XOCompany report
Exxon Mobil Corporation XOM
The bet you're really making is that oil stays high enough for Exxon's cheapest barrels, the ones in West Texas and off the coast of Guyana, to keep throwing off cash. You're betting Exxon keeps pumping more of those cheap barrels while the expensive ones run down, and that when oil crashes, and it always does, Exxon is big enough to pay its dividend and buy back stock straight through it. Right now it looks better than it is: profit more than doubled to $14.5 billion because oil prices jumped, while the total barrels it pumped actually shrank 2.5%. You pay about 20 times last year's earnings, near the top of what it has cost in twelve years, when it usually ran 12 to 19.
Key data
XOM · price with moving averages
Source: market data.
The business
Exxon pumps crude oil and natural gas out of the ground, then turns much of it into fuels and chemicals it sells worldwide. The money is made mostly upstream, at the wellhead: a barrel that costs Exxon a low-teens dollar figure to lift and sells for whatever the world price is that day, with the spread being the profit. The edge is owning some of the lowest-cost barrels on Earth, the unconventional acreage in the Permian of West Texas and the offshore Stabroek block in Guyana, and running them at a scale that lets the refining and chemical arms soak up the swings when crude alone stumbles. The thing a driver would recognize is the pump price; the thing that decides this stock is the gap between what a barrel fetches and what it cost Exxon to get it out.
The numbers
The last five quarters show EPS more than doubling in Q2 2026 while the prior four ran lower and flatter.
| Quarter | Revenue | Net income | Diluted EPS |
|---|---|---|---|
| Q2 2025 | $79.5B | $7.08B | $1.64 |
| Q3 2025 | $83.3B | $7.55B | $1.76 |
| Q4 2025 | $80.0B | $6.50B | $1.53 |
| Q1 2026 | $83.2B | $4.18B | $1.00 |
| Q2 2026 | $114.5B | $14.5B | $3.48 |
Revenue rose 42% in Q2 2026 versus the same quarter a year earlier and profit 105%, while EPS climbed 112% because buybacks shrank the share count underneath it. Almost none of that was more oil out of the ground. The production detail that was missing from the read a day ago now sits in the filing clean, and it cuts the other way: worldwide output was 4,514 thousand oil-equivalent barrels a day, down 2.5% from a year earlier, even as Permian-led US liquids climbed 10.6%. The swing was price, plus a natural-gas decline in Asia dragging the total down while liquids grew. The June print landed just shy of Street expectations, $3.48 against about $3.56 hoped for, so the doubling was the cycle, not a surprise from operations.
Where the low-cost barrels earn their keep is cash. Half-year numbers, from the filing:
| 1H 2026, to June | $B |
|---|---|
| Operating cash flow | 32.3 |
| Dividends paid | 8.6 |
| Buybacks | 10.0 |
| Total returned | 18.6 |
That is about $37B a year handed back, near a 5.6% cash-return yield on a $661B company, on top of a fortress balance sheet: net debt about 0.4x cash earnings, interest covered dozens of times. The variant here is unglamorous. The market is pricing this $14.5B quarter close to a run-rate, but roughly half of it is an oil price Exxon does not set, and the buybacks are being run at the richest valuation the stock has carried in a decade. The single print that settles which it is: next quarter's per-barrel realization at a lower crude price against the barrel count.
Management
The record is capital-return discipline more than empire-building. Insiders bought nothing over the last twelve months and sold about $2.4M across seven small dispositions, the largest an officer's $0.6M in February, plan status not disclosed in the vendor feed, so read it as routine rather than a signal either way. The dividend has been paid and raised through every oil crash for decades, and $8.6B went out the door in the first half. The one thing to weigh is the $10.0B of stock bought back in that same half at near the top of a twelve-year valuation band: real cash spent shrinking the count at a full price, which flatters EPS today and costs more per share than it did through most of the last decade.
How it fails or surprises you
Oil below the line. At crude in the low $60s, upstream cash thins and the roughly $37B a year of dividends and buybacks stops being covered by operations. Exxon has funded the payout through past troughs on its balance sheet, but a multi-year slump forces a choice between the buyback and the debt line. Watch quarterly operating cash flow against the $8.6B half-year dividend.
The cheap-barrel ramp (right tail). Guyana and the Permian are the lowest-cost barrels Exxon owns, and their share of output keeps rising. If those volumes grow through the back half of the decade while unit costs fall, the company throws off more cash at the same oil price than the market pays for today. The print: US liquids volume and per-barrel cost, each quarter.
Profit that is price, not barrels. The fact the read fits least: profit more than doubled while total output fell 2.5%. That gap is the oil price, which Exxon does not control. If crude normalizes next quarter and volumes stay soft, earnings halve from this base. Watch Q3 realization per barrel alongside worldwide oil-equivalent production.
Closing thoughts
The oil price swings the outcome and Exxon does not set it. The question is whether the company survives the next crash, and the balance sheet says yes: net debt near 0.4x cash earnings, interest covered many times over, and a dividend paid and raised through every downturn in living memory. The left tail is a long oil slump, and it dents returns without threatening the company. The right tail is the low-cost barrels of Guyana and the Permian compounding volume as the rest of the field declines. For a multi-year hold the right tail is the fatter one, because those barrels are real and growing while today's earnings are partly borrowed from a high crude price.
The bet is still that oil stays high enough for Exxon's cheapest barrels in West Texas and off Guyana to keep throwing off cash, more of those barrels pumped each year, and a balance sheet strong enough to pay you through the next crash. It breaks when the oil price and the barrel count fall together, so the one pair to watch is realization per barrel against worldwide production. The next low-oil quarter will show how much of this profit was the company and how much was the price.
Methodology
The year-to-date row is the sum of the 2 reported quarters of the current fiscal year, diluted EPS included; the five-quarter and five-year tables are the vendor income statements.
Sources: Q2 2026 10-Q (filed 2026-08-03, period ended 2026-06-30), as-filed XBRL for Q2 2025 and Q2 2026 revenue, net income, EPS, production, cash flows, dividends and buybacks; consensus actual EPS for Q3 2025, Q4 2025, Q1 2026; market, valuation and insider data from the evidence pack current to Sep 6, 2026.
Basis: revenue, net income and diluted EPS are as-filed for Q2 2025 and Q2 2026; Q3 2025, Q4 2025, and Q1 2026 revenue and net income not in the 10-Q and cannot be derived from fiscal year less nine months (FY 2025 data not in evidence pack); EPS for those three quarters from consensus actual results. Production and 1H cash figures read from the filing this run; multiples derived from the pack and tie out within rounding; FY28 forward P/E uses consensus EPS of $10.36.
Fact check: Q2 2026 revenue $116.0B, net income $14.5B, diluted EPS $3.48, and Q2 2025 comparatives verified as-filed. Production 4,514 kboe/d (-2.5% YoY) and US liquids +10.6% read from the 10-Q. 1H operating cash flow $32.3B, dividends $8.6B, buybacks $10.0B from the cash flow statement. Insider sales $2.4M/7 sales from vendor data; no 10b5-1 status disclosed. All numerical claims reconciled to filed 10-Q or evidence pack. 0 errors found. Final analysis verified as of Sep 6, 2026.
Lens (energy): emphasis on cash returned versus reinvested, cost position of the marginal barrel, and balance-sheet strength through a downturn; reserve-life and replacement figures were not in the pack this run.
This is analysis, not advice.
Bid Cap
Daily ideas, a 390-name database, and a model long/short book from an investor who mostly covers financials. $70 a month or $700 a year.
Subscribe on Substack


