Maintaining Hold: A Record Permian Print and an Early Yellowtail Are the Right Operational Story; the Chemical Drag and the Unresolved 45V Question Keep Us From Upgrading

Rating Action: Maintaining Hold

This is the second report in our XOM coverage arc. We initiated at Hold with a constructive bias on Q2 2025 and we are maintaining that rating after Q3. The arc walk:

• Q2 2025 (Initiating at Hold, constructive bias). Hold-anchor rationale: (1) the Stabroek arbitration ruling against XOM let Chevron into the block via the Hess transaction, introducing a non-zero contractual-precedent risk into the long-dated joint-operating-agreement value algorithm; (2) the Baytown low-carbon hydrogen project was put under genuine review after the 45V tax-credit eligibility window for construction start was shortened from 2033 to early 2028, with Woods stating "if we can't see an eventual path to a market-driven business, we won't move forward with the project"; (3) Pioneer integration was tracking $2B → $3B/yr synergies with management telegraphing further upward revision in December; (4) Permian was at 1.6 mboed and the 2030 path to ~2.3 mboed was reaffirmed; (5) Yellowtail first oil was telegraphed as "imminent" but not yet at first oil. The operational story was clean but the equity already priced it; the Stabroek + 45V overhangs argued against an Outperform start.
• Q3 2025 (Today — Maintaining Hold). Two of the five Q2 anchors have now resolved favorably: Yellowtail delivered four months early at full design intent, and the Permian moved from 1.6 to 1.7 mboed with the proprietary-proppant rollout running ahead of schedule. Hammerhead was newly sanctioned; eight of ten flagship 2025 startups are now complete; the lightweight-proppant claim now has third-party (Wood Mackenzie) validation. But: the Stabroek precedent question is unresolved (no new ruling, no new commentary from XOM on the legal angle); the Baytown 45V FID has not been taken (still the same conditional Woods set in Q2); and Chemical Products dropped ~$1.4B YoY on the same persistent trough management flagged in Q2. The constructive-bias half of "Hold (constructive bias)" has tightened; the price-judgment half has not loosened. We hold the rating and tee up the December 9 corporate plan update as the natural re-rate inflection.

What gets us to Outperform out of Q4 / December plan: (a) the December 9 corporate plan update recuts the FY30 algorithm meaningfully higher on Pioneer synergy revision + Permian technology pipeline + Yellowtail-pulled-forward, with explicit accountability for the >$3B 2026 startup uplift; (b) further FPSO conversion at-or-above design on the next 1-2 Guyana developments; (c) Chemical Products margin inflection on Chinese capacity rationalization; (d) a 10-15% drawdown without thesis impairment that compresses the integrated-major premium. What gets us to Underperform: (a) the corporate plan reveals the >$3B 2026 startup uplift was at "high end" rather than a base case; (b) Brent sustains sub-$55/bbl with dividend-coverage strain; (c) Baytown hydrogen formally cancelled with material write-down language; (d) further negative ruling on the Stabroek precedent that affects future JOA rights.

Results vs. Consensus

Q3 2025 was a modest EPS beat on a top-line miss — the same algorithmic shape as Q2, but with the magnitude on both sides compressed. The story is cost discipline + Permian volume + early Yellowtail offsetting softer realizations, with a $1.4B YoY chemical drag absorbing some of the upside.

