terms: profit of €725m for the half, up 33.3% in constant euros, on total income up 12.0% and costs down 2.1%. Loan-loss provisions rose 74.4% in constant euros, which management attributed in part to portfolio normalisation and in part to balance-sheet optimisation initiatives.

"Then we see that TSB is a great addition to what we are doing. TSB gave us 4 million customers more. We have just finished the submission of the Part seven on last week, on Thursday. With that, we enter into the full process. We do believe that the cost saving that we promised, about EUR 400 million, are in line of what we are going to be able to deliver. I think it is too soon to say if we are going to be above that, given that we need to see the exercise of the Part seven and how those updates comes out."
— Héctor Grisi, Chief Executive Officer

Assessment: this is the first hard checkpoint on the capital-allocation bear point and it cleared. Closing on time at the guided capital cost with the legal transfer already filed is the kind of unglamorous execution that de-risks a thesis. The synergy delivery test is still 2027 and 2028, and the 74% rise in United Kingdom provisions is worth watching against a market management itself described as intensely competitive in both mortgages and deposits.

7. Webster Cleared the OCC and the ECB

The second checkpoint also cleared. Management confirmed that the Office of the Comptroller of the Currency approval was already in hand and that European Central Bank authorisation was received the day before the call, leaving the transaction on track to close in the second half. The estimated CET1 impact was quantified at approximately 150 basis points, against the roughly 140 basis points indicated in February.

"In terms of Webster, what I would tell you is that yesterday we got the authorization by the ECB to continue with the transaction. We have a very constructive engagement with all the supervisors and the approvals needed to close the transaction during the quarter. Also, as you will know, we received the OCC approval. We have just received, as I said, the ECB approval. The process is proceeding fully in line with our expectation, to close on the second half of the year, as we have said."
— Héctor Grisi, Chief Executive Officer

Management also noted that Webster reported the previous day with a 17% return on tangible equity excluding transaction costs, and characterised those results as in line with market expectations.

Assessment: the political-interference risk that was live in the question is now largely spent, because the two regulators that could have blocked the deal have not. What remains is the part that was always going to be hard: an $800m synergy programme with no revenue synergies and no dis-synergies in the return calculation, full run-rate only by the end of 2028. The 10 basis point widening of the capital impact from 140 to 150 is small but it moves the wrong way, and it lands in a half where CET1 has to fall from 14.0% to 12.8–13.0%.

8. Argentina Took the Provisions, Diversification Absorbed Them

Argentine loan-loss provisions rose 183.2% for the half to €592m, and Argentine profit fell 26.0% to €194m. That is a country contributing under 3% of group profit taking roughly 9% of group provisions. Sequentially it improved: Argentine provisions fell 18.9% in the second quarter and profit rose 172.9% to €142m.

"First of all, 12 months cost of risk in Q2 2026, close at 1.15%. All right? As I said, it's the higher provisions in Argentina and the single names in CIB. Excluding Argentina, actually, asset quality improved 2 basis points to 1.07%. All right? There is no underlying deterioration of the portfolios."
— Héctor Grisi, Chief Executive Officer

Management stopped new lending in Argentina, expects further improvement in the second half, and expects that improvement to be offset by the usual third and fourth quarter seasonality in United States auto. Group cost of risk is guided to remain around 1.15% for the year, moving to a 1.0–1.1% range across 2027 and 2028 once TSB and Webster change the mix.

Assessment: this is the diversification pillar working exactly as advertised, and it is the most defensible part of the thesis. A country-level credit event was absorbed without denting group growth. The caveat is arithmetic: the 2026–28 plan assumed a 1.0–1.1% cost of risk and the group is running at 1.15%, so the improvement to plan is being outsourced to two acquisitions that have not yet been integrated.

9. Openbank: Motor Finance Recedes, Credit Metrics Do Not

United Kingdom motor finance dealer-commission provisions totalled roughly €245m gross in the half, of which €207m was booked in the first quarter and €39m in the second. That is the fourth consecutive year the issue has taken a charge. The receding trajectory is the good news; the credit trajectory underneath is not. Openbank's non-performing loan ratio rose 64 basis points year on year to 5.61% while coverage fell seven points to 70%, and the division's return on tangible equity of 7.7% is less than half the group's.

