The Balance-Sheet Bear Case Closed in a Single Quarter. The Valuation Bear Case Did Not.

Results vs. Consensus

Phillips 66 filed no pre-announcement ahead of this print. The Item 2.02 8-K that reset expectations three weeks before the first-quarter release has no second-quarter counterpart, and the gap between the Q1 8-K on May 14 (annual meeting results) and the earnings 8-K on August 5 contains nothing. The Street therefore modelled a war-driven margin quarter with no company reset, and the estimate dispersion shows it: four vendors published consensus adjusted EPS between $6.76 and $7.68, a spread of $0.92 on a number that printed at $9.41.

Q2 2026 Scorecard

Consensus note. The four vendor readings of the same consensus were $7.68, $7.50, $7.02 and $6.76, implying surprises between 22.5% and 39.2%. We use the most conservative and most widely syndicated of the four, which states the smallest beat. Any headline putting the surprise near 40% is quoting the lowest available estimate.

Revenue note. Unchanged from last quarter and unchanged in conclusion. The income statement carries sales and other operating revenues of $51,004M and total revenues and other income of $52,044M. Three data vendors reported three different "actuals" for this quarter, at $52.04B, $51.93B and $42.1B, and disagreed on the direction of the surprise. For a refiner, the top line is overwhelmingly a crude-price pass-through. We report the actual and decline to score it.

Year-over-Year Comparison

The year-over-year comparison is clean in a way the Q1 comparison was not. Neither period carries a material disposition gain in the adjusted figures, both periods reflect the post-WRB Central Corridor consolidation on the adjusted-throughput basis used for margin and cost per barrel, though the West Coast cessation still falls between the two periods and does not lap until the fourth quarter. On that basis adjusted earnings quadrupled on a refining margin that doubled, at slightly lower utilization and 6.9% more throughput.

Sequential Comparison

Prior-period note. The supplemental data filed with this release restates first-quarter 2026 adjusted EBITDA from the $1,268M reported on April 29 to $1,230M. The entire $38M sits in Chemicals, where the proportional share of CPChem's income taxes was revised from $21M to $13M and the proportional share of depreciation and amortization from $144M to $115M, taking segment adjusted EBITDA from $250M to $212M. Segment pre-tax income is unchanged. This report uses the restated figures throughout, so the Q1 EBITDA comparatives here differ from our first-quarter note.

Assessment: Revenue

Total revenues and other income of $52,044M rose 55.3% year over year and 57.7% sequentially, and the driver is price rather than volume. WTI averaged $93.21 against $71.98 in Q1 and $63.86 a year ago, and Brent averaged $104.52 against $80.61 and $67.82. Total processed inputs of 2,053 MBD were up only 6.9% year over year and 2.2% sequentially. For a business that buys crude and sells product, a 55% revenue increase on a 7% volume increase is the crude market passing through the income statement. The line to watch is the spread, and the spread is reported directly as realized refining margin.

Assessment: Margins

Realized refining margin of $24.08 per barrel is the highest the company has reported in the six quarters shown in the supplemental data, and it is more than double both the $11.25 of a year ago and the $10.11 of last quarter. What matters for the forward model is the decomposition. Market capture fell 40 points to 98%, so essentially all of the margin expansion came from the market crack rather than from commercial outperformance. That is the mirror image of Q1, where an unremarkable crack was converted into a 138% capture. The company earned this quarter's margin by owning refineries in a tight market, not by out-trading anyone.

Regional dispersion widened sharply. Central Corridor realized margin went from $4.60 to $29.56 per barrel, Gulf Coast from $11.31 to $24.25 and West Coast from $13.12 to $29.65. Atlantic Basin and Europe moved the other way, from $15.62 to $14.44, the only region where realized margin fell in a quarter when the worldwide figure more than doubled. Clean product yield slipped one point to 86% and crude utilization rose one point to 96%.

Assessment: EPS

Reported adjusted EPS of $9.41 overstates the operating quarter, which is the precise opposite of the problem three months ago. The $450M mark-to-market gain and the $100M tariff refund together are worth roughly $1.37 of pre-tax EPS and about $1.06 after tax, putting the clean figure near $8.35. Even that is an extraordinary number: annualized, it is $33.40 against a $202.55 share price.

The more useful anchor is the first half, because it nets the two mark-to-market swings against each other. Clean first-half adjusted EPS of approximately $10.44 on a first-half realized margin of $17.21 per barrel is the cleanest earnings-power observation this company has produced in the two quarters we have covered it. It is also, by construction, an observation about a war. The tax rate normalized to 22.4% from a Q1 figure of 15.7% that was depressed by a small pre-tax base, which removes one of the reasons the Q1 print was hard to read.

Grading Last Quarter's Commitments

The first-quarter call produced an unusually specific list of forward statements. This is the scorecard against them. It is the part of the quarter that says something about the company rather than about the crack spread.

Assessment: Ten of thirteen were met or beaten, and the two largest, the debt path and the working-capital reversal, were delivered in a single quarter rather than over the four the company had allowed itself. The three that were not met are all disclosure failures rather than operating failures: a project without a capital number, an acquisition without an earnings contribution, and a headwind the company named but has never sized. That pattern matters for how much credit to extend the 2027 targets, and it is the reason we treat the Midstream bridge as evidence rather than as arithmetic.

Segment Performance

Refining

Refining produced $3,086M of adjusted pre-tax income, more than double the entire company's adjusted pre-tax income in any of the five prior quarters the supplemental data discloses, on a realized margin of $24.08 per barrel and 2,053 MBD of processed inputs. Three of four regions delivered their best realized margin in at least five quarters. The fourth was the region that carried the first quarter.

Regional figures are on an adjusted basis. The only special item in the segment this quarter was a $24M legal accrual charged to Atlantic Basin and Europe, so that region's reported pre-tax income was $352M against the $376M adjusted figure shown. Central Corridor comparatives to periods before October 1, 2025 remain distorted by the WRB buy-in, which moved Borger and Wood River from proportional equity accounting to full consolidation. The realized-margin and cost-per-barrel lines use adjusted throughput that grosses up prior periods for the equity-affiliate share, so those two metrics are comparable even where the dollars are not.

Central Corridor: the largest single-quarter regional swing in the portfolio

A $2,017M sequential swing, from a $418M loss to $1,599M of pre-tax income, on realized margin going from $4.60 to $29.56 per barrel. Turnaround expense in the region fell from $105M to $81M and throughput rose from 760 to 834 MBD at 101% crude utilization. Our first-quarter note said a swing of $400M or more was available here without any change in the macro. The macro changed, and the swing was five times that.

The disclosed Canadian heavy sensitivity did not do the work. WTI less WCS averaged $14.61 against $14.13 in Q1, roughly $0.48 wider, not the near-$18 level management described on the April call. At the company's stated $140M of annual EBITDA per dollar of widening, that is worth about $67M annualized. The Central Corridor result is a crack-spread result.

"Well, our general view on WCS is that differentials are going to wider structurally over time, driven primarily by growing heavy crude supply out of Canada and Venezuela."
— Brian Mandell, Marketing, Commercial and Renewable Fuels

Assessment: The mechanical reversal we expected happened and the macro amplified it. The forward point is that the heavy differential thesis is still unbanked. A structural widening toward $18 remains available and is worth roughly $540M annualized at the company's own sensitivity, concentrated in exactly this region. Nothing in the second quarter proves or disproves it.