Quality of Beat

• EPS-ex-items: The $0.07 beat is fundamentally cost-and-volume rather than price — same algorithm as Q2, but the magnitude is compressed because the Chemical Products YoY drag absorbed about half of the volume tailwind on the bottom line. The non-GAAP / GAAP gap of $0.12 ($1.88 vs. $1.76) is the largest identified-items spread we've seen in recent XOM prints; we await the 10-Q for the line-item attribution but the press-release framing reads as a clean operational quarter with one-time below-the-line items rather than operating noise.
• Revenue: The ~1.7% top-line miss is consistent with the Q3 macro pricing setup (continued softer crude realizations vs. the model deck) and is the second consecutive quarter of revenue lightness on the same root cause. Volumes were a record — this is a price-deck issue, not a market-share or volume issue. Same pattern as Q2.
• Earnings progression: Earnings-ex-items of $8.06B in Q3 vs. $7.08B in Q2 is a +$0.98B sequential expansion despite continued softer realizations. The math implies the Permian volume add + project startup contribution + structural cost takeout more than offset the price compression and the chemical drag. YTD 2025 GAAP earnings of $22.34B vs. YTD 2024 $26.07B is a -$3.73B (-14%) YoY decline; H1 was -$2.67B and Q3 alone was -$1.06B vs. PY, so the YoY pace is essentially flat-lining at the H1 trajectory (no further deterioration).
• Cost takeout: Cumulative structural cost savings now $14.3B since 2019, up from $13.4B at H1, implying ~$0.9B added in Q3 alone. Per Mikells, "our track record in this regard is bar none." The flywheel continues to compound at the high end of management's $2-2.5B/yr framing.
• Capital return: $9.4B distributed in the quarter ($4.2B dividends + $5.1B buybacks) is consistent with the post-Pioneer $20B/yr buyback program. FCF of $6.3B / distributions of $9.4B is a 67% coverage ratio — XOM is funding distributions out of a combination of cash flow and balance sheet drawdown for the moment, consistent with the buy-the-dip framing the board has communicated for several quarters.
• Stock reaction: Shares fell marginally on the print, with investors focused on the chemical drag and revenue light rather than the production beats. The relative-strength flag — XOM trading flat-to-down on a quarter that for any other integrated major would be a clean beat — tells you the equity was priced for stronger headline numbers.

Segment Performance

Headline segment dollar disclosures from the press-release excerpt available to us are limited to YTD comparisons; we use the segment color from the call as the primary input below and will refresh against the 10-Q dollar splits when published.

Upstream: Two Records and an Early FPSO

Upstream is, again, the cleanest piece of the print. Three things stand out: a record Permian, a record Stabroek, and Yellowtail delivered four months ahead of schedule.

Permian: "Nearly 1,700,000 oil-equivalent barrels per day" in the quarter — a record and a step up from Q2's ~1.6 mboed. The proprietary-proppant rollout is now running ahead of plan: ~25% of new wells use the proppant in 2025, scaling to ~50% in 2026. Wood Mackenzie published an independent study during the quarter validating up-to-20% recovery uplift — an external endorsement of an internal claim that previously had to stand on management's say-so. The Sinakin Petroleum bolt-on (~80,000 net high-quality Midland acres) extends drilling-location inventory and provides incremental locations for technology deployment.

"In the Permian Basin, another advantaged asset, we set yet another production record of nearly 1,700,000 oil-equivalent barrels per day. We also acquired more than 80,000 net high-quality acres in the Midland Basin from Sinakin Petroleum. The transaction provides control of drilling locations and opportunities to further deploy our technology to drive greater returns. It is another example of bringing our portfolio advantages to an acquisition, ensuring that one plus one equals three or more." — Darren Woods, CEO

Woods then delivered the most direct peer-comparative line we've heard from XOM management this cycle: "This is an important point, as it clearly differentiates us from our competitors who are talking about reduced investments, peak production, or a shift to harvest mode." We read this as XOM positioning the December 9 corporate plan to extend the Permian production trajectory beyond the current 2.3 mboed-by-2030 target. That extension — if numerically meaningful — is a candidate upgrade trigger.

Guyana: Q3 Stabroek production "more than 700,000 barrels per day," up from ~650 kboed in Q2. Yellowtail (250 kbd nameplate, fourth and largest FPSO) came online four months ahead of schedule "in nearly the same time as previous FPSOs despite a 70% increase in facility weight." That phrasing — same execution time, 70% bigger facility — is the project organization's most concrete proof point that XOM's FPSO build-cycle has structurally compressed. Hammerhead, the seventh development, was newly sanctioned with first oil in 2029 (250 kbd capacity); this completes a meaningful chunk of the eight-development build to 1.7 mboed gross / 1.3 mboed net by 2030.