Management attributed the cost of risk pressure to the macro environment and to credit quality deterioration in corporates within Openbank Europe, partly offset by strong performance in the United States where used car prices have stayed high and capital optimisation measures are working.

Assessment: the segment's recurring pre-tax profit growth of 15% excluding motor finance is the number management wants read, and it is legitimate. But a consumer book where the non-performing ratio rises 64 basis points and coverage falls seven points in twelve months is not a segment that has stabilised, and it is the one place in the group where the second derivative on credit is clearly negative.

10. The Share Count That Has Not Moved Since December

Shares outstanding were 14,689m at 30 June, 31 March and 31 December, and 14,885m a year earlier. The €5.0bn programme announced in February, comprising €1,830m against second-half 2025 results and €3,200m representing roughly half the capital generated by the Poland disposal, is running. The European Central Bank has now approved a further €1.8bn against first-half 2026 results, which management said would take cumulative buybacks to around €9bn against a commitment of at least €10bn for 2025 and 2026 combined.

The executive chair framed the programme in capital terms rather than in share-count terms.

"we are on track to execute a total of €6.9 billion in share buybacks, equivalent to c.114 basis points of CET1, while continuing to invest in profitable growth opportunities across our footprint."
— Ana Botín, Executive Chair

None of that is in dispute. The observation is narrower: six months into the year, with roughly €5bn of programmes running, the register shows the same number of shares it showed in December, and the Webster stock consideration remains ahead. The 19.9% growth in underlying earnings per share against 14.9% growth in underlying profit is the averaging effect of buybacks completed in the prior twelve months plus the AT1 coupon deduction, not retirement during this half.

Assessment: we flagged this pillar as at risk in February on the argument that the Webster stock issuance would neutralise the buyback inside a twelve-month window. Nothing has happened to change that, and the register confirms it. Investors underwriting per-share compounding at Santander should model the share count flat through 2026 and be pleasantly surprised rather than the reverse.

11. The Capital Walk From 14.0% to 12.8–13.0%

CET1 fell 40 basis points sequentially to 14.0%, entirely because of the 55 basis point TSB charge; excluding it the ratio rose 20 basis points. Net organic generation was 27 basis points in the quarter, comprising 61 basis points from attributable profit, 33 basis points of capital distributions consistent with the 50% payout target, and 1 basis point of net risk-weighted asset growth. A further 4 basis points came from model updates and 11 basis points went out through deductions related to deferred tax assets. Distance to the maximum distributable amount was 401 basis points and the management buffer 418 basis points.

The second-half walk has to accommodate roughly 150 basis points for Webster and, on management's own guidance, a further 15 to 20 basis points of supervisory charges, against roughly two quarters of organic generation at the current pace. That arrives at the guided 12.8–13.0% with very little to spare.

"We still think that we will have around 15 to 20 basis points charge, negative impact from supervisory actions in the second half of the year. If you look at our capacity to generate capital, and as we are improving profitability, we think that from the 12.5% post Webster, roughly, that we are at the moment, 14% minus 150 basis points for Webster, roughly, is 150. We think we should be in line or above our target of 12.8 by the end of the year."
— José García Cantera, Chief Financial Officer

Assessment: the capital position is strong and the walk is credible, but it is tight, and it is tight in a half that also contains the closing of a $12.2bn acquisition. Management also poured cold water on the idea that European regulatory easing will help, saying there is nothing concrete and no change in day-to-day supervisory behaviour, and confirming that the Danish Compromise, when it comes, adds no capital to the existing balance sheet.

12. Balance-Sheet Mobilisation as a Recurring Source of Capital

Risk-weighted assets rose roughly €16bn in the quarter, of which €13.1bn was TSB, leaving organic growth muted against healthy lending. The reason is a mobilisation programme that management sized precisely: €13.4bn in the quarter, of which securitisations were exactly half (a quarter cash, a quarter synthetic), asset sales 37% and mostly non-performing, and guarantees and other actions 13%. The programme is guided to a similar amount in the second half and to a run-rate thereafter.

