Listen to the audio version of this article (generated by AI).
Earnings season has a way of cutting through the noise. This week, it cut the economy in two.
On one side: AI infrastructure stocks posting blowout results, raising guidance, and struggling to keep up with demand.
On the other: traditional companies cutting forecasts under the weight of rising costs.
Same economy. Radically different realities.
The AI Boom is not just surviving the current macro turbulence. It is accelerating through it. And everything tethered to old-economy inputs — energy costs, consumer discretionary spending, rate-sensitive balance sheets — is getting crushed under the weight of them.
A great “AI Bifurcation” is beginning to split the economy — and market — in two. And the gap between them is widening fast.
AI Infrastructure Stocks Just Delivered Blowout Earnings
Four companies. Four different corners of the AI infrastructure stack. One unmistakable signal.
Vertiv: Data Center Demand Is Outpacing Capacity
Vertiv (VRT) is an essential supplier within the AI infrastructure stack. It makes the cooling and power management systems that keep AI data centers humming smoothly. And in the company’s first-quarter earnings call, CEO Giordano Albertazzi put it plainly: the urgency from customers has increased, the scale of deployments is larger, and the technical complexity is greater than it was six months ago.
The numbers back it up. Vertiv posted organic sales growth of 23% in Q1. The Americas — where the hyperscaler buildout is most concentrated — was up 44% organically. Free cash flow more than doubled, up 147% year-over-year to $653 million. At $1.17, earnings per share (EPS) beat guidance by $0.19 and was up 83% from a year ago.
The numbers are strong. The guidance is stronger.
Vertiv raised its full-year EPS guidance to $6.35 — up 51% from 2025 — and its full-year revenue guidance to $13.75 billion, reflecting 34% growth. Its adjusted operating profit is now expected to hit $3.2 billion, up 53%.
Vertiv is running out of room to grow faster. GE Vernova, which supplies the power that makes Vertiv’s cooling systems necessary, has a similar problem.
GE Vernova: The Power Behind the AI Boom
CEO Scott Strazik opened his latest quarterly remarks by noting that the company’s total backlog has grown from $116 billion at its spin-off to $163 billion today. And it now expects to hit $200 billion in backlog by 2027 — a full year ahead of schedule.
In Q1 alone, GE Vernova (GEV) booked $18.3 billion in orders — up 71% year-over-year — at a book-to-bill ratio of approximately 2X. That means that for every dollar of product it shipped, customers ordered two dollars more.
Free cash flow in the quarter was $4.8 billion — more than the company generated in all of 2025. Likewise, the company’s electrification segment’s data center orders reached $2.4 billion in Q1 alone — a full year’s worth of 2025 volume, in a single quarter.
Gas turbine pricing is now running 10 to 20 percentage points higher on new orders than where the backlog was sitting in Q4 of last year. GE Vernova raised its full-year free cash flow guidance to $6.5- to $7.5 billion, up from $5- to $5.5 billion previously.
Its April order intake alone has already exceeded all of Q1.
That’s no ordinary growth rate. It’s the product of a massive — and growing — wave of demand.
Teledyne: Defense and AI Demand Converge
The Iran conflict that’s strangling United Airlines is, for Teledyne Technologies (TDY), a revenue catalyst.
The firm makes sensors, imaging systems, drones, and defense electronics. And on its latest earnings call, executive chairman Robert Mehrabian reported record Q1 sales, record EPS, and record operating margins — the company’s 10th consecutive quarter of a book-to-bill ratio above 1.0.
Defense orders are surging, with the U.S. government actively investing in response to the Iran conflict and broader geopolitical pressures. Counter-drone orders came in the “tens of millions” in Q1 alone. And the company’s Black Hornet nano-drone — the smallest autonomous combat drone on the market — is shipping to customers worldwide. Their total defense-related business is approaching $2 billion in annual revenue, growing at high-single-digit organic rates, and the pipeline is building further.
Geopolitical instability and AI infrastructure demand are, it turns out, not competing forces. For Teledyne, they’re the same order book.
ASM International: Chip Demand Is Maxing Out Supply
And then there’s ASM International (ASM), the Dutch semiconductor equipment maker whose atomic layer deposition (ALD) tools are essential to manufacturing the chips that run AI models.
ASM posted Q1 revenue of €863 million (~$1.01 billion USD) — at the high end of guidance — with a record operating margin of 33.1%. Its Q2 guidance calls for €980 million (~$1.15 billion USD), approaching the €1 billion (~$1.17 billion USD) quarterly run rate.
CEO Hichem M’Saad wasn’t shy: “We are fully booked for this year.”
The company confirmed it will outgrow the broader wafer fabrication equipment market in 2026 — a market that Gartner now estimates is tracking toward 25% growth.
Every segment — logic, memory, power — is growing simultaneously, a rare setup in semis. The reason? Chipmakers are moving to increasingly complex transistor designs that require more of ASM’s tools at every new generation. The current leading-edge chips are already in volume production. The next generation begins pilot production in the second half of this year — and ASM has already locked in key equipment wins there.
The demand runway isn’t measured in quarters. It’s measured in chip generations.
The Old Economy Is Getting Hit by Energy and Cost Pressures
While companies across the AI infrastructure stack were posting blowout results, the past quarter proved to be a slog for firms mired in the “old economy.”
Sonoco, for instance, cut its full-year profit guidance, citing elevated costs it can’t fully pass through to customers. Similarly, United Airlines revised its 2026 outlook downward, grappling with softer demand signals and higher fuel costs largely driven by elevated oil prices.
These disappointing results aren’t the result of strategic errors. These companies are just exposed to the wrong inputs at a time when the Iran War has driven up energy costs. And in a world of razor-thin consumer discretionary margins, higher costs translate directly into lower earnings.
This is the “Hormuz toll” thesis playing out in real time.
Higher oil prices function as a structural tax on energy-intensive, consumer-facing businesses like airlines, packaging, logistics, and shipping.
For hyperscalers and AI infrastructure builders, that tax simply doesn’t apply. Their contracts are priced in compute demand, not jet fuel.
The Bottom Line: AI Infrastructure Stocks Are Pulling Away
The divergence we’re noting between the “old” and “new” economy is not a one-quarter anomaly driven by backlog timing or easy year-over-year comparisons.
AI infrastructure requires years of continued construction; and orders placed today won’t be delivered until 2027 or 2028. The backlog is locked in. The capex commitments from hyperscalers are growing. And the demand from sovereign governments, data center developers, and enterprise customers alike is accelerating.
The old economy, meanwhile, is navigating a world of higher energy costs, softer consumer demand, and the lingering uncertainty of a geopolitical conflict that has permanently altered oil price floors. Anyone whose business model touches physical goods, fuel, or the rate-sensitive consumer is feeling real pressure.
We’re watching two movies on a split screen. And AI has a better plot, a bigger budget, and — for now — all the action.
The structural tailwinds behind AI infrastructure — the power, the chips, the grid, the defense electronics — aren’t waiting for the macro to settle. They’re compounding through it.
Position accordingly.
Today, capital is flowing into infrastructure.
But the biggest gains tend to show up in the next phase, when the platforms built on top of that infrastructure take over.
That shift is already underway.
And one company sits right at the center of it — still private, but moving steadily toward what could be the most important AI IPO of this decade.
I’ve found a way to position ahead of that moment — before it becomes obvious.