"Last December, we reviewed our corporate plan under the theme of 'A League of Our Own.' The results we have delivered since then continue to support that theme... in the third quarter, we delivered our highest earnings per share compared to other quarters in a similar price environment." — Darren Woods, CEO — the relative-strength framing he is asking the buy-side to adopt

Discovery Six — the new HPE/NVIDIA supercomputer commissioned in the quarter (the world's seventeenth most powerful) — is described as cutting seismic processing from "months" to "weeks" and delivering "more than a billion dollars in potential value capture from increased resource recovery at our first six FPSOs in the Stabroek Block." We treat the dollar number as a directional management framing rather than a firm uplift, but the broader point — that XOM is converting compute capability into recovery uplift across an already-producing portfolio — is a real and durable advantage.

Stabroek arbitration: No new commentary from management on the call this quarter. The ruling remains as previously discussed; CNOOC's stake unchanged, XOM's share unchanged, operational continuity intact. The contractual-precedent overhang we flagged in Q2 has not been re-tested this quarter. We continue to monitor for follow-on litigation or signals on future joint operating agreements.

Energy Products (Refining): Margin Tailwind, Reliability Record

Refining margins were "pretty supportive this year" and a contributor to sequential earnings growth. Per Woods, the macro setup is a looser crude balance (cheaper feed) coupled with a tighter product balance (capacity offline, supply disruptions) — the cleanest configuration for refining earnings. XOM-specific: "in the third quarter, we saw the highest reliability that we have ever had." That reliability claim is materially load-bearing for the structural-cost narrative because it implies the centralized global operations organization is delivering both lower maintenance cost AND higher run-rates simultaneously — a rare both-axis-improvement claim.

"We are taking the low-value products that come out of a barrel of crude and putting in the conversion capacity to make those high-value products. The most recent and significant example of that is what we did in Singapore with our CRISP project, where we took the lowest value product, residue fuel oil, and converted that to some of our highest value products with a brand new to the world lube base stocks and additional diesel." — Darren Woods, CEO

The Singapore Resid Upgrade — which produced first on-spec base stocks in Q2 — is now at "around 80%" utilization, ramping to full capacity by year-end. That is the textbook XOM project ramp shape and de-risks one of the larger items in the >$3B 2026 startup-earnings framework. Baytown was newly FIDed in the quarter (refining margin uplift project, distinct from the at-risk hydrogen project); Woods framed it as "a continued step in [high-grading molecule conversion]" with "very good returns and very resilient returns."

Chemical Products: The Trough That Won't Quit

Chemical Products is the segment of the print under the most macro pressure and the one that prevented us from upgrading. The segment was down ~$1.4B YoY — the largest single drag on the YoY earnings comparison — and management has stopped framing the recovery in terms of timing. The Q2 framing of "going to be with us for longer than anybody would like" has hardened into "good demand for chemical products around the world, but a lot of supply chasing that demand."

The structural offsets management cites are unchanged from prior quarters: feed flexibility allowing daily margin optimization, structural feed cost advantages, high-value product mix shift via Proxima and other product innovation. We agree those keep XOM Chemicals at the top of the peer set in absolute earnings terms; the question we keep posing is what level of mid-cycle earnings power the segment recovers to, and management's answer this quarter was effectively "we don't know, but we'll get there." The China Chemical Complex completed startup; full ramp is now expected over the coming quarters.

Assessment: No incremental data point this quarter to either inflame or relieve our Q2 view. The segment continues to anchor the Hold case alongside LCS uncertainty.

Specialty Products and Low Carbon Solutions: Mixed; the Forward Story Is Carbon Materials

The most newsworthy item in this segment basket is the Superior Graphite asset acquisition. Per Woods, XOM bought "key assets" plus technology rights in the graphite/specialty-carbon space, with the strategic logic being a battery-anode product that delivers "30% faster charging, 30% more range, and four times longer battery life" in early OEM testing. The TAM Woods cited is "up to $40 billion" with explicit competitive framing on the cost-curve angle: "outcompete the Chinese." That last phrase is rare in XOM management commentary and we read it as a strategic-narrative shift on the carbon-materials platform.

CCS: No major incremental update this quarter; the third-party offtake portfolio continues to ramp consistent with prior framing.

