The two American oil majors reported the same morning, on the same quarter, with exactly the same oil price. One beat consensus by about 4%, the other missed by about 4%.
And the difference between them explains something about this quarter that the headline number hides.
The Backdrop: What Happened to the Oil Price
Brent averaged $104 a barrel in the quarter - against $81 in the first quarter and $68 a year ago.
That is a jump of more than 50% within a year, and it stems from geopolitical escalation: U.S. strikes on Iran, severe disruption to traffic through the Strait of Hormuz - through which roughly a fifth of global oil shipments pass - and risk spreading to the Red Sea as well. In parallel, the U.S. strategic petroleum reserve fell for an 18th consecutive week, to its lowest level since 1983.
That is the backdrop that made this one of the sector's strongest quarters in years. And it is also the backdrop that separated the two companies.
What Was Reported
| Metric | Chevron | Exxon |
|---|---|---|
| Net earnings | $12.1 billion | $14.5 billion |
| EPS (GAAP) | $6.11 | $3.48 |
| Adjusted EPS | $6.06 | $3.52 |
| Consensus | about $5.81 | $3.68 |
| Against the estimate | a beat of ~4% | a miss of ~4% |
| Cash from operations | $22.6 billion | $23.6 billion |
| Free cash flow | $18.1 billion | $17.2 billion |
| Return on capital employed | 21.4% | - |
| Returned to shareholders | about $6.6 billion | $9.4 billion |
Note the first row against the second: Exxon earned $2.4 billion more - and yet its earnings per share is nearly half of Chevron's.
Why EPS is inverted relative to total earnings
The explanation is simple and structural: the share count.
Exxon is a much larger company - a market capitalization of roughly $650 billion against about $383 billion for Chevron - and it has roughly twice as many shares. So the same dollar of profit is divided across a wider base.
The practical conclusion: when comparing these two companies, EPS does not tell you who earned more. It tells you how much of it reached each share.
Which is why the correct comparison is against each one's own expectations - and there the picture is unambiguous: Chevron cleared the bar, Exxon did not.
Chevron: Record Production, and Refining That Worked
Chevron's earnings break down as follows:
| Segment | Q2 2026 | Year ago |
|---|---|---|
| Upstream - international | $4,641 million | $1,309 million |
| Upstream - U.S. | $3,541 million | $1,418 million |
| Downstream - international | $2,457 million | $333 million |
| Downstream - U.S. | $2,411 million | $404 million |
| All other | $(978) million | $(974) million |
Two things stand out here.
The first - refining. Chevron's downstream earnings rose roughly sixfold in the U.S. and about sevenfold internationally. The company reports 97% utilization at its U.S. refinery crude units and record throughput - meaning it captured the high margins at full output.
And the second - production. Worldwide output rose 20% to 4,070 thousand barrels of oil-equivalent per day, with record U.S. production. International liquids realizations jumped from $58.88 to $96.41 a barrel.
In other words: Chevron got both the price and the volume.
Exxon: Same Price, but the Disruption Hit It
And at Exxon the quarter looks different - and the company says so itself.
Its official headline cites "the highest Upstream production in more than two decades" - and immediately adds five words: "excluding the Middle East disruptions."
The five words that explain the gap
A footnote in the release clarifies what that means: the disruption analysis is based on excluding Middle East country volumes across all periods. In other words - to reach a two-decade high, an entire region has to come out of the calculation.
And in the quarterly walk the company writes explicitly that upstream earnings improved thanks to record Permian production and the absence of disruptions in Kazakhstan, "partly offset by the Middle East disruptions."
And that is the point: the very same geopolitical event that pushed oil to $104 is the one that hit Exxon's volumes.
Chevron got the price without paying the price. Exxon paid both.
And in full fairness: Exxon still earned $14.5 billion in the quarter and returned $9.4 billion to shareholders. A "miss" here is relative to expectations, not to performance.
And what did work at Exxon:
- Record Permian production - above 1.8 million barrels of oil-equivalent per day, consistent with a planned 9% compound annual growth rate through 2030
- The fifth Guyana FPSO set sail, with production startup planned for the fourth quarter - adding 250 thousand barrels per day of capacity
- Record second-quarter diesel production
- Cumulative structural cost savings of $16.3 billion
And CEO Darren Woods summed it up: "The second quarter was shaped by disruption, but defined by execution."
And the Figure That Is Not Oil at All
And inside an oil company's results, a line appeared that is effectively a power infrastructure deal.
Chevron signed a 20-year power purchase agreement with Microsoft, to supply about 2.67 gigawatts of dedicated behind-the-meter electricity from a facility to be developed in West Texas, for a Microsoft data center.
Why this matters far beyond Chevron
"Behind the meter" means the power is generated at a dedicated facility and reaches the customer without passing through the public grid.
And that bypasses precisely the bottleneck we described yesterday: in the electricity sector piece we saw AEP reach 69 gigawatts of contracted demand through 2030, and regulated utilities required to commit enormous capital and win regulatory approval before they see a return.
