The Best Quarter of the Plan, Sold Off on £450m of Cost Management Chose to Take

Results vs. Consensus

Barclays does not file a consensus document, and the reference number in the UK market is the company-compiled poll circulated to the sell side. On that basis the quarter beat, and it beat on the line management is measured against.

Q2 2026 Scorecard

Year-Over-Year Comparisons

Quarter-Over-Quarter Comparisons

Segment Performance

Excluding the c.£225m disposal gain, the US Consumer Bank's return was 10.5% and its cost:income ratio 49%, and the group's return was 14.9%. Every division improved its return on tangible equity year on year except Private Bank and Wealth Management. That is a better distribution than Q1, when the group return fell despite four of five divisions improving.

Investment Bank

This was the quarter that made the year. Income of £3,958m was up 20%, the return on tangible equity reached 16.0% against a c.12% full-year target, and income over average risk-weighted assets rose a full point to 7.7%. Risk-weighted assets went to £204.1bn from £196.4bn, and management noted that roughly 40% of that increase was currency translation rather than risk. The division has now delivered nine consecutive quarters of positive operating jaws.

The composition matters more than the total. Equities up 45% against FICC up 1% is exactly the rebalancing management has been promising: equities is now 32% of Investment Bank income against 22% in 2023. Within markets, financing income grew 16% year on year for the eighth consecutive quarter and supplied around a third of the growth, with management attributing the gain to prime balances in Asia and noting that about 80% of it came from the top hundred markets clients.

"We continue to drive stronger and more consistent returns by maintaining RWA discipline, improving income stability and delivering operating leverage for nine consecutive quarters."
— Anna Cross, Group Finance Director

Assessment: This is the pillar that has done what it said it would. The bank has held a broadly stable capital base in the Investment Bank for three years, lifted the return on that capital from single digits to 16%, and shifted the income mix toward the businesses that recur. The one caveat is that a quarter with equities up 45% and equity capital markets up 94% is a cyclical peak by any reading, and management repeated the c.12% divisional return target for the year rather than raising it, which tells you they think so too.

Barclays UK

Income of £2,259m was up 7% with total costs exactly flat year on year at £1,195m, which produced a 53% cost:income ratio and a 20.4% return. Net interest income reached £2,000m, up 8% year on year and 1% sequentially, and the net interest margin was 3.68% against 3.55%. Lending grew 5% and deposits rose £1.7bn in the quarter, with the mortgage book at £176.7bn on gross lending of £10.0bn, described by the finance director as the largest quarterly completions the bank has recorded.

The blemish is the product margin. Management had told the market to expect the product-margin contribution to be broadly flat sequentially. It was not.

"We said that we expected that product margin cash impact to be broadly flat quarter-on-quarter, and you can see it was a little worse, and that's essentially down to that fact."
— Anna Cross, Group Finance Director

The fact in question is the ISA season, which management sized at 7% larger than the prior year and unusually competitively priced. Barclays' own maturities fell mostly in Q1 while the market's fell in Q2, which allowed the bank to price selectively, and 85% of year-on-year ISA balance growth came from Premier customers. Impairment rose to £160m from £79m, though the comparative benefited from a favourable recalibration in the retail cards book, and 30-day and 90-day card arrears at 0.9% and 0.3% remain low in absolute terms.

Assessment: A 7% income line against flat costs is the cleanest operating-leverage print in the group, and the full-year net interest income guide was narrowed to the middle of the £8.1–8.3bn range rather than cut. But this is the second consecutive quarter in which the deposit line has come in slightly behind what management told the market to expect, and the explanation has moved from swap volatility to ISA competition. The claim that the second half is neutral-to-positive on product margin now needs to be right.

UK Corporate Bank

The strongest relative performance in the group and the least modelled. Income rose 8% to £558m with net interest income up 15%, costs down 11% on the non-repeat of a prior-year conduct charge, and the return on tangible equity at 21.3% against 16.6%. The cost:income ratio was 45%, against 54% a year ago. Loans grew 12% year on year and deposits 4%, taking the loan-to-deposit ratio to 35%.

Management devoted a full slide to why. Before the 2024 re-segmentation this business sat inside the Corporate and Investment Bank and had neither the capital to lend nor the technology budget it needed, leaving it with a 31% loan-to-deposit ratio against a 50–75% peer range and a 9% lending share against a 22% deposit share. Lending share has since risen 70 basis points, roughly 1,400 new clients have been added with about 40% of them borrowing, and those new clients have driven around 70% of loan growth at a default-grade mix similar to the existing book.

Assessment: This is a share-recapture story with a visible mechanism, a stable risk profile and a nine-point improvement in cost:income in twelve months. The finance director named it in Q1 as a division where consensus was too low and confirmed on this call that consensus has since moved up. The pillar is intact and, on the evidence of a 21.3% return against a high-teens full-year target, still conservatively guided.

US Consumer Bank

The headline is spectacular and the underlying is flat. Reported income of £1,136m was up 38% and the return on tangible equity was 30.2%. Strip the c.£225m gain on the American Airlines exit and income was £911m, up 11% year on year but down roughly 7% from the £983m the division earned in Q1, and adjusted attributable profit was £87m. That is the identical figure the division earned in Q2 2025, on an adjusted return of 10.5% against 10.2%.

Three things happened at once. The American Airlines exit removed roughly $6.5bn of reported receivables in April, which mechanically lifted the margin 40 basis points and cut the net interest income line 11% sequentially. Best Egg completed on 1 May and added around $11bn of managed balances plus c.$45m a month of cost. And higher benchmark rates compressed the market-wide pricing of originate-to-sell loans, which pressured the fee line at exactly the moment Best Egg was supposed to start filling it.