Baytown low-carbon hydrogen: No FID this quarter. The 45V haircut and the Woods "won't move forward" conditional from Q2 are unchanged. The market has absorbed the optionality impairment over the past 90 days; we don't think Q3 introduces incremental risk on this question, but we also don't see an incremental positive data point. This is one of the unresolved Q2 anchors.

Hyperscaler / data-center decarbonization power: A more substantive thread on Q3 than on Q2. Woods framed XOM as "very, very engaged with most of the hyperscalers" on carbon-abated natural-gas power for data centers, leveraging the Denbury CCS infrastructure. He was careful to emphasize XOM's role is the molecule side (decarbonized natural-gas power + CCS) rather than the electron side: "we are also working with independent power producers to, with them, provide the electron side of the equation while we provide the molecule side of the equation." Status: "We are early in the game. We will see what gets translated into actual contracts and then into construction." The right characterization — this is real interest, not yet contracted, on a multi-year time frame. We would not value the optionality at zero and would not ascribe meaningful FY26-27 earnings to it either.

Key Topics & Management Commentary

Overall Management Tone: More relaxed and more peer-comparative than Q2, with the "league of our own" framing more confident on the back of the operational record print. Woods opened with the framing explicitly: "in the third quarter, we delivered our highest earnings per share compared to other quarters in a similar price environment." That is a clear management positioning — "watch the relative print, not the absolute one" — and it is internally consistent with the Q2 messaging. The December 9 corporate plan update was teed up multiple times during the call as the moment when the multi-year story will be recut.

The 2025 Project Start-Up Wave — Eight of Ten Done

Management's framing on the 2025 start-up cohort tightened from Q2's "five projects coming online" to Q3's "eight key startups we have highlighted to date" out of ten. Two remaining: Proxima systems expansion and Golden Pass LNG, both on track for year-end. Aggregate guided contribution unchanged: ">$3 billion in earnings contributions next year at constant prices and margin." The 2025 cohort gross capital is "on the order of $50 billion" per Woods — a number that, if delivered as guided, sets the 2026 starting line for the FY30 algorithm at a meaningfully higher base.

"You have seen it this year with our global projects organization and the eight key startups we have highlighted to date, which include some of the industry's largest and most complex projects... Together, these 10 projects establish an important foundation for our 2030 earnings and cash flow growth plans. They are expected to drive more than $3 billion in earnings contributions next year at constant prices and margin." — Darren Woods, CEO

Capital Allocation: Disciplined Underrun, Not Activity Cut

Cash CapEx is now expected to come in below the $27-29B low-end of the guided range (excluding the $2.4B of in-quarter M&A). Woods's framing: "I would really just caution everyone to take the changes in the CapEx that you are hearing today as a reflection of, in my mind, what we refer to as disciplined capital spending, which is not cutting CapEx, but spending it in a wise way." Mikells added that LCS pacing reflects the rate of market development rather than a strategic retreat: "the market is not developing as fast as we had planned for, and so we are pacing the spend in that."

This is the right framing for the buy-side; the underrun is project-timing variance plus the LCS market not maturing at the pace of original plan, not a strategic CapEx cut. Importantly, Woods explicitly distinguished this from the inorganic activity in the quarter ($2.4B of M&A across the Sinakin and Superior Graphite deals): "we obviously do not plan for those transactions, so that is why we excluded them" from the CapEx guide.

Lightweight Proppant: From Internal Claim to Third-Party Validated

The most interesting technology-narrative shift this quarter was the move from an internally-claimed proppant uplift to a third-party-validated one. Wood Mackenzie published a report during the quarter validating XOM's lightweight-proppant resource-recovery claim and acknowledging that "our upstream integration with refining operations creates a strategic advantage that is difficult for others to replicate." This is the first independent endorsement of the proppant claim and meaningfully de-risks the central technology underpinning the Permian production-growth-into-the-next-decade framing.

