Halliburton Is Selling 2027 Before Its Customers Have Written a Budget

HAL front-runs budget season with a margin promise that leans more on SAP roll-off than on activity — and quietly shelves the technology story that carried its 2025 calls.

HAL · Earnings Call · 2026-10-20

The tell is the calendar

Halliburton's third quarter was fine, not great: revenue of $5.7B came in above plan, adjusted operating margin of 12.8% did not, C&P missed on both line items while D&E beat. Then management did something it has not done in this cycle — it handed out a 2027 framework in October, before a single major customer had set a budget. JPMorgan's Arun Jayaram said the quiet part out loud: “fairly unusual for how to... provide kinda outlook comments at this time of the year, particularly prior to budgeting season from the big oils integrated and E and Ps.” — Arun Jayaram, Analyst · 2026-10-20 Jeffrey Miller's answer was about as candid as a CEO gets: “I do not like where margins are today.” — Jeffrey Allen Miller, Chairman, President and CEO · 2026-10-20 That is the whole call in one line, and it is why the top-ranked keyword on this transcript is early view — a phrase that exists because management decided the market needed persuading now rather than in January. The setup matters: HAL has spent the last 90 days falling 15.5% from a June peak of $39.73, so this was a company talking up its margin future into a tape that has been voting the other way.

Half the margin math is bookkeeping

Eric Carre walked the bridge with unusual precision. Roughly $22.5B of 2026 revenue, about $1B of growth, and 200 basis points of margin is $500M–$600M of operating profit.

200 basis points in margin, means about $500 to $600 million in operating profit. So about half of it is going to come from cost savings, operating cost savings, structural cost savings and the ramp down of the SAP S/4 program.
Eric J. Carre, Executive Vice President and CFO · 2026-10-20

Which means the hard, contractually scheduled part is the $100M–$150M of rightsizing plus $120M–$150M of S/4 savings — roughly $250M–$300M a year that arrives whether or not the oilfield gets busier. The other half rides on incremental margins of just 20% to 30%. That is the number to sit with, because in July, Scott Gruber asked whether HAL could run above its historical 30%–35% incrementals, and Miller answered: “your incremental expectations are not wrong. Those are my expectations as well.” — Jeffrey Allen Miller, Chairman, President and CEO · 2026-07-21 Pressed further, he added “Can we do better than normal on incrementals... Yes. I mean, I think so.” — Jeffrey Allen Miller, Chairman, President and CEO · 2026-07-21 Three months later the working assumption embedded in the flagship 2027 guide is a haircut to exactly that metric. The incremental margin assumption has quietly been de-rated at the same time the headline margin target was upgraded. Carre was explicit that “we are not assuming meaningful pricing” — Eric J. Carre, Executive Vice President and CFO · 2026-10-20 — so 2027 is a cost-plus-modest-volume story, deliberately engineered so that even a flat world produces the number.

North America as an assumption, not a thesis

The structural claim of the call is that Halliburton no longer needs hydraulic fracturing to heal.

Halliburton's path to growth and improved profitability that I just described does not require any recovery in North America frac. To us, it is purely upside.
Jeffrey Allen Miller, Chairman, President and CEO · 2026-10-20

That is a real strategic statement and it is also a hedge dressed as confidence — the guide assumes flat North America, while the commentary insists “we run it for absolute returns, and we will raise price or we lose share.” — Jeffrey Allen Miller, Chairman, President and CEO · 2026-10-20 The anecdotes are genuinely interesting (a tender where only one service company bid; a customer extending an e-fleet at higher price and adding another), but note the keyword migration: white space was the single top-ranked phrase on the July 2025 call, back when the debate was how quickly the market would tighten. Three quarters of deliberate choices later, the conversation has migrated from a thesis to a guidance assumption. When a management team stops arguing a market will improve and starts proving it doesn't need one, that is informative in both directions.

What stopped being said

The loudest signal on this call is an absence. ZEUS IQ was the number-one keyword on the April 2025 call; data centers topped the list in October 2025 with electric fleets right behind. In this quarter's keyword set, distributed power, VoltaGrid, e-fleets and closed-loop fracturing are essentially gone from the top of the transcript — replaced by cost reduction, structural cost, and S/4. AI does appear, but as a back-office lever — "adopting AI at scale in our back office functions" — not a revenue product. That is a demotion of the growth narrative from offense to overhead. It may also be coincident with the tape: AI data center is the single largest theme in the market's 360-day curated movers, yet over 90 days it is down 5.5% with six names up and forty-seven down. The theme rolled over and Halliburton stopped selling it in the same window. Nothing on the call resolves whether the power venture stalled or the market did. Also worth noting: the global top keywords this quarter skew heavily financial (fee earners, mortgage universe, retail deposits, loan beta) with no energy theme at all — so HAL is not riding a market wave here. This is an idiosyncratic margin reset, not a sector rotation.

The Venezuela number that moved ten-fold

Shannon Slocum, on the optionality: “before we left Venezuela, it was greater than a $5 billion business.” — Jeffrey Shannon Slocum, Executive Vice President, Global Business Development and Marketing · 2026-10-20 In January, Miller described the same pre-exit business: “A decade ago, it was probably a half $1 billion business for us.” — Jeffrey Miller, Chairman, President and CEO · 2026-01-21 Both statements describe the size of HAL's Venezuela book before the withdrawal. That is an order of magnitude, from the same management team, nine months apart. Venezuela is a legitimate optionality — facilities open, a dozen-plus ESPs going to a major this quarter — but when a number inflates ten-fold in the retelling, treat it as sentiment rather than forecast. The Middle East is more soberly framed: roughly 3 billion barrels of supply lost since the conflict began, about 1 billion drawn from inventories, and a 2027 assumption of activity flat to the second half of 2026. Miller's answer on resolution — "dare to dream" — was the most honest thing said all morning.

What the numbers actually confirm

Revenue has been parked near $5.7B for six consecutive quarters, which is why the 2027 case needs a step up rather than a trend continuation. Total revenue of $5.7B, up 4% year over year but 32% below the 2014 peak. The margin maths is similarly a round trip: Operating margin of 13.6% in the June quarter, guided to 12.8% for Q3 and "about 15%" for 2027 — against a 16%–18% run rate through 2022–24, and a peak above 18% in 2011. "About 15%" is not virgin territory; it is the restoration of a level the company held two years ago. The balance sheet is doing its job: Effective net cash of -$5.1B, with roughly $1.6B of free cash flow guided for the year and $250M of stock retired in the quarter. The equity has re-rated on a depressed base rather than on growth.

The bottom line

The 2027 framework is credible precisely because it does not require a market: half of the promised margin is cost the company controls, and the other half assumes below-normal volume leverage, which is a low bar to clear. What has been traded away is the story. Halliburton spent 2025 selling oilfield services as a technology franchise — Zeus, e-fleets, closed-loop automation, distributed power — and in October 2026 it sold a cost programme. That may be exactly the right pitch for a stock down 15% in 90 days with margins below where the CEO wants them. But investors should be clear which security they now own: a scheduled cost cut with a free option on North American frac, rather than a growth compounder waiting on a cycle.