Net Debt Down $10.9B and the Buyback Restored to $4.2B: Shell Clears the Test We Set It

Results vs. Consensus

Q2 2026 Scorecard

¹ Each SHEL American Depositary Share represents two Shell plc ordinary shares. The company reports Adjusted Earnings per share of $1.76 on an ordinary-share basis; the ADS figure is the mechanical translation, shown because vendor EPS estimates are quoted on the ADS basis. Adjusted-earnings estimates clustered in an $8.79–8.92B range going into the print; the scorecard uses the independently compiled figure. Guided ranges in the third column are the ones Shell issued with its Q1 2026 results on May 7, 2026.

Year-Over-Year Comparison

Quarter-Over-Quarter Comparison

Assessment: Earnings

The headline is emphatic and, unusually for this company in this cycle, the composition mostly holds up. Adjusted earnings of $9,836M beat the compiled consensus by 11.9% and more than doubled against a year ago. Income attributable to shareholders of $10,821M is the largest quarterly figure Shell has produced since the second quarter of 2022.

What separates this print from the Q1 print is where the incremental dollars came from. In Q1 the entire sequential swing traced to trading, optimisation and margin, and we said so. This quarter Upstream contributed $1,108M of the sequential increase on realised prices and a Brazilian production record, Chemicals and Products contributed $952M with a record refinery utilisation number underneath it, and Integrated Gas contributed $872M while producing 31% less gas. That last combination is the single most informative fact in the release: the gas business earned more money on a third less molecule, because the portfolio absorbed the Qatari loss through Nigeria, Trinidad and a fully ramped LNG Canada, and because third-party volumes were bought in to keep customers supplied.

The caveat has not gone away, it has changed shape. Management now says explicitly that Trading and Supply is running at the top of the 2% to 4% ROACE uplift band it has previously described. That is a helpful disclosure and an uncomfortable one, because a business running at the top of its own stated range has, by construction, more room to fall than to rise. A reader annualising $9.8B is making the same mistake a reader annualising $6.9B made three months ago, in the same direction and with more conviction.

Assessment: Cash and the Balance Sheet

Three months ago the cash statement was the reason the stock fell 3.4% on an 8.7% beat. This quarter it is the reason the stock rose 2.5% on an 11.9% beat, and the reversal is not subtle. Net debt fell $10,852M in a single quarter to $41,754M. Gearing fell 450bp to 18.7%, back below where it stood a year ago. Net debt excluding lease liabilities, which is the figure the CFO chose to lead with on the call, fell from $22,012M to $12,127M.

The bridge is clean: free cash flow of $17,524M against $3,001M of cash share repurchases, $2,164M of dividends to Shell shareholders, $29M to non-controlling interests and $1,167M of interest paid. Total debt fell $2,569M and cash rose $8,257M. Lease liabilities, the item that added $3.2B to net debt in Q1 through an IFRS 16 remeasurement, fell $967M to $29,627M, so part of that mark has already come back.

The discipline required here is to separate the durable from the timing. $3,446M of the operating cash flow is working capital coming back, and $1,288M is emission-certificate and biofuel payment timing. Neither recurs. What does carry forward is the balance sheet position itself: $12.1B of ex-lease net debt against $76.9B of annualised first-half adjusted EBITDA is 0.16x, and that is a company with the capacity to fund ARC, complete a $4.2B buyback and absorb a materially worse macro without touching the distribution.

Assessment: The Working-Capital Commitment, Graded

This was the specific thing we said we would grade, so it gets its own paragraph rather than a footnote. Last quarter Shell built $11,179M of working capital and management committed to a reversal of "a significant amount" over an unspecified period. This quarter $3,446M came back. That is 30.8% of the build, and the first half of 2026 remains a $7,733M working-capital outflow.

Partial credit is the honest grade. The direction is right, the mechanism is confirmed, and the reversal arrived alongside the inventory and receivable balances falling from their Q1 peaks. But management gave no timeline for the remaining $7.7B, was not asked for one, and volunteered nothing. The reason this matters less than it did in May is that the reversal is no longer load-bearing: with $17.5B of free cash flow in the quarter and a $4.2B buyback funded, Shell no longer needs the working capital back in order to distribute. It would simply be additive. That is the difference between a timing risk and a solvency question, and this quarter moved it firmly into the former.

Segment Performance

Segment Adjusted Earnings are stated including non-controlling interest; the group figure excludes it. The group total is the income-statement figure, not the sum of the segment column. Chemicals and Products was $118M in Q2 2025 and Renewables and Energy Solutions was $(9)M, so year-over-year percentages for those two are not meaningful.

Integrated Gas

Adjusted earnings of $2,691M rose 47.9% sequentially and 54.9% year-over-year while total production fell 31% to 631 kboe/d, entirely because of the Middle East conflict's effect on Qatari volumes. The arithmetic the company disclosed is that higher trading, optimisation and realised prices added $1,359M and lower volumes subtracted $907M. LNG liquefaction of 7.73Mt exceeded the top of the guided 6.8–7.4Mt range, helped by LNG Canada reaching full capacity and passing 100 cargoes shipped in its first year, and by strong Australian performance offsetting the Qatari and planned-maintenance losses. LNG sales volumes of 17.96Mt were 6.3% lower sequentially but 1.1% higher than a year ago.

"So what we saw, whilst we were really struggling, having lost the Middle East volumes, we were able to compensate from elsewhere. On top of that, what the team did really well was almost record volumes from third party."
— Sinead Gorman, Chief Financial Officer

Segment operating cash flow of $4,629M recovered from $483M in Q1, helped by $883M of working-capital inflow. Cash capital expenditure rose to $1,269M from $1,014M.

Assessment: this is the quarter that proves the portfolio argument Shell has been making for three years, and it is the strongest single piece of evidence in the release. A third of the gas went away and the earnings went up by half, because the marketing and trading layer bought third-party cargoes and redirected the book. The catch is that the offsetting mechanism, buying volumes in the market to serve contracted customers, works precisely when prices are dislocated and works less well when they normalise.

Upstream

Adjusted earnings of $3,485M more than doubled year-over-year and rose 46.6% sequentially, the largest absolute contribution of any segment. Production of 1,824 kboe/d finished 4 kboe/d above the top of the guided range, with liquids at 1,367 kb/d up from 1,346 kb/d in Q1 and Brazil setting another quarterly production record. The company attributed a $1,134M increase to higher realised prices, offset by $242M of oil export levies in Brazil, with a further $317M of favourable tax movements.