"Higher benchmark interest rates in the quarter compressed market-wide pricing of originate-to-sell loans, creating pressure on the non-NII line. Depending on the persistence of this effect, quarterly income may not fully build back to the Q1 level during '26, despite volume momentum."
— Anna Cross, Group Finance Director

The operational metrics that management manages to are all going the right way. Core deposits reached 82% of funding, comfortably through the 75%-by-2026 target that the division set itself, retail deposits grew 16% year on year, retail cards rose to 24% of card balances, and delinquencies fell in a quarter when the American Airlines exit was expected to lift them by 30 and 20 basis points mechanically.

Assessment: The franchise build is on schedule and the earnings are not. This is the pillar we rated on track in April on the strength of an 18.8% quarterly return, and a quarter later the clean number is 10.5% with a c.12% full-year target that now requires the second half to average roughly nine. The strategic case for a 25-million-customer branchless US digital bank with a 39% cost:income ratio is unchanged. The income trajectory has been pushed out by at least two quarters, and management said so.

Private Bank & Wealth Management

Income rose 5% to £366m against costs up 12% to £267m, taking the cost:income ratio to 73% from 68% and the return to 26.9% from 31.9%. Client assets and liabilities grew 8% year on year to £230.2bn and assets under management rose 15% to £55.8bn, but net new assets under management were only £0.3bn in the quarter against £0.9bn a year ago. Management cut the 2026 cost:income guide from high 60s to mid 70s.

Assessment: This is the fourth consecutive quarter in which income growth has trailed cost growth in this division, and it is now the only guidance line in the plan that has been formally moved in the wrong direction. Management frames the spend as investment in Planning and Advice and in removing custody charges on the direct-investing platform, both of which are defensible propositions with no near-term revenue attached. The division is 4% of group income, so it does not move the group return. It does show that the cost discipline is not uniform.

Key Topics & Management Commentary

Overall Management Tone: Assured and, for the first time in the four quarters we have observed, willing to spend from a position of strength rather than defend from one. The prepared remarks led with the cost decision instead of burying it, and the answers on strategy and on the divisions were specific and unhedged. The exception was 2027, where three separate attempts to obtain any absolute cost anchor produced only comparative adjectives, and where the reluctance was pointed enough that it became the dominant theme of the call.

1. The £450m Cost Decision, Which Is Why the Stock Fell

Barclays raised its 2026 income target by £500m and committed roughly £450m of incremental second-half cost against it. The components are up to £500m of structural cost actions in the second half against a normal annual run-rate of about £300m and c.£106m already recognised in the first half, which is roughly £300m of genuinely incremental spend, plus £100–150m for a change in the pay mix of Material Risk Takers toward higher variable and lower fixed compensation from the 2027 grant year.

"Therefore, after our increase in distributions, we will use some of the capacity from our stronger first half profitability to take further cost actions later in 2026. We anticipate that this will create greater cost flexibility from 2027 onwards."
— C.S. Venkatakrishnan, Group Chief Executive

Management asserts a return on investment of around 100% within twelve months and says the 2026 cost:income target of high 50s, the distribution plans and every financial target are unaffected. The pay-mix change is presented as a response to regulatory change rather than a competitive decision, and management expects it to land disproportionately in the Investment Bank.

Assessment: Everything about this is defensible on a three-year view and nothing about it helps 2026. The market was handed a £500m income upgrade and an unchanged cost:income guide in the same breath, which is a way of saying the upgrade has already been allocated. Whether that is good capital allocation depends entirely on a 100% return claim that cannot be verified until 2027, made by a management team that then declined to quantify 2027 costs.

2. The Structural Hedge Is Now £19.5bn Contracted Across 2026 to 2028

This remains the load-bearing disclosure of the investment case, and it improved again. Gross structural hedge income is now shown as £7.1bn for 2026 with more than 95% already locked in, £6.7bn locked in for 2027 and £5.7bn locked in for 2028. Those three years sum to £19.5bn against the £18.3bn disclosed a quarter ago. Gross structural hedge contributions in the first half were £3,456m against £2,778m a year earlier, and the hedge accounted for around 45% of group net interest income in the quarter.

The mechanics behind the roll-up are unchanged and still favourable. Maturing yields run at roughly 1.5% in 2026, 2.1% in 2027 and 2.7% in 2028, about £35bn matures each year, the notional is broadly stable at £239.8bn against £236.1bn at December, and the average duration is around three and a half years. The plan assumes reinvestment at 3.5%.

"Structural hedge income growth is predictable and benefits all divisions, accounting for circa 45% of Q2 Group NII. It will drive around half of the planned Group income growth from '25 to '28 and remain a meaningful tailwind beyond. As a planning matter, these expectations are based on a 3.5% reinvestment yield. Swap rates were above this level again in Q2 at 4.3%, further supporting NII in future years."
— Anna Cross, Group Finance Director

One correction to how we read Q1. The £6bn increase in hedge notional we flagged three months ago turned out to be largely a reallocation rather than incremental income. Management explained that the end-of-Q1 top-up "led to a switch from product margin to hedge income, with minimal net effect" in Barclays UK. The locked-in income figure is the number that matters, not the notional.

Assessment: The gap between a 4.3% achieved reinvestment rate and a 3.5% planning assumption is now two quarters wide and is compounding into the outer years, where the maturing yield is lowest and the roll-up is largest. This is the most falsifiable, most contracted and least discussed source of income in a UK bank, and it got bigger.

3. The Finance Director Confirms the Outer-Year Consensus Moved

In April the finance director told the market for the second consecutive quarter that 2027 consensus looked light and 2028 very light, and named the US Consumer Bank and UK Corporate Bank as the two divisions where she thought the Street was wrong. We wrote at the time that if consensus did not move over the following two quarters despite the disclosed hedge position, the first pillar of the thesis was wrong in public. Asked on this call whether the Street had in fact moved, and specifically whether the US Consumer Bank had reached roughly the right level while the UK Corporate Bank remained too low, she answered directly.