Gulf Coast: the pricing lag turned from headwind to tailwind

Gulf Coast delivered $953M on a $24.25 realized margin at 96% utilization, against $204M and $11.31 last quarter. The April pre-announcement had quantified a roughly $300M pre-tax drag from the standard two-week lag in Gulf Coast clean-products pricing in a sharply rising market. Our first-quarter note said that lag would keep working against the region into Q2 and that the benefit would arrive with a delay. Crude peaked inside the quarter and the region printed a $17.80 per barrel pre-tax result, against $3.87 last quarter. Operating expenses in the region fell from $298M to $247M as turnaround expense dropped from $41M to $15M. Clean product yield of 81% remains the lowest of the four regions, which is structural.

Assessment: This is the region where the second-quarter print is most straightforwardly a market result and least a company result. It is also the region that will show the lag reversing first if the crack rolls over, because the mechanism is symmetric.

Atlantic Basin and Europe: the region that gave back Q1's capture

The only region where realized margin fell sequentially, from $15.62 to $14.44 per barrel, and the only one whose pre-tax income was roughly flat, at $376M against $367M. Management disclosed the reason directly, and it is the most useful piece of capture data on the call: Atlantic Basin capture was 182% in the first quarter and 79% in the second.

"When we looked at the Atlantic Basin, Q1 capture rate was pretty high at 182%. And then Q2 was on the other end of that spectrum at 79%. I really think the way you have to look at this with all the noise in the system over those 2 quarters is you got to look at it on a first half basis. And the first half capture rate was just well over -- not well over, but 112% on average."
— Rich Harbison, Refining

The region also absorbed a Humber turnaround inside that window, at $25M of expense against $30M in Q1, and utilization fell from 96% to 92%. Management noted the 2023 to 2025 average capture for the region was about 96%, so the first-half 112% is above trend even after the second-quarter reversal.

Assessment: This single disclosure retrospectively explains most of the 138% worldwide capture in Q1 and most of the 40-point fall this quarter. The commercial value in the Atlantic Basin is real on a half-year view and violently noisy on a quarterly one. Anyone who built a Q2 estimate off a Q1 Atlantic Basin run-rate was modelling a number that had already half-reversed by the time they published it. The honest framing for this asset is a two-quarter average, and on that basis it is running roughly 16 points above its own three-year norm.

West Coast: still a small business, now a profitable one

$158M of pre-tax income on a $29.65 realized margin at 93 MBD of throughput, against $55M and $13.12 last quarter. Regional depreciation and amortization stayed at $16M and operating and SG&A expenses were $63M. The realized margin is the highest of the four regions and it is being earned on a volume base nearly six times smaller than the next-smallest region.

Assessment: Our first-quarter view was that the West Coast should be modelled at roughly 96 to 120 MBD with a near-zero depreciation base rather than as a recovering 230 MBD business. The second quarter ran 93 MBD, at the bottom of that band, and produced $158M because the margin environment tripled. The asset is now a small, high-torque, cost-stripped position rather than a problem, and the associated burden continues to sit in Corporate and Other at $38M of idling costs for the quarter. Nothing here changes on a rate basis; it changes with the crack.

Midstream

Midstream produced $785M of pre-tax income and $1,046M of adjusted EBITDA, against $591M and $860M last quarter and $731M and $972M a year ago. NGL fractionated volumes of 1,020 MBD were a record, as were LPG export volumes. NGL pipeline throughput of 943 MBD rose sequentially from 930 but remains below the 956 of a year ago. Wellhead volume reached 4.5 Bcf/D, transportation pipeline volumes 3,142 MBD and terminal volumes 3,433 MBD, all higher than any period in the five quarters disclosed. Dos Picos II, a 220 MMCFD Permian gas plant, reached full production. Two new projects were announced inside the quarter: the 300 MMCFD Zeus gas plant and a 100 MBD Coastal Bend NGL fractionator at Corpus Christi.

The composition is where the caution belongs. Of the $186M sequential increase in adjusted EBITDA, $176M came from the NGL business, where the weighted-average Mont Belvieu NGL price rose from $0.62 to $0.73 per gallon. Transportation, the fee-based half that the "low-volatility earnings floor" argument actually rests on, rose $10M, from $306M to $316M.

"I think we're still hovering around in the Midstream segment around that $1 billion a quarter mark with some ups and downs depending on commodity prices and depending on just some volume variances. But largely, we're solid in that run rate."
— Don Baldridge, Midstream and Chemicals

Assessment: The $4.5B adjusted EBITDA target for year-end 2027 looked demanding at the first quarter, when the run-rate annualized to roughly $3.4B and required a third of growth in seven quarters. At $1,046M this quarter the annualized run-rate is $4,184M and the gap is $316M. That is a genuine upgrade to the pillar. It is also, on this quarter's evidence, mostly a commodity-price upgrade rather than a project upgrade, and the projects that are supposed to close the remaining gap (Iron Mesa, the Coastal Bend expansion, the two new sanctions) do not contribute until 2027. We now treat the target as reachable rather than stretched, while noting that the quarter closed the gap with the one input the segment is not supposed to depend on.

Chemicals

Chemicals produced $404M of pre-tax income and $528M of adjusted EBITDA, against $85M adjusted and $212M last quarter and $20M and $148M a year ago. CPChem's own 100% net income was $806M against $228M. The ethylene-to-high-density-polyethylene chain cash margin went from 10.7 cents per pound to 43.6, roughly a fourfold move on top of the fourfold move it made in the first quarter. Global olefins and polyolefins utilization was 91%, against guidance of the low 80s and against 94% in Q1.

Management put a number on the structural claim it made last quarter, which is the more valuable disclosure.

"And so we see things relatively more stable, though they'll be below mid-cycle. They shot above mid-cycle temporarily. They'll be coming back down to something around $0.14, $0.15 a pound."
— Mark Lashier, Chairman and CEO

The anchor management supplied for that number ties to the filings exactly: full-year 2025 chain margin was 7.1 cents per pound and full-year 2025 Chemicals adjusted EBITDA was $845M. Management's argument is that the loss of deeply discounted Iranian, Russian and Venezuelan crude has raised the Chinese naphtha cost basis permanently, so the floor resets roughly seven cents higher than the 2025 trough.

Assessment: Last quarter we called this the most interesting structural claim on the call and noted it received the least scrutiny. This quarter management quantified it and tied it to a disclosed base. A durable doubling of the cycle floor, from roughly seven cents to roughly fourteen, is worth considerably more to the equity than one quarter at 43.6 cents, and it lands before Golden Triangle and Ras Laffan add capacity in 2027. The claim is still an argument rather than an observation. Its testable form is whether the chain margin settles nearer fourteen cents than seven when the Strait normalizes, and that test is a 2027 event.

Marketing and Specialties

Marketing and Specialties produced $514M of adjusted pre-tax income against a $141M adjusted loss last quarter and $660M a year ago. Realized marketing fuel margins recovered from negative in both regions to $1.82 per barrel in the US and $4.63 internationally, against $2.83 and $7.11 a year ago. US refined product sales rose to 2,115 MBD from 1,948, and worldwide sales to 2,329 MBD. The GAAP figure of $583M includes a $110M gain on post-closing adjustments to the December 2025 sale of 65% of the Germany and Austria retail business, offset by a $41M legal accrual.