"The old adage of big fields get bigger, of course, applies in the context of the Tupi fields, the Iracema fields, the Mero fields."
— Wael Sawan, Chief Executive Officer

Cash capital expenditure fell to $1,633M from $2,159M sequentially and $2,826M a year ago, a 42% year-over-year reduction that is not obviously sustainable alongside a portfolio the company describes as growth-restored. Segment operating cash flow of $6,835M was struck after $2,061M of tax payments, the largest tax outflow of any segment.

Assessment: Upstream is doing exactly what a price-levered segment should do in a $104 Brent quarter, and the incremental information is in the volume line rather than the earnings line. Beating the top of the guided range while running higher planned maintenance, with Brazil at a record, is the kind of result that makes the 4% production growth rate through 2030 look underwritten rather than aspirational. The capex reduction is the item to watch: two consecutive quarters of falling Upstream spend against rising production is a good outcome and also a depleting one.

Chemicals and Products

Adjusted earnings of $2,877M compare with $1,925M in Q1 and $118M a year ago, and split $2,523M Products and $354M Chemicals. Refinery utilisation of 102% is a company record and sits three points above the top of its guided range, with processing intake of 1,267 kb/d up 3.9% sequentially. Chemicals plant utilisation of 83% was 200bp lower sequentially on higher maintenance but 600bp above the year-ago half-year rate. The company attributed a $454M increase to chemicals margins and $429M to products margins, both driven mainly by higher trading and optimisation, against a $156M increase in depreciation.

"Margins, you know as well as I do, how strong those have been this quarter. And of course, those helped significantly. But what the weighting is much more towards the fact of the cost takeout that we've managed and the operating capability of the assets."
— Sinead Gorman, Chief Financial Officer

The segment generated $7,941M of operating cash flow after a $2,308M outflow in Q1, of which $2,185M was working-capital inflow. That is a $10.2B sequential swing in one segment's cash generation, and it is the largest single component of the group's cash recovery.

Assessment: the Chemicals number is small and the trajectory is what matters. Three quarters ago this business was the reason the bear case existed; Monaca has now run at its best-ever rate, the Singapore divestment is behind it, and the CFO's answer on the margin-versus-self-help split weighted the self-help side. Products at $2,523M in a $24 per barrel indicative refining margin environment is a cyclical peak and should be treated as such. The durable read is the utilisation, which is an operating outcome the company controls, not the crack, which it does not.

Marketing

Adjusted earnings of $1,329M were flat sequentially and 10.8% higher year-over-year, and the flatness conceals a deterioration. Margins fell $268M sequentially, on lower trading and optimisation and lower Lubricants margins from both volume and unit-margin pressure, and were offset almost exactly by $288M of comparatively favourable tax movements. Sales volumes of 2,570 kb/d fell 2.2% sequentially and 8.6% year-over-year, which the company attributed to market impacts from the Middle East conflict. The Jiffy Lube divestment closed on June 30 for $1.3B and produced most of the segment's $282M of post-tax divestment gains.

Assessment: Marketing is the one segment where the reported line flatters the underlying, and the offset is a tax movement rather than an operating one. Mobility unit margins improved and Lubricants did not, and Lubricants faces a specific known headwind next quarter because it depends on Pearl GTL base-oil volumes that will not arrive. A flat print with a $268M margin decline underneath it is a warning that the downstream-marketing recovery narrative has a soft edge.

Renewables and Energy Solutions

The only segment moving backwards. Adjusted earnings of $79M fell from $348M sequentially on a $265M decline in margins, mainly from trading and optimisation. Operating cash flow swung to an outflow of $65M from a $2,937M inflow in Q1, driven by $1,025M of net derivative outflows. Identified items included $536M of post-tax impairment charges related to renewable generation assets in Asia and Europe, and the segment absorbed $429M of cash capital expenditure, more than Marketing.

The company's own disclosure is unusually blunt: "Most Renewables and Energy Solutions activities were loss-making in the second quarter 2026, these were more than offset by positive Adjusted Earnings from trading and optimisation and energy marketing."

"Remember, some of that capital today is sitting unproductively because we are still building up, take CCS, for example, take Holland Hydrogen I in Rotterdam. So this is capital that will start to show a return likely in 2027 onwards. As I've said in the past, we will expect a return on that part of the business to be north of 10% before the end of the decade."
— Wael Sawan, Chief Executive Officer

Assessment: a segment whose entire positive earnings contribution is trading, which took $536M of impairments on the assets it does own, and which still consumed $429M of capital in the quarter. The Sprng Energy divestment for $1.8B and the impairments together read as a portfolio being cut back rather than built out, which is the right decision and an expensive one. The 10% return promise is dated to before the end of the decade, which is a long way from a business that will show a return "likely in 2027 onwards."

Corporate

A net expense of $617M, improved from $908M in Q1 and inside the $600–800M guided range, though 33% worse than the $463M a year ago. The improvement came from $250M of favourable net interest movements plus $94M of favourable tax. Interest expense fell to $1,114M from $1,473M.

Assessment: the deleveraging is already visible in the interest line, and this is the most mechanical, most forecastable good news in the release. Guidance for Q3 is $500–700M, a further improvement at the midpoint. If net debt holds near $42B, Corporate becomes a roughly $2.4B annual drag rather than the $3B-plus run-rate implied by the Q1 exit.

Operating Metrics and Guidance Delivery

Shell issues seven quantified operating ranges alongside each quarter's results. Grading the Q2 outcome against the ranges issued on May 7 is the cleanest available test of whether the guidance credibility we flagged at initiation is a durable feature or a single good quarter.

Four of seven inside the range, three above the top end, none below. Combined with Q1, where all nine ranges the company set were met and two were met at the ceiling, that is two consecutive quarters without a single downside miss on a quantified operating commitment.

The structural cost reduction figure for Q1 2026 is the 2025-versus-2022 cumulative disclosed at that date; the company reports the programme on a cumulative basis and disclosed $690M of delivery across the first half of 2026 rather than a quarterly split.

Key Topics & Management Commentary

Overall Management Tone: assured to the point of being relaxed, and noticeably less pre-emptive than last quarter. Three months ago management led with the balance sheet before anyone asked; this time the prepared remarks ran barely six minutes and left the numbers to speak, and the Q&A ranged across exploration, M&A appetite and reserve life rather than circling the distribution. Where the call was least convincing was on disclosure rather than performance: the trading contribution, the forward capex envelope and the chemicals-exit process were each left where they were, and management was pressed on the first of those and deferred it by nine months.