"Cash securitizations are actual sales. Between sales, securitizations of both types and guarantees, can we keep on doing something like EUR 35 billion to EUR 40 billion a year at least? The answer is yes."
— José García Cantera, Chief Financial Officer

Management tied the sustainability of that run-rate explicitly to continued strong demand for private credit. The leverage ratio, put at 4.8% and trending down in the question that prompted the answer, went unaddressed.

Assessment: mobilising €35bn to €40bn a year is a legitimate and increasingly standard European bank tool, and Santander is executing it more evenly through the year than it used to. It is also a capital source with a market-dependent supply curve. A meaningful widening in private credit spreads would not break the balance sheet, but it would remove a component of the organic capital generation that is currently helping fund both the buyback and the Webster walk.

Guidance & Outlook

Every 2026 target was reaffirmed and none was raised, which is itself a choice given the chief executive described performance as running slightly ahead of plan.

Implied second-half ramp. Underlying profit of €7,328m in the first half means the second half can come in 7.6% lower and still clear the €14.1bn bar. That is a comfortable setup, and management went further, saying it expects pre-tax profit to improve in the second half over the first. Hold the €14.1bn comparison in mind, though: it is the 2025 statutory attributable profit, which included Poland's earnings and net non-recurring gains, whereas the 2026 measure is underlying and Poland-free. The bar is harder than it reads.

Where consensus sits. The company-compiled poll was effectively spot on the quarter's profit and roughly 2% light on net interest income, which is the pattern to expect for the rest of the year if the rate path holds. The Street's published twelve-month targets cluster around €13, roughly 8% above the reaction-day close, with a spread from €8 to €14.40 that is unusually wide for a name of this size and reflects genuine disagreement about whether the 2028 target is a plan or an aspiration.

Guidance style. Conservative and consistent. Santander has now beaten or met its own annual scorecard for three consecutive years, and this management team reaffirms rather than raises mid-year as a matter of habit. The one place the habit slipped was the cost target, where the chief executive volunteered the possibility of slippage against 2028 without being pushed.

Analyst Q&A Highlights

Spanish net interest income and the size and duration of the ALCO portfolio

The opening question set the agenda for the whole session. Spanish net interest income had run well ahead of low-to-mid single digit guidance, loans were up 8% at the aggregate level against 0.6% in retail, and the ALCO bond portfolio had grown by almost €20bn in six months. The response separated the client business from the balance sheet, credited the commercial strategy for the deposit mix, and then disclosed the ALCO position, its yield, its duration and the sensitivity target it exists to hold, along with a commitment not to grow it further.

Q: "I want to start with Spain, particularly NII. If you can update on your guidance for the year, the low to mid single digit growth, because first half has been above expectations. Particularly, also comment on two points here, which is loan growth, which I see is up 8%, but retail is just 0.6%. Strong CIB this Q2. How sustainable into the second half? Also, on the ALCO bond portfolio is up almost EUR 20 billion in the first half of the year. You can update on your size and ALCO strategy, size and duration."
— Francisco Riquel, Alantra

A: "The first half NII was up 7.7%, and we see these trends continuing to the second half. We see no reason to see a different trend in the second half relative to the first half."
— José García Cantera, Chief Financial Officer

Assessment: management answered fully, which is to its credit, and the answer is the most important disclosure of the quarter. The commitment to cap the ALCO book at €60bn is a useful boundary, because it means the duration bet is now sized and will not be added to. It also means Spanish net interest income growth from here has to come from repricing and volumes, and the 0.6% retail loan growth embedded in the question is the reason that is not trivial.

Single-name provisions in CIB and the shape of the cost-of-risk year

A recurring line of questioning tried to size the single-name provisions that had been mentioned in passing for Europe and Brazil and to reconcile a 12-month cost of risk that had drifted up with a narrative of stability. The answer gave the ex-Argentina figure, restated full-year guidance, and set out the macro assumptions underpinning the 2026 to 2028 range, but never sized the single names.