Roughly $160M of the segment's result was mark-to-market. Other realized margins and revenues not included in marketing fuel margins, the part of this segment that does not swing with spot, were $279M against $287M a year ago, so the non-spot half of the business shrank slightly.

Assessment: The first-quarter view was that a marketing business whose realized fuel margin goes negative in a sharply rising spot market is behaving as designed, and that the effect reverses when prices stop rising. It reversed. What it did not do is return to the year-ago level: at $1.82 per barrel the US realized fuel margin is still 36% below the $2.83 of a year ago, and the segment's adjusted result is 22% below. In a quarter when refining margins doubled, marketing recovered to well under its own prior-year run-rate. That is the correct shape for an integrated model and a reason not to double-count the cycle across segments.

Renewable Fuels

Renewable Fuels produced $544M of pre-tax income against a $41M loss last quarter and a $133M loss a year ago. Production of 53 MBD was the highest in any period the supplemental data discloses, and Rodeo ran at 106% of nameplate. The credit environment did most of the work: the biodiesel RIN averaged $2.12 against $1.44 last quarter and $1.08 a year ago, CARB ULSD in San Francisco averaged $4.19 against $2.93 and $2.52, and CBOT soybean oil averaged $0.73 per pound against $0.58 and $0.49.

Two items inside the result are not operating. Management disclosed a $100M one-time benefit primarily from tariff refunds and just under $50M of mark-to-market carried over from Q1. Underlying pre-tax income was therefore closer to $394M.

"We also had a onetime help in Q2 of $100 million, primarily due to tariff refunds. As Mark mentioned, Rodeo ran above nameplate capacity with record utilization of 106%."
— Brian Mandell, Marketing, Commercial and Renewable Fuels

Management also flagged a specific policy risk and a modelling change: the concern that foreign feedstock RIN generation is cut in half after the end of next year, and an update to the company's renewable diesel indicator beginning in August to include $0.40 per gallon of production tax credit benefit under the 2026 45Z guidelines released in June.

Assessment: Rodeo has gone from the segment that lost money in five consecutive quarters to one contributing roughly $394M of clean quarterly pre-tax income, and management was asked directly whether that resets the old mid-cycle framework of roughly $700M annually. It declined to reset it. Given that the underlying quarterly figure annualizes above $1.5B, a refusal to raise the mid-cycle number is management telling investors the credit environment is not the run-rate. We read it the same way, and we note that the same policy machinery that doubled RIN values can halve them, on a timetable set in Washington and Sacramento rather than in Houston.

Corporate and Other

The pre-tax loss narrowed to $407M from $451M, against guidance of $430M to $450M. Net interest expense fell to $233M from $255M as debt came down and interest income rose to $81M from $31M on the elevated cash balance. Corporate overhead and other fell to $168M from $187M. Idling costs for the Los Angeles and San Francisco refineries were $38M against $43M in Q1, for $81M in the first half. Third-quarter guidance of $325M to $350M implies a further $57M to $82M of improvement, almost all of it interest.

Assessment: This line is now a deleveraging read-through rather than a cost line. The Q3 guide implies a roughly $1.35B annualized run-rate against $1.6B annualized this quarter, and the delta is the interest saving on $6.6B of retired debt showing up with a one-quarter lag. It is the cleanest confirmation available that the balance-sheet repair is real cash rather than presentation.

Key Operating Metrics

Key Topics & Management Commentary

Overall Management Tone: Management was assured and, for the first time in the two calls we have covered, willing to make forward commitments rather than only forward observations. The balance-sheet answers were quantitative and volunteered where three months ago they were framework-level and defensive, and the cost target moved from striking-range language to an explicit expectation of achievement. The one place the posture stayed careful was the margin environment itself, where the argument for durability was made on mechanism rather than on numbers, and where the company again declined to convert the environment into a target.

1. The Mark-to-Market Reversed, and This Time It Flattered the Print

Three months ago an $839M pre-tax mark-to-market loss made a genuinely good quarter print at $0.49. This quarter the same mechanism ran in reverse and contributed roughly $450M of pre-tax gain to a print of $9.41. The company runs a net short derivative position as an economic hedge against physical inventory that LIFO accounting never marks up, so the paper leg moves and the physical leg does not.

"The company's second quarter financial results were impacted by mark-to-market gains of approximately 50% of the first quarter mark-to-market losses."
— Kevin Mitchell, CFO

The segment split was given when asked: approximately $240M in Refining, approximately $160M in Marketing and Specialties and just under $50M in Renewable Fuels. The 10-Q discloses the position behind it. The net short in crude oil, refined petroleum products, NGL and renewable feedstocks was 36 million barrels at June 30, against 33 million at December 31, 2025 and roughly 50 million at March 31. The gross derivative book moved the other way and grew substantially, from $2,804M of assets and $2,711M of liabilities at year-end to $9,800M and $9,404M at June 30, netting to a $248M carrying value.

Assessment: The first-quarter conclusion holds and now has two data points instead of one. Reported earnings for this company are a poor signal in any quarter with a violent price move, in either direction, and the correct discipline is to strip the mark-to-market in both. The reversal arrived faster than management guided, which is the good news. The less-noticed disclosure is that the gross derivative book more than tripled while the net position shrank by roughly a quarter from the March peak. That is a larger gross exposure being run at a smaller net position, and it is the balance-sheet footprint of a commercial organization that has been leaning into volatility.

2. Six and a Half Billion Dollars of Debt Retired, and a Target Reset That Was Not Filed

This is the quarter's most consequential fact and the one that changes the investment case. Total debt fell from $27,124M to $20,565M, net debt from $21,974M to $16,466M, debt-to-capital from 48% to 39% and net debt-to-capital from 43% to 33%. All outstanding commercial paper was repaid, leaving nothing drawn against the $5B program at June 30 against $200M at year-end. The $2.25B 364-day term loan drawn on March 18 was repaid $1B in the quarter and the remaining $1.25B in July.

"We ended the quarter with total debt of $20.6 billion and net debt of $16.5 billion. This positions us better than where we started the year. And using current consensus estimates, we expect net debt to be less than $16 billion by the end of this year."
— Kevin Mitchell, CFO

Pressed on whether the $17B target should now be reset, the CFO gave a new number verbally.

"And so on a net debt level, I think of a next sort of target is something around about $13.5 billion, which would equate to a $15 billion or thereabouts balance sheet debt number. I don't want to reset the target in absolute debt level terms just because we're also restricted by the -- or impacted by the maturity schedule of the debt we have out there."
— Kevin Mitchell, CFO

The 10-Q filed the same day still states the old commitment: the company is "targeting reductions of total debt to $17 billion and reductions of our debt-to-capital ratio by the end of 2027." The reset exists on the call and not in the filings.

Assessment: The single argument that made our initiation a Hold was that the balance sheet risk and the cycle risk were the same risk, because the deleveraging path required margins to stay strong. That argument has been settled empirically rather than deferred. The gross-up unwound in one quarter, exactly through the mechanism management described, and it unwound while the company was also buying back $379M of stock and paying $508M of dividends. Bear point one is closed. The caveat is procedural rather than financial: a target reset that lives only in a transcript is a softer commitment than one that lives in an SEC filing, and the maturity-schedule caveat attached to it is the CFO reserving the right not to retire debt uneconomically. That is correct capital allocation and it is also a reason to hold the $13.5B figure loosely.