The "Buy Value, Not Volume" Refrain on M&A

Sam Margolin asked whether XOM's balance sheet and capital efficiency now support a step-up in inorganic activity. Mikells: "We look at a lot of things. A lot of things. As you would expect us to. We transact on very few things because they have to meet our criteria." Woods followed: "I think the bottom line in that is we buy value, not volume. I think that differentiates us from many in the industry." That answer reads as a deliberate signal that XOM is not in scoping-call mode on a near-term major transaction — a useful read on the 12-month inorganic backdrop.

Forward Guidance / Outlook

Formal forward guidance was limited; the December 9 corporate plan update will recut the FY26-30 framework. Specific Q3 forward signals worth flagging:

• 2025 cash CapEx: below the low end of $27-29B (excluding $2.4B of in-quarter M&A).
• 2025 startup wave: 8 of 10 delivered; Proxima systems expansion and Golden Pass LNG to complete by year-end. Aggregate >$3B 2026 earnings contribution at constant price/margin reaffirmed.
• Permian production: well above 1.7 mboed exit-2025 implied; new "well into the next decade" framing on the production trajectory teed up for the December plan.
• Singapore Resid Upgrade: full capacity by year-end; currently ~80% utilization.
• Structural cost savings: similar ~$2.5B add expected for 2025 on top of the $14.3B running cumulative.
• Hammerhead: first oil 2029, 250 kbd nameplate; seventh of eight Stabroek developments to first oil.
• Mozambique LNG: Woods characterized the project as "in a very good place" with the security situation "improved dramatically"; in the process of lifting force majeure following Total. FID timing not specified.

Analyst Q&A — What Mattered

The Q&A was dominated by capital deployment, Permian technology depth, and exploration. We summarize the exchanges that most affect the thesis (paraphrased; analyst names + firms cited for context, no quotation):

• Neil Mehta (Goldman Sachs) opened on the CapEx underrun. Woods's answer reframed the underrun as deliberate pacing of LCS spend in line with market development — not an activity cut. Mikells added the in-quarter $2.4B of M&A is excluded from the cash CapEx framework, a useful clarification for modelers.
• Devin McDermott (Morgan Stanley) asked for unpacking on the Permian production beat and the technology drivers. Woods answered with the proppant + cube-development + technology-pipeline framing and teed up the December plan to "build that improvement into the plan outlook" — signaling Permian forward-trajectory upside in the December plan.
• Arun Jayaram (JPMorgan) asked about the global outlook through 2050 (20% gas growth, doubling of LNG demand). Woods's answer was foundational rather than incremental — outlook informs strategy, depletion math drives the need for continued investment.
• Doug Leggett (Wolfe Research) challenged the "pedestrian" dividend growth rate as potentially holding back valuation recognition. Mikells deflected with the 43-year consecutive-growth track record and the multi-criteria "sustainability + growth + investor-feedback" framing. Woods added the commodity-cycle resilience argument. Read: dividend cadence is unlikely to change materially; the buyback remains the principal capital-return signal.
• Bob Brackett (Bernstein) drilled into Superior Graphite. Woods provided the most quantitative TAM framing of the call ("up to $40 billion") and the strategic logic on outcompeting Chinese supply. Useful incremental detail on a small-but-strategic platform.
• Sam Margolin (Wells Fargo) on whether the balance sheet now supports a step-up in inorganic activity. The "buy value, not volume" answer is the salient line; XOM is not telegraphing a near-term transformative deal.
• Paul Cheng (Scotiabank) on whether the project pipeline has hit organizational-capability limits. Woods: not at a limit; the binding constraint is the high return-criteria filter rather than execution capacity. This matters — it tells you XOM has runway to absorb additional inorganic activity if the right asset shows up.
• Biraj Borkhataria (RBC) on Mozambique. Woods characterized the project as "in a very good place" with security improved; FID early-2026 not directly confirmed but implied as plausible.
• Ryan Todd (Piper Sandler) on exploration. Woods's three-lever framing — squeeze more from existing fields + grow advantages for inorganic + find new resources — is the cleanest articulation of the upstream long-term framework we've heard from this management.
• Betty Jiang (Barclays) on hyperscaler power contracts. Woods reaffirmed XOM's molecule-side / carbon-abated framing — not interested in offering traditional power without the carbon capture leg.
• Jean Ann Salisbury (Bank of America) on proppant granularity. Woods: not at the end of the learning curve; pipeline of additional technologies; patent and feedstock supply both protected.
• Jason Gabelman (TD Cowen) on the industry-wide pivot to exploration. Woods: industry is now seeing what XOM has been investing toward for years; competitive advantage is in execution capability and technology rather than acreage alone.
• Paul Sankey (Sankey Research) on the AI CapEx boom and the comparison to XOM's own 2013 peak CapEx. Woods declined to opine on hyperscaler capital deployment but reaffirmed XOM's deployment of AI internally (Permian + global operations + Discovery Six supercomputer).
• Philip Jongworth (BMO) on refining margins and the Baytown FID. Woods's answer was the cleanest articulation of XOM's "shift the molecule make" strategy that we have heard.