"Clearly, the analyst community has listened, and we've seen a movement in that consensus for 2028. We have confidence in all of our businesses, and we have confidence in the UK Corporate Bank, as well as the US Consumer Bank. So, we've seen both of them move up."
— Anna Cross, Group Finance Director

Assessment: The test passed on the first of the two quarters we allowed for it. That does not make the forecasts right, but it does remove the specific failure mode we were watching for, which was a management claim about outer-year income that the market simply ignored. The claim was heard, and the numbers moved.

4. Income Guidance Raised Twice, and What the Second Half Actually Implies

The 2026 group income target went from c.£30bn at the February 2024 investor update to c.£31bn in February 2026 and now to c.£31.5bn. Group net interest income excluding the Investment Bank and Head Office was raised to more than £13.7bn from more than £13.5bn. First-half income was £16,501m, so the new full-year target implies roughly £15.0bn in the second half against £14,244m in the second half of 2025, or about 5% growth after 11% in the first half.

Assessment: The guide embeds a material deceleration, which is the correct posture after a quarter in which equity capital markets income rose 94% and advisory rose 53%. Management is not extrapolating a cyclical markets peak into the outer quarters. On our reading the target is beatable if banking activity holds anywhere near current levels, and this is the third consecutive quarter in which the income guide has proved conservative.

5. Distributions: A Doubled Buyback and a Change in Dividend Shape

The quarterly buyback went from £500m to £1.0bn and the interim dividend to 5.9p per share, or c.£800m, against 3.0p a year ago. First-half distributions of £2.3bn are up 61% year on year. Buybacks announced in the first half total £1,500m against £1,000m in the first half of 2025. Management also disclosed a deliberate change in dividend phasing from roughly one-third in the first half to a 40:60 split, on the grounds that it is more consistent with US and European peers and gets cash to shareholders faster.

Against the at-least-£10bn 2024 to 2026 pledge, the running total is £9bn with £1.2bn of dividend and two more quarterly buybacks still to come.

Assessment: In April we wrote that a £500m quarterly buyback annualised to roughly £4bn against a 2026 to 2028 pledge that needs to average above £5bn a year, and that the pledge therefore looked back-end loaded. One quarter later the pace has doubled and the phasing has been pulled forward. The arithmetic objection has been answered faster than we expected.

6. Credit: 51 Basis Points, and the Post-Model Adjustments That Were Consumed

The group impairment charge of £571m equated to a 51 basis point loan loss rate, the low end of the through-the-cycle 50 to 60 basis point range and 23 basis points better than Q1. The Investment Bank charge fell to £44m from £67m a year ago, at a 13 basis point loan loss rate. First-half impairment of £1,394m at 62 basis points still carries the £228m single-name charge taken in Q1, which contributed around 20 basis points to the group rate.

The accounting detail is worth noting. Consensus unemployment expectations rose as management had anticipated, which consumed the post-model adjustments recognised in the prior quarter in Barclays UK and the US Consumer Bank. A £68m post-model adjustment was retained in the Investment Bank for global macro uncertainty.

Assessment: This is the commitment from April that has been most cleanly delivered. Management said the move to the top of the range was the single name alone and contained no macro view. A 51 basis point quarter is the evidence. The one thing to watch is that the buffers built last quarter have now been spent against a deteriorating unemployment forecast, so the next adverse macro revision has less to absorb it.

7. Capital: 14.3%, an Above-Range Position, and 8 Basis Points of Regulatory Relief

The CET1 ratio held at 14.3%, above the 13 to 14% target range, and falls to 14.0% pro-forma for the announced buyback. Capital generation was 62 basis points in the quarter and 115 basis points in the half. The first-half walk shows 108 basis points from attributable profit, 68 basis points of distributions, and 40 basis points from a £9.8bn increase in risk-weighted assets excluding currency effects. Risk-weighted assets rose to £364.8bn from £356.8bn, with the American Airlines exit and Best Egg acquisition together releasing £3.2bn net.

Guidance for 2027 regulatory inflation is unchanged at c.£19–26bn, split between c.£8–15bn from Basel 3.1 from 1 January 2027 and c.£11bn from the US Consumer Bank moving to an internal ratings based model, now expected in the second half of 2027. Management continues to expect an offsetting Pillar 2A reduction that it has not sized. Separately, proposed UK leverage changes would cut Barclays' Tier 1 leverage requirement by 8 basis points.

Assessment: Holding above the top of the range while doubling the buyback is a deliberate signal that management expects regulatory clarity to be favourable rather than punitive. The unsized Pillar 2A offset remains the single largest unquantified variable in the 2027 capital picture, and it has now been unsized for three consecutive quarters.

8. Tangible Book: 423p, and a Compounding That Nobody Discusses

Tangible net asset value per share reached 423p from 409p at December and 384p a year ago, a 10% year-on-year increase. The half-year bridge is first-half earnings of 30.7p against a 6p negative movement in the cash flow hedging reserve, a 6p reduction from prior-year dividends paid in Q1 and a 6p reduction from share awards vesting.

Assessment: In April the same rates that lifted income were cutting the book through the hedging reserve. That drag continued at 6p in the half but was comfortably outrun by earnings. A bank compounding tangible book at 10% while distributing 61% more than a year ago is doing both things at once, which is the whole argument for owning it. The one qualification is that buying back stock at 1.19 times tangible book is now mildly dilutive to book per share, which was not true at 1.05 times in April.

9. UK Lending Growth Is Capability-Led, Not Price-Led

Group lending grew across every division. Barclays UK lending was up 5% year on year, UK corporate lending up 12%, and £25bn of the planned c.£30bn of UK business growth risk-weighted assets has been deployed since 2024, of which £18bn was organic. Mortgage application share exceeded stock share for a ninth consecutive quarter. Asked whether volume growth was being bought with margin, management rejected the framing.