3. Market Capture Fell to 98%, Exactly Where Management Said It Would

Worldwide market capture was 98% against 138% in the first quarter, and against a mid-90s starting point management set in April without being pressed to. It is a small thing that says something large: the company guided down the number it was being celebrated for, and then hit the guide.

"Supported by a strong contribution from our commercial organization, we captured 98% of our market indicator in the second quarter."
— Mark Lashier, Chairman and CEO

Third-quarter capture was addressed directly, and the answer moved the reference point from a one-quarter guide to a standing convention.

"But just specifically, as we think about capture rates, we've historically guided to about 95%, and we don't see any reason why that would be any different this year in terms of where we are."
— Kevin Mitchell, CFO

Assessment: This is the most important disclosure for anyone building a forward model, and it is deflationary. From here the realized margin is the market crack times roughly 0.95, and the commercial organization is a reliability story rather than an alpha story. The Atlantic Basin data makes the point sharper: 182% then 79% for a 112% half. Capture is a quarterly artifact that mean-reverts to the indicator, and the two quarters we have covered bracket it in both directions. We would capitalize this capability at close to the indicator, not above it.

4. The Cost Program Reached $5.57, and the Year-over-Year Comparison Is Worse Than the Headline

Refining adjusted controllable costs excluding turnaround expense were $5.57 per barrel against a $5.50 target for 2027. Management's framing was the most confident it has been on any topic across two calls.

"So what I see in summary is that $5.50 is well within range and I fully expect us to achieve that goal next year. And those cost improvements that we are driving for our stock owners, these are structural. They're not going to work their way back into the system."
— Rich Harbison, Refining

Three specific initiatives were named out of a pipeline of more than 200: an online boiler-tube cleaning process on the Bayway FCC boilers, a heat recovery project at Ferndale, and an acid consumption project at Wood River, each worth more than $1M a year. The target basis is a $3.00 per MMBtu Henry Hub price.

Two facts complicate the headline. First, Henry Hub averaged $2.93 this quarter against the $3.00 target basis, so the $5.57 needs essentially no normalization. Applying the sensitivity implied by management's own first-quarter disclosure, where a $1.87 per MMBtu gas overrun was worth roughly $0.40 per barrel, the gas-normalized figure is about a cent and a half higher. That is genuinely close to target and it is a real achievement, in contrast to the first quarter where the reported $6.21 required a $0.40 normalization to look reachable.

Second, and less flattering, the year-over-year comparison went the wrong way. The same metric was $5.46 per barrel in the second quarter of 2025, when Henry Hub averaged $3.16 and throughput was 6.9% lower. Adjusted controllable costs excluding turnaround rose from roughly $955M to roughly $1,041M, or 9.0%, against 6.9% more barrels, on cheaper gas.

Assessment: Both things are true and management presented only the first. The sequential and full-year trajectory is genuinely good and the target is now more likely than not to be met. But the cost base grew faster than volume against the comparable quarter of last year, so the improvement is coming from the annual average and the turnaround calendar rather than from a falling absolute cost line. At roughly 740 million barrels of annual throughput each $0.10 per barrel is about $74M of pre-tax income. Measured against the second quarter alone the remaining $0.07 gap is worth about $50M annually; measured against the first half's $5.88, which is the annualized basis management is actually targeting, the $0.38 gap is worth roughly $280M. This is still the least macro-dependent lever in the company and we still capitalize it ahead of the refining margin.

5. The Macro Case Changed From Dislocation to Duration

Last quarter management framed the environment as a geopolitical shock the company had prepared for. This quarter the framing was explicitly a comparison to 2022 and an argument that this cycle lasts longer.

"And now when you look at what's going on, it's more of a supply shock than a demand shock. You've had significant refining capacity off-line and stocks are low. And so we see it taking a lot longer for that situation to normalize than what we saw in 2022."
— Mark Lashier, Chairman and CEO

The supporting inventory of facts was the most specific management has given: roughly 7 million barrels a day of refining capacity down in Asia and the Middle East plus another 1.4 million in Russia, Chinese product exports running at about 400 thousand barrels a day against 800 thousand, Chinese crude imports down from about 12 million barrels a day to about 8 million, low product inventories globally, high forecast turnaround activity in 2027 and 2028, and net refinery additions below expected demand growth.

The critical qualifier came in the same breath as the bull case. Management referred to the Strait opening and to peace breaking out as facts in progress rather than as scenarios, and the Chemicals discussion explicitly assumed normalization.

"And so whatever your view of mid-cycle was, we think it will be stronger going forward, and we think that it will be persistent going forward."
— Mark Lashier, Chairman and CEO

Assessment: The mechanism argument is stronger than it was last quarter because it now rests on named, countable capacity outages rather than on a chokepoint closure. A refinery damaged by ordnance comes back on an engineering timetable, not a diplomatic one, and that is a materially more durable claim than a closed strait. What management still will not do is convert the view into a number. It says mid-cycle is higher and persistent, and it does not say what mid-cycle now is, does not reset the Renewable Fuels mid-cycle framework it was invited to reset, and does not file the debt target it verbally lowered. A company that believed its own duration call at the multiple the market is paying would be doing at least one of those things.

6. Chemicals: Management Put a Number on the New Floor

The most valuable structural disclosure of the call, and a direct follow-through on the claim we flagged as under-scrutinized last quarter. Management's argument is that Chinese producers have lost access to deeply discounted crude, that their naphtha cost basis has permanently reset higher, and that the global polyethylene cost curve has shifted toward US Gulf Coast ethane.

"And we see that impact of about $0.07 per pound over where we saw the kind of the bottom of the cycle in 2025 at about $0.07 per pound. Put that in perspective, at the bottom of the cycle, our portion of CPChem's EBITDA was about $845 million at that $0.07 per pound."
— Mark Lashier, Chairman and CEO

Both anchors tie to the filings. Full-year 2025 chain margin was 7.1 cents per pound and full-year 2025 Chemicals adjusted EBITDA was $845M. The claim is therefore that the trough doubles, to roughly fourteen or fifteen cents, and management said explicitly that the current environment shot above mid-cycle and will come back down toward that level.

Assessment: This is a testable, quantified, non-macro claim, and it is worth more to the equity than the 43.6 cents the quarter printed. A cycle floor that doubles reprices the whole segment rather than one quarter of it, and Golden Triangle and Ras Laffan add capacity into it in 2027. Two cautions. The claim rests entirely on the persistence of sanctions and conflict-driven trade restrictions on Iranian, Russian and Venezuelan crude, so it is the same geopolitical bet as the refining margin wearing different clothes. And management observed on this same call that Venezuelan imports into the US are already up 300% since January, which is the beginning of exactly the normalization that would undo the argument.

7. Shareholder Returns Ran at 21% of Cash Flow Against a Greater-Than-50% Policy

The company returned $887M, comprising $379M of repurchases and $508M of dividends, against $4,317M of cash from operations excluding working capital. That is 21% for the quarter and 33% for the first half, against a standing commitment to return more than 50%. The first quarter ran at 111% on a depressed cash-flow base, so the half-year figure is the meaningful one.