1. The Buyback Is Back, and Larger Than the Run-Rate

The single most consequential announcement in the release. Shell suspended its $3.0B Q1 programme partway through because of securities-law restrictions around the ARC shareholder process, completing $1.8B of it. The new programme announced with these results is $3.000B of new buybacks plus $1.232B of catch-up from the suspended programme, $4.232B in total, to be executed under London and Netherlands contracts running to October 23 and completed before the Q3 results announcement.

At initiation we set an explicit downgrade trigger: a buyback below $3.0B at this print without a working-capital reversal of at least $5B would move us to Hold. The announced programme is 41% above that floor, and by external count this is the nineteenth consecutive quarter in which Shell has committed at least $3B to repurchases.

"And today, we have announced $3 billion of share buybacks, which we expect to complete by our Q3 results announcement in October. In addition to this new program, we will also complete the portion of the previous buyback program that was halted due to regulatory restrictions associated with the ARC transaction."
— Sinead Gorman, Chief Financial Officer

Assessment: the balance-sheet question that capped our conviction at 6 is answered for now. The buyback was never about affordability, as the CFO said twice on the call, but affordability is what the market was pricing in May and the market was wrong. The catch-up structure is also a small piece of good governance: the suspended dollars were not quietly abandoned, they were rescheduled.

2. $17.5B of Free Cash Flow, and Where It Actually Came From

Free cash flow of $17,524M is a figure that invites over-reading, so the decomposition matters. Adjusted EBITDA of $20,710M is the base. Working capital added $3,446M and emission-certificate and biofuel payment timing added a further $1,288M. Tax payments took out $2,934M and cash capital expenditure $4,237M. Operating cash flow excluding working-capital movements was $17,987M, up 4.3% from Q1's $17,241M.

"Of course, it shows up quite significantly when you're sitting on free cash flow this quarter of some $17 billion in 1 quarter."
— Sinead Gorman, Chief Financial Officer

Assessment: a genuinely excellent cash quarter that is roughly $4.7B better than its own run-rate on timing items alone. The right way to hold this is that Shell's underlying quarterly cash generation, before working capital, is now in the high $17B range against $12.3B a year ago, and that is the number worth carrying into a model. The $21.4B is not.

3. The Working-Capital Reversal Is Real and Only 31% Complete

The specific commitment from the Q1 call was a reversal of a significant proportion of the $11,179M build, described as majority price-related and therefore self-reversing, with no timeline attached. $3,446M came back this quarter, made up of a $3,739M inventory release and a $1,593M receivables release against a $1,887M reduction in payables. Inventories fell to $26,639M and current trade and other receivables to $52,938M, both down from their Q1 peaks.

No analyst asked for a timeline on the remainder and management did not offer one. The nearest thing to forward commentary came on cash tax timing rather than working capital.

Assessment: the mechanism is confirmed and the pace is slower than the language implied. Roughly $7.7B remains outstanding across the first half. The reason we are not marking this as a failure is that its purpose in the thesis has changed: in May the reversal was the funding source for the buyback, and this quarter the buyback funded itself three times over from operations. The remaining $7.7B is now upside to distributions rather than a precondition for them.

4. Integrated Gas Earned More on a Third Less Gas

Integrated Gas produced 631 kboe/d against 909 kboe/d in Q1, a 31% decline entirely attributable to the loss of Qatari volumes, and delivered adjusted earnings of $2,691M against $1,819M. Liquefaction of 7.73Mt beat the top of the guided range. LNG Canada reached full capacity and passed 100 cargoes in its first year of operation, and Nigeria and Trinidad supplied incremental volumes.

"And when you have an Integrated Gas, for example, Qatari volumes out and you're still getting roughly the same LNG output, it just speaks to the rigor with which the organization is pursuing that performance drive."
— Wael Sawan, Chief Executive Officer

The mechanism management described has three parts: portfolio substitution from other producing assets, near-record third-party purchases to keep contracted customers supplied, and price risk management through the quarter's volatility.

Assessment: the best evidence yet for the integrated-portfolio argument, and the most quotable result in the quarter. It also complicates the bear case in an interesting way. The offset was not a hedge in any formal sense, it was a trading and procurement operation, so investors who discount the trading contribution are also discounting the mechanism that just absorbed a geopolitical shutdown.

5. Qatar and Pearl GTL: Restart Timelines, Finally Dated

Management separated three assets for the first time. Pearl GTL Train 1 was never damaged and could restart "in a matter of weeks," constrained only by the ability to ship product out. Pearl GTL Train 2, damaged in the March attack, is now dated. Qatar LNG is described as straightforward to restart, subject to terminal capacity, storage, shipping availability and safe passage.

"The one that has been damaged, that second train, we expect the repairs, which are now progressing to be completed and for that facility to be ready to go, again, subject to our ability to export by end of the first quarter of next year. So by end of Q1 2027 is when we could expect that facility to be back online, subject to the conditions allowing us to ship out."
— Wael Sawan, Chief Executive Officer

The Q3 outlook explicitly "excludes any volumes from ARC Resources Ltd. and Qatar," so none of this optionality sits inside the guided ranges.

Assessment: the Q1 estimate of "around a year" for Train 2 has been converted into a date and it is consistent, not slipped. More useful for the near term is that Train 1 and the Qatari LNG train are both physically ready and excluded from guidance, which means a resolution of the shipping constraint is pure upside to the Q3 and Q4 numbers. The exposure is now well enough characterised that it should stop functioning as an unquantifiable tail risk in the valuation.

6. Chemicals and Products Delivers Its Best Quarter in Five Years

Adjusted earnings of $2,877M, refinery utilisation of 102%, chemicals utilisation of 83%, and a positive free cash flow contribution from Chemicals specifically. The CFO framed the transformation as three variables and weighted them.

"The hard work the team is putting into the transformation is starting to pay off, and combined with a more favorable margin environment this quarter's results represents the best we have seen in over 5 years, but there is much more to do."
— Sinead Gorman, Chief Financial Officer

On the refining side, management described a structural change in how run rates are set, with traders embedded alongside operators at Norco in the US and the model being rolled out across the refining system, shifting production toward middle distillates such as jet fuel in response to price signals.

"And so much more of what we see at the moment is that the run rate is being determined by commercial factors driven by our trading organization in partnership with the asset."
— Wael Sawan, Chief Executive Officer

Assessment: two different claims are bundled here and they deserve different weights. The utilisation and cost story is durable and is the reason this segment can now earn through a downcycle. The margin story is cyclical, with the indicative refining margin at $24 per barrel against $17 in Q1, and it will unwind. Management's own Q3 commentary already flags softening chemical spreads. Model the utilisation, not the crack.