"The thing that's driving the growth in NII across Barclays is not price, it's capability build. And that's true of every single product that we have... So, we don't feel like we are under margin pressure."
— Anna Cross, Group Finance Director

Assessment: The claim is partly testable and partly not. Kensington in mortgages and Tesco in cards are real distribution assets rather than price concessions, and the UK Corporate Bank client-acquisition data shows growth at a stable default-grade mix. But the group net interest margin story is still dominated by the hedge, and the same executive conceded that the product margin came in worse than guided. Capability-led is the right description of the balance growth. It is not yet the right description of the margin.

10. The Motor Finance Provision Sits at £430m With an Unquantified Legal Tail

The provision was unchanged in the quarter at £430m after the £105m Q1 top-up. On 2 July the Upper Tribunal ordered a suspension of parts of the Financial Conduct Authority redress scheme following four legal challenges to the final rules, with hearings expected in Q4 2026 or Q1 2027. Barclays declined to challenge the rules itself while stating that it "strongly disagrees with aspects of the rules which require financial redress even where customers suffered no demonstrable financial harm."

Assessment: The provision has stabilised, which is progress, and litigation and conduct charges fell to £4m in the quarter from £76m a year ago. But the filing still carries no range, no scenario weighting and no per-claim sensitivity around a number that management moved from a probability-weighted basis to a single scenario one quarter ago, and the scheme it is provisioned against is now partially suspended by a tribunal. The direction of that uncertainty is not obviously adverse. It is simply not sized.

11. Best Egg, Samsung and GoHenry: Three Acquisitions and a Partnership in One Quarter

Best Egg completed on 1 May for c.£0.6bn, adding around $11bn of managed balances that are originated to distribute and therefore do not sit on the reported balance sheet, at c.$45m a month of cost. A de novo partnership was announced to provide card products in the Samsung wallet in the US. GoHenry, a UK children's money-management platform, was announced on 12 June for completion in Q4. And on 30 June Barclays acquired a 999-year leasehold interest in One Churchill Place for £750m, described as broadly neutral to capital and earnings.

Assessment: The execution capacity is the point. Exiting a large co-branded partnership, integrating an acquisition and launching a new partner in a single quarter is a real demonstration of the platform argument that underpins the whole cost thesis. Management declined to size the Samsung economics, which is normal practice for partnership deals and also means the most forward-looking of the four items is unmodellable.

12. Market Risk Discipline, Now Verifiable in the Filing

Management cited stable value at risk and no loss days as evidence that the Investment Bank's income growth did not come with risk escalation. Unlike the first quarter, the half-year filing does contain the supporting table: average total management value at risk of £20m in the first half against £19m a year earlier, with a high of £29m against £30m. Income over average risk-weighted assets rose to 7.9% for the half from 7.2%.

Assessment: Eleven per cent more first-half Investment Bank income on flat value at risk and a broadly flat capital base is the cleanest available evidence that the returns are structural rather than a beta trade. The loss-day count management referenced is still not in the filed document, so that half of the claim rests on the presentation.

Guidance & Outlook

Implied second-half ramp: First-half income of £16,501m against a c.£31.5bn full-year target implies roughly £15.0bn in the second half, about 5% above the £14,244m earned in the second half of 2025 after 11% growth in the first half. On costs, first-half total operating expenses of £9,065m plus up to £500m of structural cost actions, £100–150m of pay-mix cost and the annual UK regulatory levy that lands mostly in Q4 puts the second half near £9.6bn and the full year near £18.7bn, which is a high-50s cost:income ratio on the new income target. The guide is internally consistent and leaves the income line with room.

Street at: Consensus for 2027 sits at £18.2bn of total operating expenses and a 55.5% cost:income ratio, both quoted back to management on the call. Management would confirm neither, saying only that 2027 should be a meaningful step from the high 50s toward the low 50s. If the 2026 exit is near £18.7bn, the £18.2bn consensus requires an absolute cost decline in 2027, which is what the second-half structural cost actions are supposed to fund.

Guidance style: Conservative and consistent. The 2026 income target has been raised twice from an original c.£30bn, the return target has never been lowered, and the divisional targets have been beaten in four of the five divisions this quarter. The one exception, the Private Bank cost:income guide, was cut rather than quietly missed, which is at least the honest version of a disappointment.

Analyst Q&A Highlights

Whether Every Income Upgrade Now Gets Spent

The sharpest exchange of the call, and the one that framed the whole cost debate in a single sentence of arithmetic. The question observed that guidance had moved income up £500m and costs up £450m, and asked the obvious follow-on: if the bank has access to investments returning 100% within a year, why would the next strong quarter not be spent the same way. Management's answer confined the spend to what it judges it can execute rather than to what is available.

Q: "So it sounds like you want a balance between distribution, growth, various different things. Well, I think at this point, it looks like in the guidance, income is now £500 million higher and incremental cost is £450 million higher. So if there is another quarter of very strong IB income, et cetera, would a lot of this incremental income be spent on SCAs? Because obviously, if you are having opportunities to invest in things that give you 100% ROI, that probably suggests that you have many other opportunities to invest and improve the long-term efficiency of the company. Is that the right way to think about it?"
— Perlie Mong, Bank of America

A: "In terms of an operational list and the level of investments that we can deploy wisely on behalf of the shareholder, we think that's the right sort of level. So clearly, as we look forward from here, Venkat and I have at least as much confidence in the RoTE of the firm as when we started the year for 2026. And you can see that in the fact that we have upgraded our income twice, first to £31 billion and then to circa £31.5 billion. So we think there is momentum in the business, but this will be a big operational lift for us in the second half."
— Anna Cross, Group Finance Director

Assessment: The answer is an execution-capacity argument, not a returns argument, and it is more credible for that. Management is saying the constraint is how much change the organisation can absorb, which implies the spend is bounded rather than open-ended. What it does not do is commit to a ceiling. An investor who believes the 100% return claim should want more of this spend, not less, and the fact that the market marked the stock down 4.8% suggests the market does not yet believe it.