A private behind-the-meter facility does not wait in the interconnection queue and does not need a rate approval.
Which is why this direction is significant: if hyperscalers can buy power directly from an oil and gas producer instead of waiting for a utility, part of this demand wave simply routes around the regulated system - and moves to players who know how to build generation quickly.
And that is also why CEO Mike Wirth framed it the way he did: to power "American AI dominance" and generate "resilient cash flows." An oil company talking about AI on its earnings release - that itself is the data point.
And one caveat: 2.67 gigawatts is a facility that still has to be built. The agreement is signed; the power is not yet flowing.
The Bull Thesis
Whoever reads it positively will see a sector generating cash at an unusual pace: Chevron with $18.1 billion of free cash flow and a 21.4% return on capital employed, and Exxon with $17.2 billion and $9.4 billion returned to shareholders in a single quarter.
Beyond that: both are using the quarter to fund long-term investment - Chevron in AI power infrastructure and framework agreements with Iraq, Exxon in Guyana and the Permian. This is not merely a quarter of passing strength; it is a quarter that funds the next decade.
The Bear Thesis
And the big risk is exactly what created the quarter: the price.
$104 a barrel is not a normal level - it is the result of a supply disruption. When the disruption resolves, the price falls, and these profits do not come back. Anyone pricing these companies off this quarter is pricing a cyclical peak.
Second, the operational risk is not theoretical - Exxon just demonstrated that the same event which lifts prices also halts volumes.
And third, the stocks have already run: Exxon +30.7% and Chevron +26.4% year-to-date. A considerable part of this quarter is already priced in.
The debate in one line
The bulls see two companies generating together more than $35 billion of free cash flow in a quarter, returning capital at an unusual scale and investing in long-term growth - including Chevron's entry into supplying power for AI. The bears see profits resting on a $104 oil price created by a supply disruption, operational risk Exxon just demonstrated, and stocks already up about a third this year. Both sides are reading the same reports.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
What catches me in these two reports is how misleading the word "sector" is.
These two companies trade in the same index, respond to the same barrel price, and sit in the same ETFs. And in one quarter, on exactly the same geopolitical event, one beat and the other missed - not because of better management, but because of where their assets sit on the map.
That distinction is worth more than any oil price forecast. Anyone who bought "energy exposure" actually bought two very different exposures to the same event: one that benefits from it fully, and one that gets the price and pays in volumes.
And what I find most interesting in these reports is precisely the line that is not oil. A 20-year agreement for 2.67 gigawatts behind the meter for Microsoft is not a side deal - it is a statement of direction. An oil company that looks at its spare generation capacity and land in Texas and sees a data center there is pricing its assets differently from how the market prices them.
And that connects directly to what we saw yesterday in the electricity sector. When the regulated utilities fell despite 69 gigawatts of contracted demand, we explained that their investment is an expense that returns through rates over decades. A private behind-the-meter facility is the model that routes around exactly that - and if it catches on, it changes who profits from this power wave.
And the point of caution I hold: $104 a barrel is not a stable state. The numbers this quarter are especially handsome because the price is especially unusual, and a quarter resting on a supply disruption is not a quarter you can multiply by four.
And what I will watch: whether Chevron's power agreement stays singular or becomes a series. If more energy companies sign direct agreements with hyperscalers, then a second revenue channel has opened for the sector - one that does not depend on the price of a barrel. And if it stays a one-off, it was a convenient deal rather than a trend.
Summary
Chevron reported earnings of $12.1 billion, $6.11 per share ($6.06 adjusted) against a consensus of about $5.81 - a beat of roughly 4%. Worldwide production rose 20% with record U.S. output, refinery utilization reached 97%, and return on capital employed 21.4%.
Exxon reported earnings of $14.5 billion, $3.48 per share ($3.52 adjusted) against a $3.68 consensus - a miss of roughly 4%. Production was the highest in more than two decades excluding the Middle East disruptions, with record Permian output above 1.8 million barrels per day and free cash flow of $17.2 billion.
And above both: Brent averaging $104 a barrel against $68 a year ago - a jump created by a supply disruption, which acted on the two companies in opposite directions.
And Chevron signed a 20-year agreement to supply about 2.67 gigawatts to Microsoft. Inside an oil company's results, that may be the line that stays relevant longest.
Sources: the official second-quarter 2026 results releases of Chevron Corporation and ExxonMobil Holdings Corporation, as filed with the U.S. Securities and Exchange Commission on Form 8-K on July 31, 2026, including net and adjusted earnings, earnings per share, the segment breakdown, production and realization data, cash flows, shareholder distributions, the power purchase agreement with Microsoft and the notes regarding the Middle East disruptions; average Brent prices from Chevron's highlights table (source: Platts); analyst consensus ahead of the reports as published in the financial press; year-to-date return figures accurate as of the time of writing. The charts are shown in real time via TradingView. Nothing herein constitutes a forecast, recommendation or advice - see the full disclaimer at the bottom of the page.