7. The Cost Programme Is Inside Its Target Band With Two and a Half Years to Run

Structural cost reductions reached $5,825M against 2022 levels, with $690M delivered across the first half of 2026. The $5–7B target is measured as annualised saving achieved by end-2028, so with two and a half years to run Shell is $825M past the bottom of the range and $1,175M short of the top. Underlying operating expenses of $17,026M in the first half rose $427M year-over-year, with $1,117M of inflation, FX and activity increases offset by the $690M of structural reduction.

"We are now halfway through that band that we talked about... There's more and more opportunities than maybe we had banked for. And so my push to the team now is we need to be able to get to the top end of this range, and that's what we're working towards."
— Wael Sawan, Chief Executive Officer

Assessment: the most macro-independent line in the release and it is running ahead of schedule. The more interesting signal is the CEO framing the programme as a culture change rather than a target, and openly discussing what comes after it. A company inside its cost band with two and a half years left, whose chief executive is already asking what the next benchmark is, is not a company about to declare victory at the bottom of the range.

8. Trading Is Running at the Top of Its Own Range, by Management's Admission

Shell has long described Trading and Supply as contributing a 2% to 4% uplift to group ROACE without disclosing an earnings figure. This quarter management placed the business within that band for the first time in recent memory, and placed it at the top.

"And so as we look to the coming quarters, what I can tell you is the 2% to 4% ROACE that Trading and Supply is able to deliver continues to hold. And as you would expect, we are at the top end of that range given the current volatility."
— Wael Sawan, Chief Executive Officer

The company's own segment commentary attributes the Integrated Gas improvement to trading and optimisation combined with realised prices, the Chemicals and Products margin increases to trading and optimisation, and the Renewables and Energy Solutions decline to a fall in trading margins. No segment-level or group-level trading contribution is disclosed.

Assessment: this is the bear point and it is now on the record from management rather than inferred by us. A business at the top of its stated 2% to 4% band mean-reverts downward by definition. Shell's group ROACE was 12.4% this quarter; if trading contributes 4 points at the top of the range and 2 at the bottom, the normalisation is worth roughly 200bp of group ROACE, and nobody outside the company can size the earnings equivalent because the company does not publish it.

9. ARC Resources: 99.54% Approved, One Regulator Left

ARC shareholders voted in favour on July 14 with approximately 99.54% of votes cast supporting the arrangement. The remaining condition is Investment Canada Act approval, and completion is expected in the third quarter of 2026. The terms are unchanged: CAD 8.20 in cash and 0.40247 Shell ordinary shares per ARC share, an equity value of approximately USD 13.6B struck at Shell's April 24 closing price. Management reiterated that ARC lifts expected production growth to 2030 from around 1% a year to some 4%, and adds roughly $1.5B a year of free cash flow once complete.

"we see line of sight to double-digit returns. But I do expect my teams to aspire to meet mid-double-digit returns if we can and really demonstrate the value that we can create whenever we choose to use paper."
— Wael Sawan, Chief Executive Officer

Management also disclosed that LNG Canada Phase 2 was not included in the base economics for the acquisition and that existing Groundbirch acreage already underwrote Phase 1, so ARC's gas is incremental optionality rather than a requirement.

Assessment: the execution risk has narrowed from a shareholder vote plus regulators to a single regulator in a country where Shell is a large investor. The return language is the item to file away. A management team that publicly commits to double-digit returns on a mostly-paper acquisition, and says it expects its teams to aim higher, has created a measurable standard against which the deal can be graded from 2027.

10. The Divestment Flywheel

Four disposals were announced or completed inside the quarter: Jiffy Lube International for $1.3B, completed June 30 with a retained lubricants supply agreement; the Na Kika working interest plus the Coulomb tieback in the Gulf of America for $1.7B, agreed in June and expected to close by end-2026; Solenergi Power, which includes the Sprng Energy group, for $1.8B agreed in July; and the South African mobility sites. Divestment proceeds recognised in the quarter were $469M, so most of the announced value is still to land in cash.

"But of course, when we did the deal for ARC, we knew that this divestment program was coming, and you can see that it more than offsets in terms of the cash coming through."
— Sinead Gorman, Chief Financial Officer

Management framed the programme against a stated $45B of underperforming capital employed, a portion of which sits in the low-carbon portfolio.

Assessment: $4.8B of announced or completed disposals in one quarter against a $3.4B cash cost for ARC is a self-funding acquisition, and the CFO said as much. The strategic content is that three of the four are exits from businesses Shell explicitly says it is not the natural owner of, which is a different and better argument than selling assets to fund a distribution. The pace is now fast enough that the $45B underperforming-capital figure should start visibly shrinking, and that is a number worth tracking quarterly.

11. The Q3 Setup Management Actually Described

The most useful forward commentary came in response to a question about the July environment rather than in the outlook table. The CFO gave a segment-by-segment read that is more granular than the guidance ranges.

"We're seeing, of course, a positive margin environment for Refining in Q3, but the chemical spreads are beginning to soften. We do see that come through. And we're seeing, of course, less volatility, which means a little bit less coming in, in terms of our Downstream business from the trading angle of things as well."
— Sinead Gorman, Chief Financial Officer

She added that Q2 had unusually few turnarounds, that more are scheduled in Q3, and that Lubricants will be harder because it depends on Pearl volumes that are not expected in the near term. On LNG, the redirection of cargoes from Asia back into a Europe entering winter with storage well below the levels normally expected is described as creating near-term tightness.

Assessment: management pre-flagged a sequentially weaker Q3 in four separate ways without saying so in a headline: softer chemical spreads, less trading contribution from lower volatility, more turnarounds, and a Lubricants hole. Refining margins and LNG tightness are the offsets. Anyone modelling Q3 off this quarter's $9.8B has been told, in detail, not to.

Guidance & Outlook

The Q3 ranges are almost all higher than the equivalent Q2 ranges, which is the correct comparison, and almost all below the Q2 actual, which is the honest one. Integrated Gas is guided to 570–630 kboe/d explicitly excluding both ARC and Qatar, so it assumes the Qatari shutdown persists through the whole quarter and gives no credit for a Pearl Train 1 restart that management says is weeks away once shipping allows. Upstream is guided wider and lower at the midpoint on higher maintenance. Refinery utilisation of 93–101% would be a strong outcome anywhere inside the band given a 102% record just delivered.

On capital expenditure, the CFO restated the full-year envelope and its composition without hedging.