Q: "The first one would be if you could provide some additional color on the single name provisions that you mentioned related with CIB in Europe and in Brazil. Basically, how much did that impact the group's cost of risk in the first half or in the second quarter, as you prefer?"
— Carlos Peixoto, CaixaBank

A: "Cost of risk, I expect it to remain around 1.15%. In H2, the better cost risk in Argentina, that is already, as I said, coming better in the second quarter, and it would continue on and on. Some benefit we're going to have from the acquisition of TSB and Webster, that is offset by the usual seasonality that we have in the U.S. Remember that in the U.S., always Q3 and Q4, we have a little bit more provisions."
— Héctor Grisi, Chief Executive Officer

Assessment: the specific question was not answered. The group cost of risk is disclosed, the ex-Argentina figure is disclosed, but the quantum attributable to CIB single names is not. Single names were cited in three separate geographies across the call and the release, and not one of them was sized. With CIB's cost of risk up 22 basis points year on year and its coverage ratio at 37%, that is now a gap worth closing.

Whether the transformation is visible in the headcount

A pointed question noted that costs are falling across many geographies while group headcount is flat sequentially, and asked when the employee number would start to come down. The answer reframed the metric rather than defending it, and the reframing is honest as far as it goes.

Q: "First one is if you can give us a bit of an overview or some opinions, basically, on when do you think headcount numbers on a group basis are going to start falling down. You're cutting costs actually in many geographies already, but the headcount numbers, I think they're stable quarter-on-quarter."
— Ignacio Cerezo, UBS

A: "Even with TSB, if we look at December 2025 to June 2026, headcount is down over 2,000 people. If we obviously take TSB into account, headcount is down 3% or 4%. The headcount reduction is the consequence of implementing ONE Transformation. As we've discussed in the past, this is a long-term trend that should continue going forward. The focus is not the headcount reduction per se. The headcount reduction is obviously the consequence of implementing ONE Transformation."
— José García Cantera, Chief Financial Officer

Assessment: the reported figures support the answer. Employees fell from 187,539 at December to 185,279 at June on the Poland-excluded basis, a reduction of 2,260, achieved while adding TSB's staff. Branches fell from 6,765 to 6,496. The exchange matters because it is the first quarter in which the transformation's headline metric has been flattered by an acquisition rather than by itself, and it will be flattered again when Webster arrives.

Webster closing timing and the risk of political interference

The question raised the possibility of delay or, in an extreme case, cancellation on political grounds. The answer disclosed that the second of two required approvals had arrived the previous day, which effectively retired the question.

Q: "The second one is on Webster. If you can give us an update, basically, on when do you think the closure of the deal is going to happen? Again, we've obviously heard the risks about the possibility of delays or even, in extreme case, actually, cancellation of the deal because of political interference. If you can give us your view, basically, about that."
— Ignacio Cerezo, UBS

A: "The process is proceeding fully in line with our expectation, to close on the second half of the year, as we have said. No news on that, Ignacio."
— Héctor Grisi, Chief Executive Officer

Assessment: "no news" is the best possible answer to this question and the disclosure that preceded it is the substance. With the Office of the Comptroller of the Currency and the European Central Bank both cleared, the transaction risk has shifted from whether it closes to whether the synergies land, which is a slower and less dramatic risk but the one that determines the return.

Capital: whether European regulatory easing is real, and what the Danish Compromise adds

With capital trending ahead of target and press coverage of possible European easing, a question asked directly whether a capital reserve was in prospect and where the Danish Compromise approval stood. The answer was notably unhelpful to the bull case, in both directions.

Q: "It looks like it is trending ahead of targets, and we are seeing signs from Brussels of potential easing of regulation. The first question for you is, do you feel like you will have a capital reserve soon? Where are you on the Danish Compromise approval? I seem to recall that you were expecting approval by June 2026."
— Andrea Filtri, Mediobanca

A: "Reality is that there is much talk about easing capital requirements, directly or indirectly, for banks in Europe. Reality is that there is nothing concrete on that. In fact, on a daily basis, what we see is no change in the supervisory actions of the ECB. I think it's premature to assume that all these, like you said, rumors or comments or talk will end up having a significant impact on capital for European banks. So far, we are not seeing it at all."
— José García Cantera, Chief Financial Officer

Assessment: a management team with an incentive to talk up a regulatory tailwind declined to. That is a credibility datapoint. It also removes an argument circulating in support of European bank multiples, and it was reinforced later when the same speaker confirmed the Danish Compromise "will not add any capital to Santander" and is already excluded from the second-half capital expectation.