What They're NOT Saying

Three notable absences from the Q3 commentary:

• No update on the Stabroek arbitration legal-precedent question. The Q2 framing — that XOM would litigate to protect contractual rights and that the ruling was a "surprise" — is the most recent management commentary on the topic. Q3 was silent. We would have expected at least a paragraph on whether the ruling has affected ongoing or planned joint operating agreements; the silence is itself a data point that legal activity is either contained or being held for non-public discussion.
• No new commentary on the Baytown low-carbon hydrogen FID timing. Woods's "won't move forward" conditional from Q2 is the live framing. We would have expected an update on whether the commercial offtake / policy environment has clarified enough to either green-light or formally cancel the project. Silence here is consistent with the project being in genuine "wait and see" rather than active development.
• No specific dollar quantification on the chemical trough recovery path. Management has stopped guiding to a recovery date or a specific mid-cycle dollar number for Chemical Products. That is appropriate given the macro uncertainty but it leaves the buy-side modeling Chemical Products at an unknown floor.

Market Reaction

XOM shares fell marginally on the print — an unusual reaction for a record Permian + early Yellowtail + EPS-ex-items beat. The market focused on (a) the ~$1.4B YoY chemical drag and (b) the ~1.7% top-line miss against ~$86.8B consensus, with the GAAP-vs-adjusted-EPS gap (the $1.76 GAAP print fell short of headline numbers some screens had set as the consensus reference) creating a brief noise-trade window in the early session.

The relative-strength flag is the read worth keeping: a flat-to-down day on a quarter that for any other integrated major would be a clean operational beat is the market's way of confirming that XOM is priced for stronger headline numbers than this quarter delivered. That read is consistent with our Q2 thesis — the equity already discounts the Permian + Guyana premium — and reinforces the price-judgment argument behind the Hold rating.

Distribution dynamics in the quarter ($9.4B in shareholder distributions, $5.1B of buybacks alone) provide ongoing technical support; we don't see the buyback cadence faltering.

Street Perspective

The Street arrived at Q3 looking for confirmation that the operational machinery was running at the framing management laid out at the 2024 Analyst Day — specifically, Permian above 1.6 mboed and progress on the Guyana FPSO conversion. The print delivered on both. The bull case being made on the Street is essentially that XOM is the integrated major with the cleanest line-of-sight to FY30 earnings and cash flow growth, supported by (1) the Permian-into-the-next-decade framing teed up by Woods on the call, (2) eight-of-ten 2025 startups completing on schedule, (3) Yellowtail four months early and Hammerhead newly sanctioned, and (4) Pioneer synergies tracking above the original underwriting.

The cautious view on the Street — closer to our Hold framing — rests on three points: chemical trough persistence, Baytown 45V optionality impairment, and the question of how much further the Permian / Guyana premium can compress relative to Chevron and the European super-majors before the relative-value math turns. The cautious view points out that the Q3 EPS-ex-items beat was only ~3.9% on a 12% earnings sequential expansion — the absolute earnings power is still YoY-down and the recovery to FY24-mid-cycle levels has not yet begun in the print itself.

The December 9 corporate plan update is the agreed-upon next inflection point on the Street; we share that view.