"Our cash CapEx outlook of $24 billion to $26 billion for 2026 is unchanged. This includes some $4 billion for the ARC Resources acquisition and associated cash CapEx."
— Sinead Gorman, Chief Financial Officer

Implied second-half spend: $8,439M was spent in the first half, so the guided range implies $15.6B to $17.6B across the third and fourth quarters, roughly double the first-half rate. Around $4B of that is ARC-related, leaving approximately $11.6B to $13.6B of organic spend against $8.4B in the first half. That is a material second-half step-up and it is the main reason free cash flow will not repeat at anything like $17.5B.

Street at: consensus entering the print sat at $8.79B of adjusted earnings against the $9.84B delivered, and the compiled range was tight at $8.79–8.92B. There is no published consensus for Q3 in the material we would treat as reliable; the more relevant anchor is that the company has now guided seven operating metrics for Q3 and pre-flagged four separate sequential headwinds on the call.

Guidance style: conservative and getting more so. Shell exceeded three of seven Q2 ranges to the upside and missed none, then set Q3 ranges that sit below the Q2 outturn on five of seven metrics while excluding Qatar entirely. Management is setting itself beatable bars, which is a defensible posture in a quarter-to-quarter macro it cannot forecast, and which means the ranges should be read as floors rather than expectations.

Analyst Q&A Highlights

The 40–50% Payout Ratio Against the Value of the Buyback

The opening question of the call, and the one that carried the most thesis weight. The framing put two of Shell's commitments in tension: a through-cycle payout ratio of 40% to 50% of operating cash flow, and a stated conviction that buying back stock at the current price is value-accretive. In a quarter with $21.4B of operating cash flow, a $4.2B distribution run-rate sits well below 40%. Management's answer separated the two and defended the framework rather than the arithmetic, and pushed back on the characterisation of the Q1 buyback reduction.

Q: "The first one is just on the distribution front. And going back to your comments in Q1, you cut the buyback -- yes, trimmed the buyback, let's say, the argument you made was you wanted to be agile and tactical. And I guess, just a view on the value of the buyback in terms of your share price. At the same time, you have a payout ratio and that calculus on the return on the buyback is not really embedded in a 40% to 50% payout ratio. So as we look forward, obviously, the impact of the war is maybe more pronounced than you thought at the time. But it looks like your run rate on distributions will be well below the 40% if you continue at this rate."
— Biraj Borkhataria, RBC Capital Markets

A: "This is definitely not about affordability in any sense. You talked about last quarter specifically and what did we do last quarter? So I wouldn't say we trimmed, I'd say we rebalanced. So as we discussed at the time, we rebalanced between both the buyback and the dividends. So we increased the dividends at the time when we moved the buyback to $3 billion. So that allowed us to stay within that payout ratio. We're very pragmatic on this and not dogmatic at all. We are dogmatic about the 40% to 50% through the cycle. But in terms of how we split it, we make that decision quarter-by-quarter, and we look through the quarter."
— Sinead Gorman, Chief Financial Officer

Assessment: management defended the commitment and declined to defend the quarter, which is the right answer and an incomplete one. "Through the cycle" is doing a great deal of work in a quarter where cash flow tripled and the distribution did not move. The exchange did establish that the Q1 reduction was a mix decision rather than a capacity decision, and this quarter's balance sheet retrospectively supports that claim. A repeat of this exchange with $21B of cash flow and a sub-25% payout would land differently in three months' time.

Whether the $5–7B Cost Target Gets Reset Upward

A question that reframed a risk as an opportunity and drew the most forward-leaning answer of the call. With the programme at $5.8B against a $5–7B band targeted at end-2028, the question was whether the target is about to be raised and by how much.

Q: "And then my follow-up is on the $5 billion to $7 billion cost-cutting target. You're about halfway there, 3 years or 2 years early. So I'm wondering if you could frame for us what the risk is that those numbers get reset and any kind of magnitude you could put around that."
— Douglas Leggate, Wolfe Research

A: "And the challenge, by the way, is not just a structural cost reduction challenge. It is a free cash flow enhancement challenge. That's what we're trying to drive. Improve reliability, improve availability, enhance business models, turn around underperforming businesses and become leaner, more focused as an organization. So that's been embraced. We are now halfway through that band that we talked about. I continue to be encouraged by what I see, Doug. There's more and more opportunities than maybe we had banked for. And so my push to the team now is we need to be able to get to the top end of this range, and that's what we're working towards."
— Wael Sawan, Chief Executive Officer

Assessment: management steered to the top of the existing range rather than announcing a new one, which is the disciplined answer three quarters ahead of a likely capital markets event. The substantive disclosure is the reframing from cost reduction to free cash flow enhancement, which widens the definition of what counts and makes the target easier to hit for reasons other than cost. That is worth noting rather than penalising, because reliability and availability gains show up in the same place cost gains do.

Sustainability of 102% Refinery Utilisation

The most important operational question asked, because the refining result is both the largest single earnings driver in the quarter and the one most likely to be dismissed as a cyclical fluke. The answer separated maintenance discipline from commercial optimisation and put the weight on the latter.

Q: "I just wanted to ask about the really remarkable operational performance in the Downstream, notably the 102% refinery utilization. ... but maybe more conceptually, how much of this high refinery utilization rate is sustainable? I mean, how long can you continue to operate above 100%, just thinking about maintenance cycles, et cetera?"
— Kim Fustier, HSBC

A: "So to give you a small example, at Norco in the U.S., we have moved into a model where the traders are tied at the hip with the operators, finding the right feedstock to be able to source given the dynamics in the market at the moment. And then the products traders finding what's the best placement and reading all the price signals to be able to then manage how much do we push into jet fuel versus -- or at the expense of diesel and gasoline, and how do we optimize for value. And so much more of what we see at the moment is that the run rate is being determined by commercial factors driven by our trading organization in partnership with the asset. We are rolling that model out in every single one of our refineries and have tested it and really been pleased with what we see."
— Wael Sawan, Chief Executive Officer

Assessment: a specific, checkable answer with a named site and a described mechanism, which is more than this question usually gets. The claim that run rates are now set commercially rather than technically is a genuine structural change if it holds, and it is being rolled out rather than already complete. Management explicitly preserved the safety-critical turnaround schedule, which removes the obvious concern that 102% was bought by deferring maintenance.

Chemicals: How Much Was Margin and How Much Was Self-Help

The single most useful question for anyone trying to work out how much of the Chemicals and Products result survives a normalising crack. The answer explicitly weighted the two.