Remaining TSB restructuring charges and whether synergy guidance has upside

A question sought the timing and size of the restructuring charges still to be taken and asked whether a year of due diligence had revealed upside to the cost-saving target. The answer sized the remaining charge precisely and declined the invitation on synergies.

Q: "Just on TSB, a clarification with regards to the restructuring costs. Could you let us know the timing and the size of the remaining charges you intend to take this year?"
— Britta Schmidt, Autonomous Research

A: "In terms of transformation charges, okay, you're going to see that in other results, you have EUR 281 million of transformation charges. That's exactly for the whole group, okay? On the line of non-recurring items, you have EUR 250 million from what we have done so far in TSB, okay? What we expect is another EUR 250 million in the following quarters. We will try to do as much as we can, the sooner the better, okay? That's the idea of what we're trying to achieve there."
— Héctor Grisi, Chief Executive Officer

Assessment: a clean, quantified answer, and the "sooner the better" framing suggests the second €250m arrives in the second half rather than being spread into 2027. That matters for the statutory profit line, which will carry both that charge and the Webster closing costs in the same period. The refusal to guide synergies higher before the Part VII outcome is appropriately disciplined rather than evasive.

Balance-sheet mobilisation, the leverage ratio and the ceiling on risk transfer

A question tied the falling leverage ratio to the sustainability of the risk-transfer programme, asking how much capacity remained over two to three years and how low a leverage ratio management would accept. The answer sized the programme and its market dependency but did not name a leverage floor.

Q: "Your leverage ratio continues to trend down. It was 4.8% this quarter. Does that in any way limit any of the SRT capacity over the next two, three years? I know you gave very helpful guidance on the second half SRTs, but does that slightly limit the SRT capacity in the longer term, and how low leverage ratio would you be happy to run with? Would you be comfortable with a 4% leverage ratio?"
— Sofie Peterzens, Goldman Sachs

A: "Well, as long as the market remains as it is, meaning there is a strong demand for private credit, I think we should be able to continue mobilizing assets on a recurring basis. As you can see, our new origination, assets that are coming in new into our books, are increasing, and this has given us new opportunities to rotate the balance sheet. Also, markets that were closed a couple of years ago are opening up to investors, like in Latin America."
— José García Cantera, Chief Financial Officer

Assessment: the leverage floor was not given, and it is the part of the question that mattered. The capacity answer is conditioned on private credit demand remaining where it is, which is a reasonable base case and an explicit dependency. Investors should treat the €35bn to €40bn annual mobilisation as a market-contingent contributor to capital generation rather than as a structural one.

What They're NOT Saying

1. The size of the CIB single-name provisions. Named three times across the call and in the report as a driver of the cost-of-risk drift, never quantified, in a division whose cost of risk rose 22 basis points year on year and whose coverage ratio fell to 37% while loans grew 22% in constant euros.
2. The composition of the €387m of Spanish other operating income in the second quarter. It rose 272.5% sequentially and offset roughly two thirds of a €408m collapse in Spanish trading gains. Neither the release nor the call explained what it consists of, and no analyst asked.
3. A quantified bridge from 15.6% underlying return on tangible equity to above 20% by 2028. Reaffirmed three times, sized zero times. The same disclosure was deferred to the February Investor Day at the last results and has still not been given as arithmetic. The components named on this call were Brazil moving from 15% to 20%, Santander UK to around 16%, and unspecified contributions from M&A and ONE Transformation.
4. Any explanation of why 2026 guidance was reaffirmed rather than raised. The chief executive said performance is running slightly ahead of plan and separately that the group is "on track to exceed" the profit guidance, but no target moved. Either the reaffirmation is conservative or the second half carries a headwind that has not been described.
5. A leverage-ratio floor. Asked directly whether 4% would be acceptable; not answered. With risk transfer supplying a meaningful share of capital generation, the constraint on that programme is a live question.
6. Openbank's path back to a group-average return. A 7.7% return on tangible equity in a division representing roughly €220bn of loans was addressed with "we expect profitability to improve as we continue to scale the business" and no timeline, target or intermediate milestone.
7. The buyback's effect on the share count. Cumulative buyback progress toward the €10bn commitment was presented in euros and in basis points of CET1. The share count, which has not moved in six months, was not mentioned in the release narrative and was not raised in Q&A.
8. Whether the €500m per 100 basis points Spanish rate sensitivity is symmetric. Disclosed as a target to be held, and currently running at around €450m. The downside case, in a market pricing further European Central Bank easing, was not walked through.