Model Implications

We make the following directional model adjustments after the print:

• Permian production trajectory: Roll the 2025 exit rate from ~1.6 mboed to ~1.7 mboed. Hold the 2030 target at 2.3 mboed but flag upside risk pending the December plan recut. The proppant adoption curve (25% of new wells in 2025 → 50% in 2026) implies recovery uplift compounding through the model period; if the Wood Mackenzie validated 20% uplift holds, the per-well economics improvement supports a higher 2030 target with the same well count.
• Guyana production trajectory: Pull forward Yellowtail contribution by ~1 quarter (4-month early start). Add Hammerhead at 2029 first oil with 250 kbd nameplate; risk-adjust at 70% capacity in year-one ramp. 2030 gross target of 1.7 mboed is now better-supported.
• Chemical Products earnings power: Mark down the FY26 mid-cycle assumption by ~$0.5-0.7B from prior given the trough has now persisted longer than the median cycle and management has stopped guiding to recovery timing. Hold the FY28+ recovery path but at lower confidence.
• Energy Products: Roll Singapore Resid Upgrade to full capacity by year-end 2025; Baytown refining FID adds a multi-year value-mix uplift starting in 2027-2028. Cumulative startup contribution to 2026 earnings remains the >$3B framework.
• Low Carbon Solutions: Hold the conservative LCS earnings model from Q2 (Baytown hydrogen at ~30% probability-weighted FID); add Superior Graphite carbon-materials platform as a small-but-real long-dated optionality line item ($40B TAM framing per management; ascribe FY30+ contribution at single-digit %).
• Capital return: Hold the $20B/yr buyback program through 2025; FCF coverage of distributions improves on the chemical recovery and the FY26 startup uplift. Dividend growth assumption unchanged at low-single-digit (consistent with 43-year framing).
• Cumulative structural cost savings: Roll cumulative target tracking from $13.4B → $14.3B; on pace to $18B by 2030 framework holds.
• FY30 algorithm: Stable on the $20B-earnings / $30B-cash-flow framing through Q3; the December 9 plan update is the next inflection point. We expect the plan to recut the FY30 framework higher but reserve judgment on magnitude until December.

Thesis Scorecard — Walked from Q2 Initiation

Overall: Five of six bull points are confirmed or strengthened, with one new bull added (proppant validation). On the bear side, the chemical trough has worsened and the LCS impairment is unchanged; the Stabroek precedent risk has gone latent (no new information). The central tension from Q2 holds: operational execution is in the top decile of the integrated-major peer set and the multi-year setup has further de-risked, but the equity has continued to price the constructive view and two of the three Q2 Hold-anchors remain unresolved. The Hold rating is, again, a price judgment on a high-quality franchise.

Action: Maintaining Hold. Operational execution this quarter was at the high end of the bull case — a record Permian, an early Yellowtail, Hammerhead newly sanctioned, eight of ten 2025 startups complete, third-party validation of the lightweight-proppant technology, and the structural-cost flywheel adding ~$0.9B in the quarter. The price-to-thesis math at current levels still leaves modest upside under the base case and meaningful downside under a sustained low-realization scenario or further LCS optionality impairment, and the chemical trough has worsened rather than improved. The December 9 corporate plan update is the natural next inflection point for a re-rate; we want to see the FY30 framework recut before changing the rating. Upgrade triggers: (a) December 9 plan recuts the FY30 algorithm meaningfully higher with explicit accountability on the >$3B 2026 startup uplift and a Permian trajectory extending beyond 2.3 mboed; (b) chemical margin inflection on Chinese capacity rationalization with a quarterly dollar lift; (c) further FPSO conversion at-or-above design on the next 1-2 Guyana developments; (d) a 10-15% drawdown without thesis impairment that compresses the integrated-major premium. Downgrade triggers: (a) the corporate plan reveals the >$3B 2026 startup uplift was high-end rather than base-case; (b) Brent sustains sub-$55/bbl for two-plus quarters with dividend-coverage strain; (c) Baytown hydrogen formally cancelled with material write-down language; (d) further negative ruling or precedent on Stabroek arbitration affecting future JOA rights. We will revisit on the December 9 corporate plan update and again on the Q4 2025 print.