There are reports whose top line misleads, and this is one of them. Baker Hughes (NASDAQ: BKR) reported revenue down 2% and GAAP EPS down 3% - and yet this is one of the most significant reports of the season. The reason sits in an entirely different line: orders jumped 49%, and the energy-infrastructure division doubled its own.
About the Company: Two Businesses Under One Roof
Baker Hughes is publicly perceived as an "oil company," but in practice it consists of two businesses currently moving in opposite directions - and that is the whole thesis of the report.
Oilfield Services and Equipment (OFSE) - the traditional business: drilling, well completion, equipment and services for oil companies. A business dependent on the oil price and on producers' exploration budgets.
Industrial and Energy Technology (IET) - and here sits the story. This division manufactures gas turbines, compression equipment and systems for generating and moving energy. These are precisely the products required to build new electricity generation capacity - the same world in which GE Vernova operates.
To understand why this matters: a large data center consumes electricity on the scale of a small city. It cannot simply be plugged into the existing grid; additional generation capacity must be built, and a gas turbine is the fastest way to do that. Hence the demand.
What Was Reported: The Split That Tells Everything
The overall picture - mixed. Revenue fell 2% to $6.7 billion, and GAAP EPS fell 3% to $0.68. But adjusted EPS rose 2% to $0.64, and adjusted EBITDA reached $1.231 billion - above the high end of the company's own guidance.
And the cash. Free cash flow jumped to $1.109 billion, versus just $239 million in the comparable quarter.
And the figure that decides: the orders.
| Metric | Amount | Change |
|---|---|---|
| Total orders | $10.5 billion | +49% |
| Total backlog (RPO) | $40.1 billion | +$4.0 billion in the quarter |
And here is the split that is the heart of the report:
| Segment | Orders | Revenue | EBITDA |
|---|---|---|---|
| Oilfield services (OFSE) | $3.41B (-3%) | $3.45B (-5%) | $605M (-11%) |
| Energy technology (IET) | $7.09B (+101%) | $3.29B (flat) | $678M (+16%) |
The energy division doubled its orders in a single year. Its backlog reached a record $37.1 billion - meaning almost the entire backlog of the whole company is concentrated in it. And for the first time, its EBITDA is larger than the oil division's, even though its revenue is lower.
What the CEO Said
And the company left no room for interpretation about the source of the demand. The CEO summed up: "another strong quarter, reflecting the breadth of the portfolio and continued momentum in data center markets, gas infrastructure and upstream."
An oilfield services company listing data centers first among its engines - that, in itself, is a sign of where the market is moving.
The Outlook Was Expanded
The company raised the annual order guidance for the energy division, and expanded its horizon outlook for 2026-2028 to more than $45 billion in orders for that division - against what it describes as broadening demand and decisions to expand production capacity.
The Connection to Our Thesis
This report is the third confirmation within a week of the thesis we have been tracking all summer.
First, GE Vernova jumped its gas-turbine backlog from 100 to 116 gigawatts in a single quarter. Second, NextEra added 3.6 gigawatts to its backlog. And now Baker Hughes - a company not perceived as an electricity player at all - reports that orders on its energy-infrastructure side have doubled.
Three independent companies, three separate reports, the same conclusion. This is no longer one company's story; it is an industry pattern.
And no less important - the internal split. That same company shows the traditional oil side contracting while the energy-infrastructure side surges. This is exactly the split we saw Friday at Schlumberger, where the international business was hurt by the Middle East conflict while North America jumped. The world is not leaving energy - it is moving money within it, from the drill to the grid.
The Bull Thesis
Whoever reads it positively will point to an especially strong combination: orders that jumped 49%, a $40 billion backlog providing visibility for years, and free cash flow that grew more than fourfold. At an equipment company, the backlog is the most important figure - it tells what has already been sold and is due to be delivered.
Beyond that: the company sits at a junction receiving structural rather than cyclical demand. Building electricity generation capacity for data centers is a multi-year project, and the $37 billion backlog in the energy division embodies exactly that. Expanding the order horizon to $45 billion by 2028 widens the picture further.
The Bear Thesis
Whoever reads it critically will point out that actual revenue is still falling. Orders are a promise; revenue is realization. The gap between the two is large today, and converting the backlog depends on production capacity, supply chains and project timetables.
Second, the oil side is genuinely contracting. EBITDA falling 11% in a division still representing about half of revenue is not negligible, and it depends on the oil price - a variable outside the company's control, and one last week showed to be volatile.
Third, demand concentration. If data-center investment moderates, the division that led the quarter will be hurt first, and recorded orders may be postponed.
And fourth, execution risk. Doubling orders also means needing to expand production capacity. Equipment companies have historically struggled to grow fast without hurting margins.
The debate in one line
The bulls see orders that jumped 49%, an energy division that doubled its own with a record $37 billion backlog, free cash flow up fourfold, and a CEO naming data centers as the engine. The bears see revenue still falling, an oil division contracting 11% in EBITDA, and a large gap between order and realization. Both sides are reading the same report.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
This report is, to my eye, the strongest confirmation our thesis has received this summer - precisely because it came from the unexpected place.
When an electricity equipment company reports strong data-center demand, that is expected. When an oilfield services company reports that orders in its energy division doubled, and that the CEO opens the list of engines with data centers - that is something else. It means the demand is strong enough to shift the center of gravity of an entire company.
And the internal split is what I take from here. In that very same report: drilling down 11%, energy infrastructure up 16%. This is a map of capital movement inside a sector - not money leaving energy, but money migrating within it, from the well to the socket.
And what I will follow: the gap between orders and revenue. $10.5 billion of orders against $6.7 billion of revenue is a machine being loaded faster than it fires. If in coming quarters revenue starts climbing toward that pace - the thesis materializes. If the backlog grows while revenue stays put, that will be a sign of a bottleneck in production - not in demand.
Summary
Baker Hughes delivered a report whose top line misleads. Revenue fell 2%, but orders jumped 49%, the backlog reached $40 billion, free cash flow grew more than fourfold, and the energy-infrastructure division doubled its orders and reached a record $37.1 billion backlog.
Beyond the company itself, this is a data point that joins GE Vernova and NextEra and completes a pattern: demand for infrastructure that generates and moves electricity is not one company's story, and it is already shifting the center of gravity of the entire energy industry. The question that remains open is not whether the demand exists - but how quickly it converts from orders into revenue.
Sources: Baker Hughes's official results announcement for the second quarter of 2026 (July 26, 2026), including revenue, GAAP and adjusted earnings, adjusted EBITDA, orders, backlog, segment breakdown, free cash flow, updated guidance and the CEO quote. Data accurate as of the time of writing. The chart is shown in real time via TradingView. Nothing herein constitutes a forecast, recommendation or advice - see the full disclaimer at the bottom of the page.