Q: "And then secondly, just on the Chemicals result, which was strong this quarter, very positive to see that. Could you just comment on the relative split of this improvement between the self-help work you've been doing since the Singapore divestment and also the margin environment that we saw in the quarter?"
— Fergus Neve, Rothschild & Co Redburn

A: "Margins, you know as well as I do, how strong those have been this quarter. And of course, those helped significantly. But what the weighting is much more towards the fact of the cost takeout that we've managed and the operating capability of the assets. So what the team did very, very well was to be able to actually ensure that those assets were up and running. So in Pennsylvania at Monaca, they managed to ensure that it actually hit record performance as well."
— Sinead Gorman, Chief Financial Officer

Assessment: management put the greater weight on self-help, which is the answer that supports the transformation case, and did so without producing a number. That is a judgment offered rather than a disclosure made, and it should be treated accordingly. The verifiable half is the utilisation and the Monaca record, both of which are real; the unverifiable half is the split. Chemicals earned $354M in a quarter where the group earned $9,836M, so the segment is still small enough that the argument is about direction, not magnitude.

How Integrated Gas Absorbed the Qatari Volume Loss

A question about whether the portfolio contains a natural hedge, in which lost Middle East volumes are offset by better margins elsewhere. Management rejected the framing and offered a more specific and less flattering mechanism.

Q: "I'd like to ask you first about Integrated Gas, very strong numbers in the second quarter despite the impact of the Qatari assets. I wondered if you could just expand on the extent to which in conditions you're seeing a degree of natural hedge almost within the business insofar as the downtime or lost volume in the Middle East being mitigated by the rest of the portfolio and perhaps stronger margins as a result that you're able to extract through the rest of that portfolio, particularly the third-party component."
— Matthew Lofting, JPMorgan

A: "So whether you call it a natural hedge or not, it was a portfolio management approach. So what we saw was, particularly in Nigeria, we saw more volumes coming out of Nigeria, also from Trinidad, but also as Wael mentioned in the video as well earlier on today, specifically around Canada. So we saw LNG Canada come into its own. We're now more than 100 cargoes out from that facility... On top of that, what the team did really well was almost record volumes from third party. So indeed, they went out into the market. They looked at where they could cover and in some cases, buying back some of our own cargoes that we had sold to them to be able to distribute elsewhere, i.e., taking from those customers who weren't as impacted by the Middle East and being able to push them to those who were."
— Sinead Gorman, Chief Financial Officer

Assessment: the correction of "natural hedge" to "portfolio management approach" is the most honest moment on the call. A natural hedge is structural and repeats; buying back your own cargoes from customers to redistribute them is an active operation that depends on having the balance sheet, the shipping and the trading desk to do it. It worked, and it will work again, but it is an execution capability rather than a property of the asset base, and it should be valued as the former.

Restart Timing for Pearl GTL and Qatar LNG

A short exchange that converted the quarter's largest open operational question into three dated answers.

Q: "So let me ask regarding the Middle East assets, yes, obviously, Qatar, but also Pearl GTL. If a normalized shipping environment comes, could you remind us on the time to get those 2 facilities back to their expected capacities, please?"
— Mark Wilson, Jefferies

A: "The LNG assets are typically easier to start up and to start the shipping out, of course, subject to terminal capacity, subject to storage, subject to shipping availability and the like. The biggest thing we're watching out there for is just access and safe passage through the Straits. Similarly, on one of the trains at Pearl GTL, so Train 2 -- Train 1, where it would actually take in a matter of weeks to be able to get the facility back up and running... So by end of Q1 2027 is when we could expect that facility to be back online, subject to the conditions allowing us to ship out."
— Wael Sawan, Chief Executive Officer

Assessment: management gave the answer with the date attached and repeated the shipping caveat three times, which is the appropriate weighting. The commercially useful content is that two of the three assets are restart-ready and constrained only by the strait, and that none of that potential is in the Q3 guide. The physical damage has a completion date; the geopolitics does not, and the company was careful not to pretend otherwise.

Capex Confidence Against 5–6% Supply-Chain Inflation

A question testing whether the $24–26B envelope survives an inflating cost base, particularly for offshore rigs, in a quarter where several operators have flagged rising day rates.

Q: "And then my next question is on CapEx. How confident are we in maintaining the CapEx guidance this year, please? Especially given the disruption in the Middle East. We have heard companies talking about higher cost for -- higher cost to get the rigs FPSO. I wonder what are you seeing in the market right now?"
— Naisheng Cui, Barclays

A: "So our range is $24 billion to $26 billion. We are confident in our ability to be able to deliver within that range, and we continue to maintain that range at the moment. We absolutely see inflation in the system, which is what you're referring to at the moment. That varies per category. But overall, we're seeing it around that 5% to 6%, but we're able to offset much of that given our scale and those framework agreements we have, but also because we have locked in many things because we saw some of this coming as well. So we have confidence in the ability to do that. You asked specifically about rigs. That's less of a problem for us at the moment because we had locked those in, in advance."
— Sinead Gorman, Chief Financial Officer

Assessment: a direct answer with a quantified inflation figure and a named mitigant, which is what this management team does well. The unaddressed part is that pre-locked rig contracts and framework agreements expire, so this is a defence of 2026 rather than of the years that follow. Nobody asked what the 2027 and 2028 capex envelope now looks like with ARC inside it, and management did not volunteer it.

Whether Trading Deserves Its Own Disclosure

The last question of the call and the most consequential unresolved one. The framing was that Shell's trading is asset-backed and therefore structural rather than speculative, and so should be reported with more granularity than it currently is.

Q: "And is there a case there for providing a bit more of detail within the financial reporting about your trading activities, particularly given that they are properly backed by assets, therefore, it's something more structural rather than what would be a trading activity of a financial institution?"
— Maurizio Carulli, Quilter Cheviot

A: "What we do, of course, is for us, trading is actually much more around being able to optimize around the assets of the various businesses. So it is not a segment in its own right. It is the other businesses that it pulls on for the volumes, et cetera. We allocate capital to those segments rather than specifically to trading as well, and that's why we report it in the manner we do... We've told you as an example, that we've never lost money in any quarter in the last decade, as we said in our Capital Markets Day 2025... We're looking to give you a bit more detail on our next Capital Markets event, which we're hoping will be at some point in the first half of 2027."
— Sinead Gorman, Chief Financial Officer

Assessment: the answer defended the current reporting on a defensible technical basis, offered a reassurance that cannot be independently verified, and deferred the substantive disclosure by roughly three quarters. That is a dodge conducted with unusual politeness. Trading is the single largest unexplained variable in Shell's earnings, management has now confirmed it is running at the top of its stated range, and the company's position is that shareholders should wait until the first half of 2027 to learn more. This remains the clearest gap between what Shell knows and what its owners are told.