"We expect to achieve our debt target, while also returning greater than 50% of net operating cash flow to shareholders through dividends and share repurchases. This is a core strategic priority, and we expect to increase share repurchases in the second half of this year."
— Kevin Mitchell, CFO

The 10-Q carries a disclosure the call did not. In July 2026 the Board approved a $10B increase to the share repurchase authorization, taking cumulative authorizations since July 2012 to $35B. That is not mentioned in the press release, in prepared remarks, or in any answer, including the answer to a direct question about the balance between buybacks and dividends. The company repurchased 3.9 million shares in the first half at an average of roughly $166.

Assessment: Last quarter we criticized management for defending a balanced capital allocation policy in words without funding it in dollars. This quarter it did the opposite: it under-returned against its own policy and put the money into the balance sheet, which was the better decision and the one we argued for. The awkward part is arithmetic. Returning more than 50% for the full year now requires a very large second-half payout on top of a first half that ran at 33%, and it lands in a period when management itself expects margins to normalize. A $10B authorization increase approved before the call and disclosed only in a filing is a strange way to communicate a capital-allocation intention to a market that spent the call asking about it.

8. The Commercial Organization Disclosed More Than It Ever Has

With capture back at the indicator, the commercial narrative shifted from results to capability, and the specifics were new. The time-charter fleet has expanded fourfold in two years and now supports roughly 40% of asset-backed demand while generating third-party business. The company has been granted about 20% of the Jones Act waivers issued since the current waiver took effect in March, used to substitute WTI-based crudes for foreign grades at Bayway. It is now the third-largest buyer of Venezuelan crude worldwide. Distillate production was increased by roughly 35 thousand barrels a day inside the quarter, and secondary unit utilization set a record.

"And then finally, as a result of our time charter fleet growth and increased Panama Canal transits, we now hold a favorable canal ranking. You haven't heard a lot of people talk about this. We're 26 out of 556, which allows us to schedule transits well in advance, avoid high auction fees and reduce waiting times and improve on-time reliability."
— Brian Mandell, Marketing, Commercial and Renewable Fuels

Assessment: The capability disclosures are more concrete than last quarter's and they are the kind of thing that is hard for a peer to replicate quickly: charter positions taken before rates spiked, a waiver share, a canal queue ranking. They are also, on this quarter's own evidence, worth about 98% of the indicator rather than 138%. The right way to hold both facts is that the commercial organization protects the downside of capture rather than generating a durable premium, and a business that reliably captures its indicator in a violent market is worth something. It is not worth a multiple premium on a peak indicator.

9. Midstream Growth: New Sanctions, and a Target That Now Looks Reachable

Two projects were announced inside the quarter, the 300 MMCFD Zeus gas plant in the Permian and a 100 MBD Coastal Bend NGL fractionator at Corpus Christi, and Dos Picos II reached full production at 220 MMCFD. Management reiterated the $4.5B adjusted EBITDA run-rate target for year-end 2027, of which $500M is the Midstream half of a $1B growth commitment across Midstream and Chemicals.

"So when Iron Mesa turns on, I expect us to be able to readily fill that capacity within the first part of 2027. And then those NGLs would obviously flow into our Coastal Bend pipeline expansion that comes online end of this year and fills that capacity up."
— Don Baldridge, Midstream and Chemicals

Asked whether M&A was needed to sustain growth beyond 2027, management said the opportunity set is dominated by organic projects and that the bar for bolt-ons is high, while declining to rule out divesting non-operated assets.

Assessment: The bridge to $4.5B is now $316M wide rather than roughly $1.1B, which is a genuine improvement in the pillar and the reason we move it to on track. The composition caveat stands: this quarter closed the gap on NGL prices, and Transportation, the fee-based half, grew $10M sequentially. The target is now plausibly reachable on projects alone if NGL prices merely hold, and it becomes hard again if they mean-revert with crude. The recontracting rate concession that we flagged last quarter as the number that matters for this bridge was not raised on this call by anyone.

10. Renewable Fuels Printed $544M, and Management Would Not Reset Mid-Cycle

The segment produced its best result on record by a wide margin and did so with an old mid-cycle framework of roughly $700M annually still standing. Asked directly whether that framework needs updating, management described the operational improvement and the credit environment and left the number alone.

"We had strong earnings in Q2, driven by credits and strong diesel margins, largely a function of the Iran war. Renewable diesel prices in RINs roughly doubled versus 2025 due to the Iranian situation. RINs remain an important driver to the segment's profitability. And there's ongoing regulatory policy risk, including the concern that foreign feedstock RIN generation will be cut in half after the end of next year."
— Brian Mandell, Marketing, Commercial and Renewable Fuels

Assessment: Management attributing its own best segment quarter to a war and then flagging the policy risk in the same answer is the clearest statement of non-durability anywhere on the call. Strip the $100M tariff refund and the mark-to-market and the segment still annualizes above $1.5B against a $700M mid-cycle frame. We take management's refusal to raise the frame at face value and carry the segment well below its current run-rate.

11. Western Gateway Slipped Again, and Still Has No Capital Number

In April the company expected to reach a final investment decision mid- to late summer. In August that became a month or so away, with definitive documents still being finalized.

"We do expect that we would be able to FID the Western Gateway project here in a month or so. We are finalizing the definitive documents and finishing up the details around scope and ensuring we have a solid project execution plan."
— Don Baldridge, Midstream and Chemicals

The in-service date remains the latter part of 2029. No capital figure was given on this call, and the project is not mentioned anywhere in the 10-Q.

Assessment: Two consecutive quarters of a project weeks from sanction with no capital number and a scope still being finalized. Nothing about this changes a model with a 2029 in-service date, and we continue to treat it as strategic optionality on the West Coast rather than as a modellable asset. The pattern is worth noting because it is the same pattern as Lindsey and as the Midstream recontracting concession: things that are not quantified stay unquantified until a filing forces the number out.

Guidance & Outlook

Phillips 66 does not guide to consolidated revenue or earnings, so the guidance set is operational. Read as a package, it is the most constructive the company has published in the two quarters we have covered. Six of the seven lines are flat to better, with capture guided slightly lower, and three carry a quantified improvement: turnaround expense falls by $3M to $23M, corporate costs fall by $57M to $82M, and net debt falls further. There is no equivalent of last quarter's guided ten-point drop in Chemicals utilization or forty-point drop in capture.

Implied third-quarter setup: With capture guided to roughly 95% against 98% delivered, and utilization and Chemicals utilization both broadly flat, the third quarter is a pure crack-spread quarter. Applying the first-half relationship between realized margin and refining pre-tax income, each $1.00 per barrel of realized margin is worth approximately $740M of annualized pre-tax income at current throughput, roughly $574M after tax and about $1.43 of annual adjusted EPS on 400 million shares. A single dollar per barrel is therefore worth more than the entire year-over-year improvement in corporate costs, midstream volumes and refining costs combined.

Street positioning: The dispersion going into this print, four vendors between $6.76 and $7.68, is the tell. With no pre-announcement and no company earnings guidance, the sell-side is modelling observed crack spreads with a capture assumption, and the capture assumption is now effectively published at 95%. Third-quarter estimates should converge much faster than second-quarter estimates did, and they will converge on the strip rather than on anything the company says.