Whether the 2027 Cost Base Has Any Anchor at All

The first question of the call went straight at the consensus operating-expense line for next year, and management declined to validate or reject it. The same ground was covered again later from the cost:income angle with the same result. Across three attempts the market obtained comparatives and no absolutes.

Q: "You talked about a commensurate improvement in Group efficiency in 2027 from the £300 million of additional SCA (Structural Cost Actions) spend in H2. So, could you comment then on how you feel about the £18.2 billion of [total operating expenses] that consensus has in for next year? Does that feel too high or too low when you think about the SCA tailwinds on the one hand, versus some of the higher compensation costs on the other hand?"
— Chris Hallam, Goldman Sachs

A: "Now what we haven't done is given absolute targets in costs. The way we think about costs is because returns are our north star, cost: income ratio is our north star, because to the extent that the business has a degree of momentum around it, we may incur further absolute costs. So that's what's really important to us. We believe that the cost: income ratio in 2028 will continue to be a low 50s print, and we expect 2027 to be a meaningful step towards that from the high 50s cost: income ratio that we expect in the current year."
— Anna Cross, Group Finance Director

Assessment: The cost:income framing is intellectually coherent, because a bank that grows income should be allowed to grow costs. It is also unfalsifiable within any single year, which is exactly why the question kept coming back. A management team asking the market to fund a 2027 benefit with 2026 cash has a stronger claim on that funding if it will say what 2027 costs. This is the one place on the call where the answer was materially weaker than the question deserved.

When US Consumer Bank Income Returns to the First-Quarter Level

Management had told the market in April that the £983m first-quarter income figure was a reasonable base for subsequent quarters. The second quarter came in at £911m excluding the disposal gain, and a question in the second half of the call sought confirmation that the prior statement had been withdrawn. It had.

Q: "I think you've previously said that the Q1 [income] figure of £983m, with maybe a little bit of a growth, was sort of a reasonable gauge into Q3 and beyond. I just want to check, I think you made an earlier comment from the call to suggest maybe that wasn't the case. So can I just sort of check that comment you made earlier in terms of when we get back to that Q1 run-rate."
— Guy Stebbings, BNP Paribas

A: "So if you recall, and clearly you do, in Q1, I said I felt that Q1 was a good basis for subsequent quarters. I think given those two rate impacts that I've called out, it feels like it'll take us a little bit longer to build back to that base, Guy, and that's really what we're saying here, primarily in terms of those sold securitisation prices. Although, remember, as a bank, we do have some optionality here. So, we have probably greater optionality than some participants in that market, in that we can choose to hold some on the balance sheet if we feel that's the right thing to do."
— Anna Cross, Group Finance Director

Assessment: Clean and prompt, which is the right way to withdraw a number. The optionality point is more interesting than it sounds: a bank that funds itself with 82% core deposits does not have to sell originate-to-distribute paper into a compressed market and can warehouse it instead. That converts a fee-income problem into a balance-sheet decision, which is a better problem to have. It also means the division's income line becomes harder to forecast, because the mix between sold and retained is now a management choice.

Whether the UK Deposit Pressure Is Seasonal or Structural

Barclays UK net interest income grew 1% sequentially against guidance for growth after a muted first quarter, and the product-margin component came in behind. The line of questioning pressed on whether a competitive ISA season is a sufficient explanation when loan growth is strong and deposit competition has proved persistent in other markets. The answer laid out a specific mechanism for why the second half converges.

Q: "The question is really, what's the level of confidence that it's just the ISA season, because loan growth is strong, presumably we've seen in other regions that driving persistent deposit competition, but you're obviously confident because your BUK NII guidance is pointing to 3% quarter-on-quarter growth, run-rate more or less for the next couple of quarters. So maybe some, if you can expand on your confidence, why this is transitional and NII should accelerate from here."
— Alvaro Serrano, Morgan Stanley

A: "Remember that term deposits are in large part one-year products. And we saw the start of this more intense competition in the second half of last year. So, as we go forward from here, what you're going to see is that the maturing margin and the prevailing front book margin are somewhat converging in the second half of the year. That was not true for the market in the first half. If pricing stays where it is, they'll be broadly neutral by Q4."
— Anna Cross, Group Finance Director

Assessment: This is a dated, mechanical and testable claim rather than an expression of confidence, which is the standard we want. The back-book to front-book convergence argument only works if deposit pricing does not intensify further from here, and management was explicit about that condition. The full-year guide narrowing to the middle of the range rather than the top is the honest reflection of the first-half shortfall.

Whether the Barclays UK Absolute Cost Commitment Survives the Investment

A commitment made in April was put back to management directly: that Barclays UK costs would fall in absolute terms in each of the next three years. The additional second-half spend put that in doubt, and the answer was a single word before the elaboration.

Q: "In terms of this additional investment in the second half, I think previously you were committed to BUK costs falling every year for the next three years, I think that's the way you articulated it. Is that still your commitment? So even with that additional investment, will costs in BUK still be down each year for the next three years?"
— Edward Firth, KBW

A: "So the answer to your first question simply is yes... For BUK, we still expect sequential falls year-on-year during this plan and for the Group as a whole, I expect positive jaws in every year of the plan. So that remains completely unchanged. That's what we meant when we said it's the momentum that allows us to do this, without disturbing the commitments that we have to the outside world."
— Anna Cross, Group Finance Director

Assessment: The commitment was reaffirmed without qualification, which matters because the prepared remarks had already conceded that the second-half spend "may lead to higher costs in H2 versus H1" in that division and that the efficiency benefit lands in 2027 and 2028. Barclays UK total operating expenses were flat year on year in the quarter and up 2% in the half, so the absolute decline has to come from somewhere. The most natural reading is that 2026 is now the flat year and the falls begin in 2027, which is a softer version of what was said in April.