What They're NOT Saying

1. The 2027–2028 capital expenditure envelope was never restated. At the Q1 print, management held FY27–28 at $20–22B with ARC absorbed, and we flagged it as the single most falsifiable statement on that call. This quarter the CFO restated only the FY26 range of $24–26B and described the $20–22B as the pre-ARC guidance, without reaffirming it for the out years. Nobody asked. With ARC closing in Q3 and $11.6B to $13.6B of second-half spend implied, the forward envelope is the most consequential number Shell did not repeat.
2. The US chemicals exit has gone silent. Three months ago the language had escalated from an intention to an active process with two named structures, a trade sale or a capital-markets transaction. This quarter the word does not appear in the release or on the call, and it went quiet in exactly the quarter the assets ran at record rates and the segment turned $354M. Either the process is progressing under confidentiality or the improvement has changed management's mind. Both are material and neither was disclosed.
3. The trading contribution is still not sized, and now has a date instead. Management confirmed the 2% to 4% ROACE uplift band and placed the business at the top of it, then deferred any further disclosure to a capital markets event in the first half of 2027. Three of the five operating segments attribute their sequential movement partly to trading and optimisation, and none of them quantify it.
4. The revised long-term commodity price assumptions were not published. The basis-of-preparation note states that future long-term commodity price assumptions "were subject to change in the second quarter 2026" and "remain under review." The new deck is not disclosed, and in Q1 the company explicitly left the assumptions unchanged. A price-deck revision in a quarter with $629M of impairments and Brent at multi-year highs is worth a sentence of explanation that it did not receive.
5. No timeline for the remaining $7.7B of working capital. $3.4B came back and management neither characterised the pace nor committed to a schedule for the rest, and was not pressed on it. The forward cash statement is a great deal easier to model with that number than without it.
6. The LNG Canada midstream sell-down has disappeared. At the Q1 print management confirmed a sell-down of the LNG Canada midstream position was "a consideration." It was not mentioned this quarter, in a call that covered LNG Canada Phase 2, ARC and four separate disposals.
7. No earnings figure for what Qatar and Pearl actually cost. Integrated Gas volumes fell 31% and the segment quantified the volume effect at $907M, but the group has never sized the total Middle East earnings impact across Integrated Gas, Lubricants and Chemicals and Products. The Q3 guide excludes Qatar entirely, which means the cost is being carried forward without ever having been stated.

Market Reaction

• Pre-print setup: the ADR closed at $88.34 on July 29, up 20.2% year-to-date against 6.9% for the S&P 500, up 23.2% over trailing twelve months and up 13.9% over the trailing thirty days. The 52-week closing range entering the print was $70.31 to $94.15, so the shares went in at the upper end of their own year but 6.2% below the high set in April.
• Reaction session: the ADR opened at $90.00, a 1.9% gap up, traded a $89.21 to $90.77 range and closed at $90.51, up 2.5% or $2.17. The S&P 500 rose 1.7% on the same session.
• Volume: 10.8 million ADS against a 30-day average of 6.5 million, 1.6 times normal.
• Comparison to the prior print: the Q1 result on May 7 drew a 3.4% decline on an 8.7% beat, closing at $84.24. This print drew a 2.5% gain on an 11.9% beat, closing at $90.51, a $6.27 advance in the ADR across the two reaction closes.

The move was smaller than the beat, and the reason is that roughly two thirds of it was the market and the sector rather than Shell. The S&P 500 rose 1.7% on a session in which crude also rose on renewed Middle East tension, so the company-specific increment is closer to a point than to two and a half. That is not a criticism of the print; it is a reminder that a large-cap integrated on a beat day in a rising tape is not a clean read on how the disclosure landed.

What the tape did reward is legible. In May the market punished a large earnings beat because free cash flow covered barely half the distribution and net debt rose $6.9B. In July it rewarded a similar-quality earnings beat because free cash flow covered the distribution three times and net debt fell $10.9B. The earnings line was not the variable in either session. The cash line was, in both. That is a market that has correctly identified which number in Shell's release is load-bearing, and it means the working-capital reversal and the buyback trajectory will continue to set the reaction function for the next several quarters regardless of where adjusted earnings prints.

The restraint in the move also carries information. A 2.5% session on the largest quarterly profit since 2022, a restored and enlarged buyback, and a 450bp reduction in gearing is not a market repricing the business. It is a market taking the quarter as evidence rather than as a new run-rate, which is the same conclusion this note reaches from the other direction.

Street Perspective

Debate: Is $9.8B a peak or a new base?

Bull view: the composition is materially better than Q1's. Refinery utilisation of 102%, Upstream production above the top of the guided range, the cost programme past its midpoint, and Integrated Gas earning more on 31% less gas are all operating outcomes rather than price outcomes, and none of them reverse when the crack does. The through-cycle base has stepped up.

Bear view: Brent averaged materially higher and the indicative refining margin rose from $17 to $24 per barrel, trading is at the top of its own stated band, and the effective tax rate fell 690bp. Strip those and the quarter looks like a good but ordinary result on an inflated macro. Management itself flagged softening chemical spreads, less volatility and more turnarounds for Q3.

Our take: the bears are right about the level and the bulls are right about the direction. Our normalised quarterly adjusted earnings estimate moves from $5.5–6.0B to $6.5–7.0B, which credits the cost programme, the utilisation gains and ARC's contribution while assuming the refining and trading windfall unwinds substantially. That is a 15% to 20% increase in the through-cycle base and roughly a 30% discount to the printed quarter, and both halves of that sentence are the point.

Debate: Does the balance sheet now permit a step-change in distributions?

Bull view: gearing of 18.7%, ex-lease net debt of $12.1B, and $17.5B of quarterly free cash flow against a distribution run-rate near $5B leaves the payout ratio well under 30% in the quarter. With $7.7B of working capital still to come back and $4.8B of disposals announced, the capacity for a materially larger buyback is obvious and the company has publicly committed to 40% to 50% through the cycle.

Bear view: the CFO explicitly declined to be pinned to a full-year payout, described the split as a quarter-by-quarter decision and stressed value over affordability. ARC closes in Q3 and the second-half capex step-up is $15.6B to $17.6B. The company has just spent a quarter deleveraging and shows every sign of preferring balance-sheet capacity to a headline distribution number.

Our take: the bear reading of management's intent is correct and the bull reading of the capacity is also correct, which is why this is a debate about timing rather than about the balance sheet. The catch-up structure of this quarter's programme is the tell: Shell chose to make the suspended dollars whole rather than to raise the run-rate, which is a company protecting optionality through the ARC close. Expect the step-up, if it comes, at the Q3 print or the capital markets event, not before.