Guidance style: Two quarters of data now support a consistent read. This management guides the operational lines conservatively and hits or beats them: capture, turnaround expense, corporate costs and Chemicals utilization all came in at or better than the April guide, and Chemicals utilization beat by roughly nine points. What it does not do is guide to anything that would capitalize the environment. The April pre-announcement was heavily conservative on the volatile segment and precise on the predictable ones. The same asymmetry shows up here in a different form: precise where the company controls the outcome, silent where the market does.

Analyst Q&A Highlights

Whether this cycle is 2022 again, and what normalization would look like

The first question of the call went straight to duration, using the last comparable margin environment as the reference point and asking both what is different about the market and what is different about the company. The answer separated the two cleanly and made the more durable of the two arguments about the market rather than the company.

Q: "Mark, I was wondering if you could talk a little bit about the environment and what you're seeing. The last time, Refining profitability was at this level for you and the industry was 2022. And I wonder if you could talk a little bit about what you're seeing versus that time. What the path to normalization looks like, if that's even possible to kind of envision at this point? And then how is Phillips 66 differentially positioned versus that time would also be helpful."
— Stephen Richardson, Evercore ISI

A: "And now when you look at what's going on, it's more of a supply shock than a demand shock. You've had significant refining capacity off-line and stocks are low. And so we see it taking a lot longer for that situation to normalize than what we saw in 2022."
— Mark Lashier, Chairman and CEO

Assessment: The demand-shock versus supply-shock distinction is the strongest analytical claim on the call and it is more durable than the closed-strait framing of three months ago. Physical capacity destroyed or damaged returns on an engineering schedule that can be estimated; a chokepoint reopens on a political one that cannot. The company half of the answer was weaker and mostly restated known items, with one new figure: management put the cumulative refining cost reduction at over $1.00 per barrel heading toward $1.50.

Whether the debt target should now be reset materially lower

The sharpest and most productive exchange of the call. The questioner observed that net debt had already fallen below the total debt target, that the original target had been justified on a multiple of stable EBITDA, and that a period of strong cash generation is the moment to reset it. The answer produced a new number, which is more than the equivalent question extracted three months ago.

Q: "Kevin, the step down in your net debt this quarter takes your net debt at least below your $17 billion total debt target. Now you've got line of sight to the end of the year. Previously, you justified this on a multiple of, call it, stable EBITDA. But I think whether you agree with elevated margins or not, elevated free cash flow currently, it seems to us that you've got an opportunity to reset that net debt target or that debt target substantially lower."
— Douglas Leggate, Wolfe Research

A: "$16.5 billion at the end of the second quarter, I expect that to go down between now and the end of the year. And so on a net debt level, I think of a next sort of target is something around about $13.5 billion, which would equate to a $15 billion or thereabouts balance sheet debt number."
— Kevin Mitchell, CFO

Assessment: A concession extracted, not volunteered, and it is worth roughly $3.5B of further net debt reduction beyond a target the company had 18 months to hit. The reservation attached to it is legitimate: management will not retire debt uneconomically and is constrained by the maturity schedule. The gap between a verbal $13.5B and a filed $17B is the thing to watch. If the next 10-Q carries the lower figure, this is a commitment. Until then it is a view.

The path to the $5.50 per barrel refining cost target

A question on the 2027 strategic priorities asked for an update on the operating cost reduction goal and the growth in mid-cycle Midstream and Chemicals earnings power. The refining half of the answer was the most operationally detailed of the call and the most committed language management has used on any topic.

Q: "I wanted to see if we could get a little bit of an update on your 2027 strategic priorities. You guys highlighted thoughts on shareholder returns and the balance sheet. But I wanted to see if you could maybe update us on your goal to reduce your operating costs by $500 million as well as the $1 billion growth in mid-cycle Midstream and Chemicals earnings power."
— Arun Jayaram, JPMorgan Securities LLC

A: "And as you can see in the second quarter here, we came in at $5.57, pretty close, pretty striking -- within striking range of the $5.50 number. But the annualized number is really what we're targeting, and that's what we want to -- our goal is to achieve next year."
— Rich Harbison, Refining

Assessment: The qualifier is the important half. The $5.57 is a quarterly figure in a light turnaround quarter with gas below the target basis, and the commitment is to an annualized number that absorbs the turnaround calendar and seasonal swings. The first half ran $5.88 on that basis, so the annual gap is $0.38 rather than $0.07. Management was clear about the distinction, which is to its credit, and the headline comparison that will circulate is the $5.57.

Third-quarter capture and the crack-spread setup

A two-part question on the near-term risks to crack spreads and on the capture assumption produced the most useful forward number on the call, and it came from the CFO rather than from the commercial head who had answered the macro half.

Q: "I wanted to ask near term, are we -- as we move through the remainder of summer driving season into the fall, which factors do you view as the most important upside or downside risk to crack spreads over the next quarter? Is it demand elasticity? Is it Chinese exports? And maybe putting a finer point on the capture discussion, what are your expectations at this point for third quarter capture?"
— Theresa Chen, Barclays

A: "But just specifically, as we think about capture rates, we've historically guided to about 95%, and we don't see any reason why that would be any different this year in terms of where we are -- in terms of what we see currently with regard to third quarter."
— Kevin Mitchell, CFO

Assessment: Two quarters, two guides, both to the mid-90s, and one delivered result of 98%. Capture is now a settled input rather than a debate. The tailwind list offered alongside it was long and specific, and it was all market rather than company: refineries down in Asia, the Middle East and Russia, low inventories, restrained Chinese exports, high forecast 2027 and 2028 turnarounds, and net capacity additions below demand growth. The implication for a model is that from here PSX is a leveraged, well-run holder of refining capacity, and the differentiation is the cost line and the balance sheet.

The segment split of the mark-to-market gain

A short, precise question that produced the disclosure needed to compute a clean earnings number, and a volunteered caveat that prevents a common modelling error.

Q: "And then the $450 million mark-to-market impact in Q2, I think you might have said that Renewable Fuels was $47 million boost. Do you have the same breakout for Refining and M&S?"
— Matthew Blair, TPH

A: "So $240 million Refining, Marketing and Specialties is about $160 million. And then Renewables is just shy of $50 million. And so you put those together, you get that $450 million total."
— Kevin Mitchell, CFO

Assessment: The volunteered part matters more than the numbers. Management noted that the Refining portion "is built into the indicator and so not a variance from a capture standpoint," which means the 98% capture is not inflated by the mark-to-market and that stripping the gain from earnings without stripping it from the indicator would understate capture. The disclosure was complete, immediate and unhedged, which is the same posture the company took when the equivalent number was a loss.

Whether the renewable diesel run-rate has permanently reset

An invitation to raise a mid-cycle framework that the current run-rate has left far behind. Management declined the invitation and instead attributed the result to the war and flagged the policy risk, which is the more informative answer.

Q: "Even if I was to normalize for margins closer to that $1.50 mid-cycle, you'd probably be above the $700 million that you guided to a while ago. And so just love your perspective on that business? And is there a new run rate of profitability at this level of utilization?"
— Neil Mehta, Goldman Sachs

A: "We had strong earnings in Q2, driven by credits and strong diesel margins, largely a function of the Iran war. Renewable diesel prices in RINs roughly doubled versus 2025 due to the Iranian situation. RINs remain an important driver to the segment's profitability. And there's ongoing regulatory policy risk, including the concern that foreign feedstock RIN generation will be cut in half after the end of next year."
— Brian Mandell, Marketing, Commercial and Renewable Fuels

Assessment: Handed the opportunity to capitalize its best segment result on record, management named the war as the cause and the regulator as the risk in consecutive sentences. That is a refusal to reset, and it is consistent with the refusal to annuitize the refining environment three months ago. The one durable item inside the answer was operational rather than commercial: the segment ran at 106% of nameplate after a cost and logistics restructuring undertaken when the asset's future was in doubt.