Whether There Is Any Net Interest Income Growth Without the Hedge

The most probing follow-up of the call inverted the structural hedge argument. If gross hedge income rises £1.2bn in 2026 and group net interest income grows by less than that, the underlying franchise is contributing nothing and the whole story is a rates position that eventually runs out.

Q: "On your Slide 10, you show £1.2 billion of additional revenue from the hedge and yet if I'm looking at your guidance, banking NII is not going to grow by £1.2 billion, it's going to be somewhat less than that. So that would sort of imply that despite all the volume growth, the underlying NII, if it wasn't for the hedge, we wouldn't be getting really any NII growth at all. And I'm just thinking about what that means as we go beyond this year and into sort of '28, '29, '30."
— Edward Firth, KBW

A: "The thing I would also say is look back over the last few quarters, Ed, and add up what's coming in from the asset balance column in BUK, and you will see that you are getting to a building momentum of asset income. I think you'll get to like £130 million to £140 million of additional income coming through in BUK alone on that metric... Of course, at the moment, this is being disproportionately driven by the structural hedge, but at some point in the future, beyond 2028, it's really important that the diversification within NII comes through, and those asset balances start to carry that momentum from there on in."
— Anna Cross, Group Finance Director

Assessment: The concession inside the answer is the important part. Management accepts that net interest income growth is disproportionately hedge-driven today and frames the asset-balance contribution as the thing that has to take over beyond 2028. That is the honest structure of the position: a contracted, visible tailwind through 2028 and a franchise that has to be built out underneath it in the meantime. It is also why the £25bn of UK growth risk-weighted assets deployed against a c.£30bn plan is a more load-bearing number than it looks.

The Shape of the Dividend

An observation that the first-half dividend has historically been 30 to 35% of the full year, and that a 5.9p interim therefore implies more than the £2bn full-year figure management has guided to, drew out a phasing change that had not been separately announced.

Q: "In the last four years, the first half dividend has tended to be 30% to 35% of the full year. Is that the sort of shape we would expect moving forwards? Because if it is, it's obviously pointing to a slightly higher dividend this year than consensus has got in and probably a little more than the £2 billion that you talked about as well."
— Jonathan Pierce, Jefferies

A: "You're right over the last few years, we've typically done one-third, two-thirds. We've made a deliberate shift towards a 40% to 60%. The reason that we've done that is we feel it's more reflective of our US and our European peers. It is more consistent with our general desire to get distributions into the hands of shareholders more quickly. See it as consistent with the move to the quarterly buyback."
— Anna Cross, Group Finance Director

Assessment: The answer closes the door on the upside implied by the question. A 5.9p interim at a 40:60 split points to roughly £2bn for the year, which is the guided figure, rather than to an upgrade. That is a fair answer and a mildly disappointing one, and it is consistent with a management team that would rather return the marginal pound through buybacks it can size quarterly than through a dividend it has to sustain.

What They're NOT Saying

1. Any absolute cost figure for 2027. Asked three times, from three angles, by three analysts. The bank will give a 2028 cost:income ratio, a 2026 cost:income ratio and the direction of travel in between, and it will not give a number for the year in which the second-half spend is supposed to pay back. The consensus figures put to management, £18.2bn of operating expenses and a 55.5% cost:income ratio, were neither confirmed nor rejected.
2. Where the £500m of structural cost actions is being spent. Management said it should not be thought of as skewed to the UK and that it relates to platforms, processes and change agility across the organisation. Asked for the divisional split, the answer was "we will come back to you on the split, as we have that detail," which means the bank is committing to spend before it has finished allocating it.
3. The cost decision does not appear in the filed document. The single item that moved the stock 4.8% is absent from the Results Announcement furnished to the SEC. The terms "structural cost actions" and "Material Risk Takers" do not appear anywhere in it. The £450m exists only in the presentation and the spoken remarks, which is the same disclosure asymmetry we flagged in April on private credit exposure.
4. The securitised products concentration statistic, still. One quarter after a £228m single-name fraud loss, the filing contains no concentration disclosure for the book that produced it, and no analyst asked. Investment Bank impairment normalised to £44m at a 13 basis point loan loss rate, which is reassuring about the quarter and says nothing about the tail. An unasked question is not an answered one.
5. The maturing hedge yield beyond 2028. Asked directly, management said it has not given that detail and will do so "at some point." Given that the entire outer-year income argument rests on the gap between maturing yields and reinvestment rates, and that the disclosed maturing yield rises from 1.5% to 2.7% across the plan, the shape of that curve after 2028 is the single most useful number the bank could publish and has not.
6. What the 2026 cost:income guide would have been without the second-half spend. Income guidance rose £500m and the cost:income target did not move. The implied improvement was absorbed rather than banked, and the size of what was absorbed was never stated as such.
7. Samsung partnership economics. Declined, consistent with the bank's practice of not sizing individual partnerships. Reasonable, and it means the most forward-looking item in the US Consumer Bank cannot be modelled by anyone outside the company.
8. Any sensitivity around the £430m motor finance provision, in the quarter a tribunal suspended part of the scheme. The filing states that the legal challenges "will delay and may otherwise affect the implementation of the redress scheme" and that outcomes "remain uncertain." No range, no scenario weighting, no per-claim sensitivity, two quarters after moving to a single-scenario basis.
9. The loss-day count. Management again cited stable value at risk and no trading loss days as evidence of Investment Bank discipline. The half-year filing does now carry the management value-at-risk table, which is an improvement on the first quarter, but it carries no daily trading revenue distribution and no loss-day count. Half of the claim is verifiable and half is not.

