Earnings season just dropped a truth bomb: the economy isn’t just splitting—it’s fracturing into two completely different universes.
On one side, you’ve got AI infrastructure companies absolutely crushing it. On the other, traditional businesses are getting hammered by rising costs they can’t pass along. Same economy, wildly different outcomes. Welcome to the “AI Bifurcation.”
The AI Winners Are Unstoppable
Let’s talk about what’s actually happening. Vertiv, which basically keeps AI data centers from melting down with cooling and power systems, just posted 23% organic sales growth. But here’s the kicker—in the Americas, where all the hyperscaler action is, they grew 44%. Their free cash flow more than doubled. And they’re raising guidance like they’re printing money. Full-year EPS guidance jumped to $6.35, up 51% from last year.
Then there’s GE Vernova, the power company behind the power company. Their backlog went from $116 billion to $163 billion, and they’re expecting to hit $200 billion by 2027—a year ahead of schedule. In Q1 alone, they booked $18.3 billion in orders. That’s not growth; that’s a tsunami.
ASM International, the Dutch chipmaking equipment maker? They’re “fully booked for this year.” Their operating margins hit 33.1%. They’re not just keeping up with demand—they’re running out of capacity to meet it.
Even Teledyne, which makes defense electronics and sensors, is riding two waves at once: geopolitical tensions driving defense orders and AI infrastructure demand. Record sales, record margins, record everything.
The Old Economy Is Getting Crushed
Meanwhile, traditional companies are getting absolutely wrecked. Sonoco cut profit guidance because costs are eating their lunch and they can’t raise prices enough to compensate. United Airlines? Revising guidance downward thanks to higher fuel costs from elevated oil prices.
Here’s the brutal part: these aren’t companies making bad decisions. They’re just exposed to the wrong inputs at the wrong time. Higher oil prices function like a structural tax on energy-intensive, consumer-facing businesses. Airlines, packaging, logistics—they all get hit.
But AI infrastructure companies? Their contracts are priced in compute demand, not jet fuel. They’re immune to the “Hormuz toll.”
The Structural Story
This isn’t a one-quarter blip. AI infrastructure requires years of construction. Orders placed today won’t ship until 2027 or 2028. The backlog is locked in. Hyperscaler capex commitments are growing. Demand from governments, data center developers, and enterprises is accelerating.
The old economy is navigating higher energy costs, softer consumer demand, and permanent geopolitical uncertainty. Anyone whose business touches physical goods, fuel, or rate-sensitive consumers is feeling real pressure.
The Bottom Line
We’re watching two movies on a split screen. AI has the better plot, the bigger budget, and all the action. The structural tailwinds—power, chips, grid infrastructure, defense electronics—aren’t waiting for the macro to settle. They’re compounding right through it.
Capital is flowing into infrastructure today. But the biggest gains? Those show up in the next phase, when the platforms built on top of that infrastructure take over.
Position accordingly.