Where Chemicals margins settle when the dislocation clears

A question about the sequential softening in chemical margins and the utilization beat drew out the quantified floor that had been an unquantified assertion last quarter, along with an explicit acknowledgement that the current level is above mid-cycle.

Q: "Can you just talk about what you're seeing in the market currently? It looked like margins have come in a bit from the peak earlier this year. Can you just talk to how you're thinking about the macro set up? And then it also looked like utilization rates during Q2 came in above guidance as well."
— Joseph Laetsch, Morgan Stanley

A: "And so we see things relatively more stable, though they'll be below mid-cycle. They shot above mid-cycle temporarily. They'll be coming back down to something around $0.14, $0.15 a pound."
— Mark Lashier, Chairman and CEO

Assessment: A rare instance of management naming a normalization level rather than describing one, and it ties to filed data at both ends. The doubled floor is the durable claim; the 43.6 cents printed this quarter is not. It is worth noting that management describes fourteen to fifteen cents as still below mid-cycle, which means the mid-cycle assumption for this segment has itself moved up. That is a bigger statement than the one about the floor and it went unexamined.

What They're NOT Saying

1. A $10 billion buyback authorization increase was approved before the call and never mentioned on it. The 10-Q discloses that the Board approved the increase in July 2026, taking cumulative authorizations to $35B since 2012. It appears nowhere in the earnings release and nowhere in the transcript, including in the answer to a direct question about the balance between repurchases and dividends. An authorization is not a commitment to spend, but a $10B increase is the largest capital-allocation signal of the quarter and it reached investors through a filing rather than through the company.
2. The reset debt target exists only in the transcript. The CFO named a next net debt target of roughly $13.5B on the call. The 10-Q filed the same day still states the $17B total debt target for year-end 2027. Until the filing changes, the lower number is guidance in the weakest available form.
3. Western Gateway still has no capital number, two quarters running. The project is weeks from a final investment decision by management's own account, has moved from a "mid- to late summer" sanction to "a month or so," and does not appear anywhere in the 10-Q. Neither the total capital nor the build multiple has been given, and no analyst asked for it this quarter.
4. The Midstream recontracting rate concession has never been quantified. Named in April as one of three causes of a year-over-year decline, presented alongside ten-year-plus renewal terms, and not mentioned again by management or any analyst on this call. It is the direct offset to the $4.5B target and it remains a described headwind with no number.
5. The derivative collateral position was not disclosed this quarter. In April the company gave $3.2B posted at March 31, falling to $2.1B by April 28, and used that trajectory to support the deleveraging bridge. No equivalent figure was given for June 30, and the 10-Q states only that there was no material cash collateral that was not offset on the balance sheet. The net short position fell from roughly 50 million barrels to 36 million, which implies a large release, and the company left investors to derive it.
6. No total cost for decommissioning the California refineries. Idling costs ran $43M in Q1 and $38M in Q2, so $81M in the first half, and the third-quarter corporate guide holds an elevated run-rate. There is still no total liability, no duration, and no estimate of the redevelopment proceeds used to justify the site strategy. The Los Angeles assets were written to a $241M salvage value at the end of 2025.
7. The Lindsey acquisition was quantified in the 10-Q and explained only in general terms. The filing discloses a $115M purchase price on April 28, $202M of property recorded against $87M of asset retirement obligations, split $171M to Refining and $31M to Midstream. On the call it surfaced only in passing, as "the Prax assets," inside an answer about a low-sulfur gasoline project at Humber. There is still no earnings contribution and no explanation of which assets are being utilized.
8. No figure attached to the promised second-half buyback increase. Management said it expects to increase repurchases in the second half and reaffirmed the greater-than-50% return commitment, from a first half that ran at 33%. Neither the size of the increase nor the arithmetic that gets the full year above 50% was offered, and nobody asked.
9. The year-over-year refining cost comparison went unmentioned. Management framed $5.57 per barrel against the $5.50 target and against last quarter's $6.21. The same metric was $5.46 a year ago on more expensive gas and lower throughput. The comparison that flatters was given; the one that does not was available in the company's own supplemental table.

Market Reaction

• Pre-print setup: The shares closed at $205.89 on August 4, up 59.6% year to date against the S&P 500's 13.0%, up 68.7% over twelve months and up 16.1% over the prior thirty days from $177.33. The 52-week closing range entering the print was $118.37 to $212.27, so the stock came in roughly 3% below its 52-week closing high.
• Reaction session: Phillips 66 reports before the open, so August 5 was the reaction session. The stock opened at $209.82, a 1.9% gap higher, traded as high as $211.43, and closed at $202.55, down 1.6% or $3.34. It gave back the entire opening gap and 4.2% from the intraday high.
• Volume: 3.8 million shares against a 2.7 million 30-day average, or 1.4 times normal.
• Peer reaction: Every pure-play refiner fell on the session. VLO closed -2.1%, MPC -4.8%, DINO -6.0%, PBF -7.2% and DK -9.7%. The integrated majors were closer to flat, with XOM -1.5% and CVX -2.1%. The S&P 500 was down 0.2%.

The headline reading, that a record quarter was rejected, is wrong. Against an equal-weighted basket of the five pure-play refiners, which averaged -5.9%, Phillips 66 outperformed by roughly 4.3 percentage points and finished the best of the group. What happened on August 5 was a sector-wide de-rating, and the pattern inside it is diagnostic: the more margin-levered and less diversified the name, the worse the day. DK and PBF, the two smallest and most crack-exposed, fell four to six times as far as PSX.

That is the exact mirror of the first quarter. On April 29 Phillips 66 beat its own three-week-old pre-announcement by $408M in the segment that mattered, and closed +5.1% alongside DINO +5.1%, VLO +4.6%, PBF +4.3% and MPC +4.0%, none of which had reported. The tape paid the sector for a macro datapoint. This quarter it charged the sector for a different one, and Phillips 66 was again the vehicle rather than the subject. Across two prints, the market has treated this stock as an instrument on the crack spread rather than as a claim on a company.

The difference is that this time the company-specific result bought something. Four percentage points of relative outperformance on a day the group was marked down 6% is the market crediting a smaller balance sheet, a diversified earnings base and a cost structure, even while it re-prices the commodity. That is a more useful signal than April's, when a $408M beat produced performance indistinguishable from peers who had not opened their books.

The pre-print setup explains the vulnerability. A stock up 59.6% year to date and 16.1% in a month, entering a print 3% below its 52-week closing high, needs the guide to extend the environment rather than confirm it. Management instead guided capture down to 95%, described the Strait opening as underway, and said chemical margins would come back toward fourteen or fifteen cents from 43.6. Every one of those is the right disclosure and every one of them argues against paying a higher multiple on the peak.

Street Perspective

Debate: Has mid-cycle permanently moved higher, or is this 2022 with a different trigger?