Market Reaction

• Pre-print setup. The ADS closed at $28.56 on 27 July and the ordinary shares at 530.40p in London, both at the top of the 52-week closing range of $19.10 to $28.56. Entering the print the ADS was up 12.2% year to date against the S&P 500 up 8.3%, up 47.1% over twelve months and up 6.9% over the trailing thirty days. The stock went into the print at its high, on a twelve-month advance of nearly fifty per cent.
• Pre-market. The ADS traded down 7.6% to $26.40 before the US open against the prior close of $28.56.
• Print session (28 July). Barclays reports before the London open, so the reaction session is the print day on both legs. The ordinary shares opened at 508.00p for a 4.2% gap down and closed at 505.00p, down 4.79%. The ADS opened at $26.50 for a 7.2% gap, traded as low as $26.15 for an 8.4% decline, and closed at $26.98, down 5.5%, recovering roughly a third of the pre-market move through the session.
• Volume. ADS volume of 12.0m shares was 2.4 times the 30-day average of 5.1m. London volume of 85.1m shares was well above the prior session's 51.3m.
• Peers and index. Barclays was the largest faller in the FTSE 100 on a day the index closed up 0.83%. Every UK peer fell, none by half as much: Lloyds down 1.44%, NatWest down 2.35%, Standard Chartered down 1.83%, HSBC down 0.18%.

The shape of the move tells you what was traded. A 7.6% pre-market decline that recovers to 5.5% by the close, on a stock that entered the print at a 52-week high after a 47% twelve-month run, is a positioning unwind meeting a genuine piece of bad news. The genuine piece of bad news was £450m, and it was disclosed in the prepared remarks rather than in the release, which meant the first pre-market prints were made by readers of a headline that said profit up 31% and income guidance raised, and the subsequent selling came from listeners to the call.

The relative move is the more useful signal. Every UK bank fell into a rising index that session, so there was a sector element, but Barclays fell more than twice as far as the worst of its peers. This was company-specific, and the company-specific fact of the day was the cost line. It is the mirror image of the Q1 reaction, when Barclays closed roughly flat on a session that saw every UK peer finish higher, and the market chose to trade the fraud charge rather than the hedge disclosure. In both quarters the market has traded the cost or charge line and left the income disclosure alone.

There is also a mechanical contributor worth isolating. A widely syndicated wire item headlined the print as an EPS miss by comparing per-ordinary-share earnings against a per-ADS consensus. Anyone screening on that headline in the pre-market saw a large miss where the like-for-like comparison was a marginal beat. That is not the reason the stock fell 4.8% in London, where the ADS ratio is irrelevant, but it is a plausible reason the ADS gapped 7.2% against the London open's 4.2%.

Street Perspective

Debate: Is the Second-Half Cost a Signal of Discipline or of Drift?

Bull view: A bank that earns 16% on tangible equity and chooses to spend part of an outperformance on permanently lowering its cost base, while simultaneously raising distributions 61%, is doing exactly what long-term owners should want. The return on investment is claimed at 100% within twelve months, the spend is self-funded from a first half that beat, and the 2028 targets are now backed by more conviction than they were.

Bear view: Every income upgrade this bank has produced has been immediately reinvested, the cost:income guide has not improved despite two income raises, and management will not put a number on the year in which the payback is supposed to appear. A 100% return on investment that cannot be verified until 2027 is an assertion, and the pattern of spending upside is how banks fail to deliver operating leverage for a decade.

Our take: The bull side has the better of it on the specifics and the bear side has the better of it on the burden of proof. The spend is bounded by an execution-capacity argument rather than an opportunity-set argument, which is the more credible constraint, and the finance director reaffirmed the Barclays UK absolute cost commitment when directly challenged. But the refusal to anchor 2027 is a self-inflicted wound. Give the market a number and the debate ends.

Debate: Has the Investment Bank Peaked?

Bull view: Nine consecutive quarters of positive operating jaws, income over average risk-weighted assets up a full point to 7.7% on a flat capital base, and a structural shift toward equities and financing that reduces the earnings' cyclicality. Value at risk was flat while first-half income grew 11%, so this is franchise improvement rather than risk-taking, and the division is now earning 16% against a 12% target.

Bear view: Equity capital markets income up 94%, advisory up 53% and equities up 45% in a single quarter is the definition of a wallet peak, and Barclays' banking fees still lag the US bulge bracket in absolute share. Management repeated a c.12% divisional return target rather than raising it, which is what a management team does when it expects the second half to be worse.

Our take: Both are right and they are not in conflict. The cycle will turn and the structural improvement will remain, because the capital base has been flat through the cycle and the mix shift is durable. The number that matters is not this quarter's 16% but the 7.7% income over risk-weighted assets, which is a productivity measure rather than a wallet measure, and which has risen in nine of the last ten quarters.

Debate: Does 1.19 Times Tangible Book Still Leave Anything on the Table?

Bull view: At 505p against 423p of tangible book the market is pricing roughly 12% of sustainable return on tangible equity. The bank delivered 14.2% in the first half excluding the disposal gain, targets more than 14% by 2028, and is compounding tangible book at 10% a year while distributing at a mid-single-digit yield. Closing even half the gap between 12% priced and 14% delivered is worth well over twenty per cent.

Bear view: The stock is up 47% in twelve months and has re-rated from 1.05 to 1.19 times book in a single quarter. The easy money was made between April and July. From here the multiple has to expand into a 2027 regulatory capital inflation of £19–26bn with an unsized Pillar 2A offset, and buybacks above book are now mildly dilutive to book per share rather than accretive.

Our take: The bear case is correct about what has changed and wrong about what it means. The character of the position has shifted from a valuation trade to a delivery trade, and the margin of safety is materially thinner than it was in April. But a bank contracted to receive £19.5bn of gross hedge income over three years, growing tangible book at 10%, and returning cash at an announced pace above 6% annualised, does not need multiple expansion to beat the index. It needs the multiple to hold.

Debate: Is the US Consumer Bank Story Intact?

Bull view: The franchise metrics are all ahead of plan. Core deposits are 82% of funding against a 75%-by-2026 target, retail deposits grew 16%, net interest margin reached 13.2% on the way to a greater-than-13% full-year guide, delinquencies fell when they were expected to rise, and the cost:income ratio is the best in the group. The income softness is a rate-driven pricing effect in one line item and the bank has the balance sheet to warehouse rather than sell.