Market Reaction

• Pre-print setup (Madrid, SAN.MC): closed at €11.91 on 21 July, up 18.3% year to date, up 70.9% over twelve months, and down 0.7% over the trailing thirty days. The 52-week closing range entering the print was €6.97 to €12.51.
• Reaction session (Madrid, 22 July): opened at €11.81, traded a €11.71 to €12.21 range, and closed at €12.08, up 1.4% or €0.17. Volume of 57.4m shares against a 30-day average of 24.2m, 2.37 times normal.
• Reaction session (ADR, NYSE): from a pre-print close of $13.65, the ADR opened at $13.89 (+1.8%), reached $13.93, and closed at $13.74, up 0.7%, on 19.9m shares against an 8.7m 30-day average, 2.3 times normal. The S&P 500 closed the same session down 0.1%.
• Positioning entering the print: the shares had gone sideways for a month, down 0.7% over thirty days after a 70.9% twelve-month run, and were sitting 4.8% below the 52-week closing high.

The shape of the day is more informative than the magnitude. The Madrid line opened below the prior close, rallied through the results presentation to a 2.5% intraday gain, and then surrendered roughly two fifths of it into the close. The ADR did the same thing more sharply, opening 1.8% higher, reaching $13.93, and closing at $13.74, two cents off the session low. That is the signature of a print where the headline reads better than the detail: the first pass on net interest income and the record-profit framing bought a rally, and the second pass on the provision line and the pre-tax miss took part of it back.

The volume tells the same story from a different angle. Turnover at 2.37 times normal in Madrid and 2.3 times on the ADR is heavy for a quarter that produced a 1.4% move, which means positions changed hands rather than prices. That is consistent with a shareholder base that agrees on the facts and disagrees on the multiple, which is precisely where this stock has been since February.

Set against the pre-print setup, the reaction is best read as relief rather than enthusiasm. A stock that has gone nowhere for a month after doubling over a year, entering a print with an acquisition mid-integration and a second one pending regulatory clearance, got clean execution on both and a 2% net interest income beat, and paid up 1.4% for it. The market took the operating news and declined to re-rate on it.

Street Perspective

Debate: Is the net interest income beat durable or a rate-path gift?

Bull view: the bull case on the Street is that the beat is structural, not cyclical. Transactional deposit growth of roughly 5% annualised has driven funding costs down independently of the policy rate, the group is adding customers at 200,000 a quarter net in Spain alone, and the ALCO and structural-hedge positions simply lock in a spread the deposit franchise created. On this reading, higher-for-longer is the occasion for the beat, not the cause.

Bear view: the bear camp points to management's own words. The profile is better "than anticipated in Investor Day guidance" because rates were higher than assumed. The group has grown a €60bn six-year-duration book in Spain and extended the United Kingdom hedge to 2.6 years, which is a position taken, and Spanish debt instruments grew 40.7% while retail loans grew 0.6%. Growth that arrives through the securities portfolio rather than the lending book is growth a competitor can replicate with a Bloomberg terminal.

Our take: the bears have the better of the mechanism and the bulls have the better of the outcome. Both effects are present, and the deposit-mix improvement is real and durable. But the incremental euro of revenue this half came from duration, and the honest way to underwrite it is to assume the deposit-franchise contribution persists and the duration contribution does not.

Debate: Is the credit drift contained or the leading edge of something?

Bull view: some sell-side desks argue this is as clean a credit picture as a bank with this footprint can produce. Excluding Argentina, cost of risk improved two basis points year on year to 1.07%, the non-performing ratio fell seven basis points sequentially, Spain and Brazil both improved, and the acquisitions will mechanically pull the group ratio toward 1.0% because a United Kingdom mortgage book and a United States commercial bank both carry lower loss content.