Bull view: The bull case on the Street is that this cycle is structurally different from prior dislocations because the supply loss is physical rather than logistical. Refining capacity damaged in Asia, the Middle East and Russia returns on a repair timetable measured in years, Chinese product exports are running at half their prior rate off a permanently higher crude cost basis, inventories are low across crude and products, and the 2027 and 2028 turnaround calendars are heavy. On this reading mid-teens realized margins are the new floor and the group deserves a higher multiple on higher earnings.

Bear view: The bear camp contends that the forward market has already voted. Nymex 3:2:1 crack spreads a year forward sit more than 35% below the front month, which is the market's own statement that it does not believe the spot environment persists. Refining margins have reverted to mid-cycle after every prior dislocation, backwardation this steep has never been a durable state, and a Gulf ceasefire that holds compresses cracks from both sides as product supply resumes while crude stays bid.

Our take: The bull argument improved materially this quarter and is now better than the one made in April, because destroyed capacity is a more durable constraint than a closed chokepoint. It is still not good enough to pay for at this price. Management believes mid-cycle is higher and persistent and has declined, on two consecutive calls, to convert that belief into a target, a mid-cycle segment figure, or a filed commitment. When a management team with better information than the market says the environment is durable and then behaves as though it is not, the behavior is the signal.

Debate: Is the balance sheet still part of the bear case?

Bull view: The optimistic case is that this question is now closed. Total debt fell $6.6B in a single quarter, net debt fell to $16.5B, commercial paper went to zero, the 364-day term loan was fully retired within four months of being drawn, and debt-to-capital returned to 39% from 48%. The CFO has verbally reset the next net debt target roughly $3.5B lower. Every element of the April bridge happened faster than promised.

Bear view: The skeptical case is that the deleveraging was funded by a margin environment nobody expects to persist, that returning more than 50% of cash flow while continuing to delever gets harder as cash flow falls, and that the lower target has not been filed. A balance sheet repaired at the top of a cycle can gross up again at the next one through the same collateral mechanism, and the gross derivative book tripled this quarter even as the net position shrank.

Our take: The bulls have this one, and it is the reason our fair value range moves up. The April concern was not that leverage was high in the abstract; it was that the deleveraging path and the margin thesis were the same bet, so the balance sheet offered no protection if the cycle turned. The company has now retired the debt while the cycle was strong, which is precisely the sequence that breaks the correlation. What remains is a policy question rather than a solvency one, and the answer to it is the buyback pace in the second half.

Debate: What is the commercial organization actually worth?

Bull view: A view still circulating is that the asset-backed trading operation deserves a structural premium. The disclosures this quarter were more concrete than ever: a time-charter fleet expanded fourfold covering roughly 40% of asset-backed demand, roughly a fifth of all Jones Act waivers granted since March, third-largest global buyer of Venezuelan crude, and a Panama Canal queue position of 26 out of 556.

Bear view: The counter is arithmetic. Worldwide capture went from 138% to 98% in one quarter and the Atlantic Basin, the engine of the first-quarter number, went from 182% to 79%. Management now guides capture to roughly 95% as a standing convention. A capability that delivers the indicator is table stakes, not alpha, and trading earnings carry a low multiple wherever they appear.

Our take: The bears have the multiple question and the bulls have the risk question, and both can be true. Two quarters of data bracket capture in both directions and average to roughly 118%, which is above the indicator but nothing like the first-quarter headline. We would underwrite this organization as insurance against underperforming the crack rather than as a source of durable outperformance, and we continue to capitalize the refining cost program, which is controllable and non-cyclical, ahead of it.

Model Update

Two quarters of data now anchor the framework we established at initiation. The changes below are material and they run in the company's favor.

Non-refining build. The $3.60B constant is the sum of Midstream at $2.90B (against $2,828M in 2025 and a $2,752M first-half annualized), Chemicals at $0.45B (against $328M in 2025 at a 7.1 cent chain margin, marked to management's mid-teens floor), Marketing and Specialties at $1.45B (against $1,841M in 2025 and roughly $1,072M annualized on a first-half clean basis), Renewable Fuels at $0.45B (against a $380M loss in 2025 and roughly $946M annualized clean, held near management's own unrevised mid-cycle frame), less Corporate and Other at $1.65B (against $1,628M annualized this quarter, with the third-quarter guide implying lower).

Earnings power by realized refining margin

The framework flexes Refining off the realized margin and holds non-refining constant. Refining pre-tax income equals 740 million barrels multiplied by the realized margin less the $8.30 per barrel cost-and-depreciation wedge. The result is taxed at 22.5% and reduced by $120M of noncontrolling interests, over 400 million diluted shares.

What the price requires. At $202.55, holding the multiple at 10x to 12x trend earnings, the shares need a sustained realized refining margin of $15.42 per barrel at 12x and $17.77 per barrel at 10x. The comparisons are $24.08 in the quarter just reported, $17.21 for the first half, $14.77 for the trailing twelve months and $10.88 for full-year 2025.

Valuation snapshot at the $202.55 reaction close: market capitalisation of approximately $80.8B on 399.0 million shares outstanding at June 30, net debt of $16,466M, and an enterprise value of approximately $97.3B. That is 7.9 times trailing-twelve-month adjusted EBITDA of $12,247M, 13.6 times trailing adjusted EPS of $14.89 and 11.5 times trailing GAAP EPS of $17.55. Every one of those multiples is lower than at the first quarter, when the same measures read 10.3x, 22.1x and 17.1x at a $173.49 share price. The stock got cheaper on trailing data while rising 17%, which is what a cycle looks like at the point of maximum earnings.

Valuation impact: Raising the fair value range to $148 to $212 from $110 to $166, corresponding to sustained realized margins of $14.00 to $16.00 per barrel at 10x to 12x trend earnings. The band moves up because the non-refining base is genuinely larger and the cost wedge genuinely smaller, and because two quarters of evidence support a sustained margin above the $13.00 to $15.00 we carried at initiation. The $202.55 close sits in the upper half of that range.

Thesis Scorecard Post-Earnings

Overall: The thesis is stronger than it was at initiation, on the specific axis that mattered most. The argument for a Hold three months ago was that the leverage and the cycle were a single correlated exposure, so the balance sheet offered no protection against the thing most likely to go wrong. That is no longer true: the company retired $6.6B of gross debt inside one quarter, while the cycle was strong, which is exactly the sequence that decouples the two risks. Two of the four bull pillars strengthened and the primary bear point moved to contained. Against that, the quarter also removed most of the ambiguity about where the earnings come from. A 98% capture on a $24.08 margin is a market result, and management spent the call describing the market's normalization while declining to reset a single target that would capitalize it.

Action: Hold. Two of the three upgrade triggers set at initiation were met, and we have raised fair value by roughly a third to $148 to $212 in response. The reason this is not an upgrade is that the share price moved 16.7% over the same interval and consumed the entire improvement: at $202.55 the shares require a sustained $15.42 per barrel at 12x against $14.77 for the trailing twelve months, which is fairly valued rather than cheap, in a market whose own forward curve prices a decline of more than a third. We would move to Outperform on a de-rating into the $160s, on the reset debt target appearing in a filing alongside a second-half buyback that carries the full year above the 50% commitment, or on evidence that the Chemicals floor reset and the Midstream project slate hold the non-refining base above $4B while refining normalizes. We would move to Underperform if realized margin falls back toward the full-year 2025 level of $10.88 while the shares hold a 12x multiple on peak earnings, which is a $104 to $124 outcome on this framework.