Hey Income traders,
Somewhere between Dow Chemical (DOW) and Union Pacific (UNP), Wall Street forgot that the future still runs on rails, copper wire, and cement.
While software stocks are getting crushed by the very AI tools they helped create, the “stuff” companies, the ones that move things, build things, and power things, are quietly having the best year in a decade. Industrials are up 28 percent over the past year. Utilities broke out to all-time highs. Meanwhile, IGV, the software sector ETF, just posted its worst month since October 2008.
Stuff is the theme. And if you’re still overweight software hoping for a bounce, you’re playing the wrong game.
The Software Apocalypse Nobody Wants to Talk About
Here’s a rule I live by: if you haven’t sat down with the latest AI tools and actually used them, don’t comment on software stocks. Not because I’m gatekeeping, but because once you do, you’ll understand why Adobe (ADBE) is untouchable and why Microsoft (MSFT) has more questions than answers for the first time in a decade.
I won’t touch Adobe with a twelve-foot pole. I sold my Microsoft stock and added downside puts. That’s not a prediction. That’s a reaction to what I’m seeing with my own eyes.
The software selloff isn’t a blip. It’s a reckoning.
In January 2026, the software sector suffered its worst three-week stretch in nearly two decades, down 15 percent in a straight line. Adobe’s market cap collapsed from $350 billion to $107 billion. ServiceNow (NOW), one of the supposed “safe” enterprise names, trades at a forward PE of 28. Its five-year average? 67.
When there’s a big selloff in a theme, everything correlates to one. They just crush everything. Then you start seeing the bounce, and it reveals which businesses are durable and which aren’t. Datadog (DDOG) ripped 15 percent on the bounce day. Salesforce (CRM)? Flat. Adobe? Dead.
The market is doing triage in real time.
Why “Boring” Is Beating “Brilliant”
While software burns, something strange is happening in the sectors most investors forgot existed.
XLU, the Utilities Select Sector SPDR, gained 20.7 percent in 2025 and is up another 15.5 percent year-to-date in 2026. XLI, the Industrials ETF, returned 28 percent over the past twelve months. The February 2026 sector rotation data tells the whole story: Industrials, Energy, and Materials are leading. Technology and Communication Services are lagging.
This isn’t random. This is money figuring out where the bottlenecks actually are.
AI needs compute. Compute needs power. Power needs infrastructure. Infrastructure needs materials. Materials need mining. Mining needs equipment. Equipment needs shipping.
Every single one of those links is a physical business that AI can’t disrupt. In fact, AI is making them more valuable, not less.
The hyperscalers, Amazon, Microsoft, Google, Meta, Oracle, are spending $600 billion on data center infrastructure in 2026 alone. That’s a 36 percent increase from last year. Seventy-five percent of that spend goes directly into AI infrastructure. The average data center now costs $597 million to build, up from $374 million just twelve months ago.
Where does all that money go? Steel. Concrete. Glass. Copper. Power equipment. Construction labor.
The Dart-Board Thesis That Actually Works
I’ll admit something: in infrastructure and materials, it really hasn’t been that hard. You could almost throw a dart at the newspaper.
I shouldn’t say that. But look at anything. Deere (DE). Union Pacific. Caterpillar (CAT). Dow Chemical. Vulcan Materials (VMC), we’re four for four in special situations with that one.
What are they gonna need to do in the industrial revitalization of the world? They’re gonna ship things around by train. They’re gonna need cement. They’re gonna need copper. They’re gonna need glass.
Corning (GLW) is a perfect example. I bought GLW in the low four hundreds and watched it run to $600. The thesis was simple: photonics. You cannot move data fast enough with electricity to meet AI compute demand. Light is the only option. And Corning makes the glass. They don’t care who wins the photonics race. They supply everyone.
The value isn’t in 22-times-earnings S&P names. It’s in the 14-PE infrastructure stocks that haven’t gone anywhere yet. Everything old is new again.
When Tech Money Rotates, Small Sectors Explode
Here’s the paradox nobody’s talking about: there’s so much money in tech that when it rotates out, it pushes these smaller sectors parabolic.
Think about how small energy is now compared to tech. Information Technology makes up 35 percent of the S&P 500 weighting. Industrials? About nine percent. Materials? Barely two percent.
When institutional money decides to hide from software disruption risk, it doesn’t have many places to go. These “boring” sectors become the only game in town. And because they’re so much smaller, relatively modest inflows cause massive price moves.
GE Vernova (GEV) went vertical. Quanta Services (PWR) ripped. The broadening play I’ve been talking about all year isn’t slowing down. It’s accelerating.
Your Move Before the Next Leg
If you’re still sitting in software waiting for the bounce, ask yourself: why?
Microsoft hit the nine-day EMA on a rally yesterday. That’s usually a powerful signal for bearish positioning. Their earnings weren’t good enough. Copilot is underperforming. Azure was “okay.” Cloud was “okay.” I’m telling you, the stuff I’m seeing out there, maybe you won’t need an operating system at some point.
That’s not a company I want to own at 28 times earnings when the narrative shifts against them.
The trade is straightforward: rotate out of software with multiple questions and into infrastructure with obvious demand. Power. Rails. Materials. Equipment. The names trading at 14 times earnings with visible catalysts instead of 22 times earnings with existential threats.
Cement is sexy now. Get used to it.
Stuff is the theme. And until AI figures out how to pour concrete, that’s not changing.
May the income be with you,
Hans