Bear view: Adjusted attributable profit was £87m, precisely what it was a year ago, on a return of 10.5% against a 12% target that now requires the second half to average around nine. Management withdrew a run-rate it gave one quarter earlier. Best Egg is costing $45m a month and the fee income it was bought to generate is being priced away by the rate environment.

Our take: The bear case owns the next two quarters and the bull case owns the next two years. A division rebuilding its funding stack, its partner mix and its product set simultaneously will produce noisy earnings, and this quarter is what noise looks like. But the specific claim we credited in April, a visible inflection in the reported return, did not hold up in Q2 on a clean basis, and that has to be marked.

Model Update & Valuation Framework

Valuation. At the reaction-session close of 505.00p against tangible net asset value of 423p per share, Barclays trades at 1.19 times tangible book, up from 1.05 times at the equivalent point after the first-quarter print. On the standard dividend-discount framing for a bank, where the price to tangible book multiple equals sustainable return on tangible equity less growth divided by cost of equity less growth, a 10.5% cost of equity and 3% long-term growth imply the market is now pricing a sustainable return of about 12.0%, against 10.9% in April. The bank delivered 14.2% in the first half excluding the disposal gain and targets more than 14% by 2028.

We hold the fair-value range at 1.20 to 1.35 times tangible book, unchanged from initiation, because nothing this quarter changes the discount we apply for the 2027 capital inflation and the unresolved structured-finance disclosure. On the current 423p of tangible book that gives 508p to 571p, or 0.5% to 13% above the 505.00p close, which is a much thinner spot cushion than the 14% to 28% available in April. Tangible book per share grew 10% year on year, and on the more conservative assumption that it compounds at roughly the retained-earnings rate net of the announced distribution, a mid-2027 tangible book near 455p on the same multiples gives 546p to 614p, or 8% to 22%. Translated at the 28 July sterling rate of 1.32886 and four ordinary shares per ADS, the spot range is $26.98 to $30.35 against an ADS close of $26.98, meaning the low end of our range is exactly where the stock already trades, and the forward range is $29.02 to $32.65.

Valuation impact: The distribution does the waiting. Period-end shares of 13,507m at 505.00p give a market capitalisation of roughly £68.2bn, against which first-half distributions of £2.3bn are 3.4% for the half, or an announced pace above 6.7% annualised. The £2bn planned 2026 dividend alone is 2.9%. An investor is paid to hold this while the 2027 cost payback either appears or does not, and the point at which we would change the rating is a 2027 that arrives with costs at the 2026 level and no ratio improvement.

Thesis Scorecard Post-Earnings

Overall: Strengthened on delivery, weakened on price. Three of five bull pillars were confirmed outright and one of them passed the specific falsification test we set at initiation. Two bull pillars moved to at risk, one because the US Consumer Bank inflection did not repeat on a clean basis and one because the shares re-rated 14 points of tangible book in a single quarter. Two bear points improved, one worsened to materialising, and one new bear point was established.

Action: Hold the position, do not add at 1.19 times tangible book. The reason to own Barclays in April was that the market was paying for a bank earning 10.9% while it earned 13.5% with two charges in the number. The reason to own it now is narrower and more demanding: that a management team which just spent its own upgrade will produce the 2027 cost step it declined to quantify. That is a worse setup and still a good one.

Bottom Line

Barclays printed the best quarter of its plan. Return on tangible equity of 16.1%, pre-tax profit up 31%, earnings per share up 43%, all three UK divisions above 20%, the Investment Bank at 16% against a 12% target, the income guide raised for the second time this year, the quarterly buyback doubled and first-half distributions up 61%. Then it told the market it would spend up to £500m on structural cost actions in the second half and another £100–150m on senior pay mix, left the 2026 cost:income guide exactly where it was, and declined three times to say what 2027 costs. The shares fell 4.79% in London into an index that rose 0.83%, while every peer fell by less than half as much.

That is the whole argument, and both sides of it are legitimate. The bank has taken a £500m income upgrade and converted it into 2028 confidence rather than 2026 earnings. Management said as much: the chief executive said the plans "increase our confidence in delivering 2028 RoTE greater than 14%," and the finance director, asked whether additional investment simply means more confidence, answered "Yes, increased confidence." An owner with a three-year horizon should want exactly this. An owner marking a book to a 2026 consensus should not, and the market that traded this print was the second one.

Underneath the argument the thesis did what it was supposed to do. In April we wrote that the most consequential number in the release was the £18.3bn of contracted gross structural hedge income across 2026 to 2028, that the plan still ran on a 3.5% reinvestment assumption against a 3.9% reality, and that if outer-year consensus did not move within two quarters, the thesis was wrong in public. Three months later the contracted figure is £19.5bn, the achieved reinvestment rate is 4.3%, and the finance director confirmed on the call that consensus for 2028 has moved up in both the divisions she named. That test is passed.

The honest marks against us are two. The US Consumer Bank inflection we credited on an 18.8% first-quarter return came back at 10.5% on a clean basis, with adjusted attributable profit of £87m identical to the prior year, and management withdrew the run-rate it had given us. And the £6bn increase in hedge notional we highlighted in April turned out to be a reallocation from product margin rather than incremental income, which management explained on this call. Neither breaks the case. Both are things we said and now have to unsay.

We maintain Outperform at conviction 6. The stock has done 18% since we initiated and the multiple has moved from 1.05 to 1.19 times tangible book, so the cushion that made the April call comfortable is largely gone. What is left is a bank compounding tangible book at 10%, distributing at an announced pace above 6.7% annualised, holding capital above the top of its own range, and sitting on £19.5bn of contracted hedge income into a 2028 target it has just spent its own upside to make more likely. The market sold the invoice. We are still buying the asset.