Debate: Should the trading contribution be discounted or capitalised?

Bull view: trading is not a bolt-on desk, it is the mechanism by which the integrated portfolio converts volatility into earnings, and this quarter proved it by absorbing a geopolitical shutdown. It is capital-light, it has reportedly never lost money in a quarter for a decade, and competitors cannot replicate it without the same asset footprint. It deserves a multiple, not a haircut.

Bear view: an earnings stream that management will not size, that sits at the top of its own disclosed range, and whose disclosure has been deferred to a capital markets event nine months out, cannot be underwritten by an outside investor at any multiple. The asymmetry is entirely downward from here.

Our take: the bear case wins on process and the bull case wins on substance, and the resolution is a discount rather than an exclusion. We treat trading as a real, durable earnings contributor and refuse to capitalise the peak-cycle portion of it, which is precisely what a normalised base of $6.5–7.0B against a printed $9.8B does. The deferral to the first half of 2027 is a genuine governance mark against a management team that has otherwise been unusually specific, and it is the reason our conviction moves to 7 rather than higher.

Model Update & Valuation Framework

Valuation

At the July 30 close of $90.51 the 5,570.9 million ordinary shares outstanding at quarter-end, equivalent to 2,785.5 million ADS, capitalise the equity at approximately $252.1B. Against that:

• Dividend yield 3.45%, on the declared $0.3906 quarterly ordinary-share dividend annualised to $3.1248 per ADS.
• Buyback yield 4.76%, on the $3.0B quarterly new-money run-rate annualised to $12.0B. The $1.232B catch-up adds a further 0.49% over the coming twelve months.
• Combined shareholder yield approximately 8.2% on the run-rate, or 8.7% including the catch-up, against 8.8% at the Q1 close on a share price 7.4% lower.
• Trailing multiple 9.9x, on $25,439M of adjusted earnings over the current and previous three quarters, or $9.13 per ADS.
• Leverage 0.54x net debt to annualised first-half adjusted EBITDA, or 0.16x excluding the $29.6B of lease liabilities.
• ROACE 12.4%, against 9.9% in Q1 and 9.4% a year ago.

Our twelve-month price target moves to $100 from $95, implying 10.5% price appreciation and roughly 13.9% total return including the dividend. Two things changed to justify the raise, and neither is the printed earnings number. First, the balance sheet is a full year ahead of where we modelled it, which removes the scenario in which the distribution has to be cut to fund ARC. Second, the operating record is now two quarters long rather than one, and it contains three upside range breaks rather than a set of narrow in-line results.

The base case assumes the conflict-driven refining and trading windfall substantially unwinds through 2027, the cost programme reaches the upper half of its band, ARC completes in the third quarter and contributes a full year, and the buyback retires roughly 5% of the share count annually against the roughly 4% of equity issued as ARC consideration. The multiple moves to 11.0x from 10.5x, which is the whole of the re-rating we are asking for and is justified by gearing 450bp lower, ex-lease net debt down $9.9B, ROACE 250bp higher and two consecutive quarters without a downside guidance miss. The bear case requires a macro reversal, a stalled working-capital recovery and a trading contribution that reverts to the bottom of its 2% to 4% band simultaneously. The bull case requires LNG Canada Phase 2 to be sanctioned before year-end, the chemicals process to resolve, and the Qatari assets to return on the schedule management described.

Thesis Scorecard Post-Earnings

Overall: strengthened. Two of the three bull pillars were confirmed without qualification for the second consecutive quarter, the third moved from qualified to confirmed, and the balance-sheet bear point was refuted rather than merely deferred. The one pillar that did not improve is the one that matters most for the multiple: the trading contribution is still undisclosed, is now confirmed to be running at the top of its range, and will not be explained for another nine months. Conviction moves from 6 to 7.

Action: hold and add on weakness, with the second-half capital expenditure step-up as the next thing to grade. We are maintaining Outperform and raising the twelve-month target to $100. The new downgrade trigger is a Q3 buyback below $3.0B or a 2027 capital expenditure envelope guided materially above $22B excluding ARC, either of which would convert the free-cash-flow story from a compounding one into a reinvestment one.

Bottom Line

Three months ago Shell beat by 8.7% and fell 3.4%, because the cash statement did not support the distribution and net debt rose $6.9B in a quarter. We initiated at Outperform on the argument that the missing cash was deferred rather than destroyed, and we set a single explicit test: a buyback below $3.0B at this print without a working-capital reversal of at least $5B would take us to Hold.

The test was cleared on the first criterion and missed on the second. The buyback came back at $4.232B, 41% above the floor. The working-capital reversal delivered $3.4B against the $5B threshold, which is 31% of what was built. On the arithmetic of the test that is one for two, and the reason the rating goes up rather than sideways is that the second criterion stopped being load-bearing the moment free cash flow of $17.5B covered the distribution three times over. The reversal was supposed to be how Shell funded the buyback. It turned out not to be needed for that, which is a better outcome than the one we tested for.

The quarter itself deserves less enthusiasm than the balance sheet does. Adjusted earnings of $9,836M were produced with Brent materially higher, the indicative refining margin at $24 per barrel against $17 in Q1, the effective tax rate 690bp lower, and Trading and Supply running at the top of its own stated range. Management pre-flagged four sequential headwinds for Q3 on the call and guided five of seven operating metrics below the Q2 outturn. Anyone annualising $9.8B is making the same error, in the same direction, as anyone who annualised $6.9B in May.

What is durable is narrower and more interesting than the headline. The cost programme is at $5.8B inside a $5–7B band with two and a half years left and the chief executive pushing for the top of the range. Refinery utilisation of 102% came from a trader-operator model being rolled out across the system, not from deferred maintenance. Integrated Gas earned 48% more money on 31% less gas, which is the strongest single piece of evidence for the integrated-portfolio argument this company has produced in years. Net debt excluding leases is $12.1B. The share count is down 6% year-over-year and the dividend is up 9%.

Against that, one thing has not moved and it is the reason conviction is 7 and not 9. Shell will not tell its owners how much of this it earns from trading. Management has now confirmed the business is running at the top of a band it describes but does not quantify, and has scheduled the explanation for the first half of 2027. That is the gap between a company that is executing extremely well and one that can be underwritten with confidence. At $90.51, on a roughly 8.2% shareholder yield and 9.9x trailing earnings, we are willing to own the gap. Maintaining Outperform, target to $100.