Bear view: the bear argument is that three separate pockets deteriorated at once. Openbank's non-performing ratio rose 64 basis points with coverage down seven points. CIB's cost of risk rose 22 basis points on single names in three geographies. Group coverage fell from 67% to 64%. Provisions came in 4% above consensus and were the reason pre-tax profit missed. A group cost of risk of 1.15% against a 2026-28 plan range of 1.0% to 1.1% is not stability, it is running above plan and relying on an unclosed acquisition to fix the mix.

Our take: the bears are right about the direction and the bulls are right about the magnitude. Nothing here is a credit event. But single names were cited in three geographies without being sized, and group coverage has fallen three points in twelve months. That combination deserves a higher burden of disclosure than it is currently getting.

Debate: Does 1.9x tangible book already discount the 2028 target?

Bull view: the argument gaining traction is that the multiple has not expanded in six months while returns improved and tangible book compounded 9.7%. The shares are worth 1.91x tangible book today against 1.93x in early February, so the entire 8.9% price return since then is book-value growth. If the group reaches above 20% return on tangible equity, the warranted multiple is above 2.1x and the shares are cheap on a two-year view.

Bear view: the sceptics observe that a warranted multiple built on a 20% return requires believing a bridge that has been reaffirmed four times and quantified never, and that the clean current return is 16.0%, not 17.4% and certainly not 20%. On 16.0% the warranted multiple is roughly 1.63x, and the shares are 18% above it. Buying that gap requires underwriting Webster synergies that will not be visible until 2028.

Our take: the bears have the arithmetic and the bulls have the trajectory. The useful observation is that the gap has narrowed, from roughly 26% above the clean warranted multiple in February to roughly 18% now, and it narrowed the right way, through earnings rather than through a de-rating. Two more halves like this one and the debate resolves itself in the bulls' favour without the price doing anything.

Model Update & Valuation Framework

The table below revises the initiation assumptions we set out in February against six months of evidence. The largest revision is to net interest income, where we were wrong.

Valuation framework. Applying a warranted multiple of (return on tangible equity less growth) divided by (cost of equity less growth), at an 11% cost of equity and 3% terminal growth, against tangible book value per share of €6.32 at 30 June:

The defensible anchor is the underlying series, because the statutory figure carries a €1,895m capital gain that will not repeat. On the Q2 underlying return of 16.0%, the shares at €12.08 sit roughly 18% above the warranted multiple and roughly 10% below the multiple full delivery of the 2028 target would support.

How that compares with February. On a consistent underlying basis, the FY2025 return of 15.2% warranted 1.525x against tangible book of €5.76, or about €8.78, and the shares were at €11.09. That was a 26% premium to the warranted multiple. The equivalent premium today is roughly 18%. The gap narrowed by nine points in six months, and it narrowed through earnings and book value rather than through a lower price: the multiple went from 1.93x to 1.91x while tangible book compounded 9.7%.

Valuation impact. Against a €10.27 downside anchor and a €13.43 upside anchor from the €12.08 reference, the distribution is roughly 11% of upside against 15% of downside before the dividend, and closer to balanced once the ordinary cash dividend is included. That is still a Hold, but it is a better Hold than February's, and it is improving in the right direction. Two further halves of this quality, or a €10.50 to €11.00 entry point, resolve it.

Thesis Scorecard Post-Earnings

The scorecard below grades the eight pillars established at initiation. No pillar has been replaced; two have changed status.

Overall: the thesis is stronger than it was in February and the balance of risk has shifted, but not enough to change the conclusion. Two of the four things we were most worried about have improved materially: the integration risk has cleared two hard checkpoints, and the net interest income decline we treated as structural turned out not to be. Against that, the two pillars that carry the equity story's per-share compounding, the fee mix shift and the shrinking share count, both weakened, and the consumer credit book deteriorated. What we have is a better bank at a similar multiple, which is worth something, and it is why the warranted-multiple gap closed nine points without the price doing much.

Action: hold. The operating case has improved and the valuation gap is closing in the right direction, through earnings rather than through a de-rating, but at 1.91x tangible book the shares still carry an 18% premium to what the clean current return supports and roughly 10% of upside to what full delivery of the 2028 target would support. We would upgrade to Outperform on either of two triggers: a de-rating toward 1.7x tangible book, which is €10.74 on today's book, or the first quarter in which Webster synergies are quantified as delivered rather than targeted. Neither is far away, and this is now a closer call than it was six months ago.