Somebody on your board has read about Caterpillar's second quarter. It was the first $20 billion quarter in the company's history, sales were up 24%, and the ISM manufacturing survey has now shown expansion eight months running. So at the next meeting they ask what sounds like a fair question. If manufacturing is doing this well, why are we flat?
It is a fair question, sort of. For a lot of mid market industrial companies it's also the wrong comparison, and I'd guess most of the people asking it haven't looked at where Caterpillar's growth actually came from.
What the briefings kept turning up
Since mid-August we've been running market briefings on industrial and manufacturing companies. Steel, piping, specialty chemicals, aerospace materials, construction services, fertilizer, energy storage. On paper they don't have much in common.
Data centers or defense spending kept coming up as a real driver of demand somewhere in their markets, a lot more often than I expected going in. For most of those companies it was their own tailwind, which is great and comes with a catch I'll get to at the end. For a couple it was a much bigger neighbor's tailwind, and that neighbor's results were the ones everybody kept quoting. And for the ones in places like fertilizer, water treatment, irrigation and industrial chemicals, neither one came up at all.
None of that is a scientific sample, and I'd be a little wary of anyone who built a macro thesis on it. The public numbers point the same direction, though, and the easiest place to see it is inside the big companies' own earnings releases.
Split the beat before you compare
Valmont is the one I keep coming back to. Second quarter revenue up 6.5% to $1.12 billion, full year guidance raised. If you compete with Valmont in irrigation, that headline reads a lot like getting lapped.
Then you get to the segment table. Valmont's agriculture business, the part that actually competes with you, fell 15.8%, and international ag was down almost 29%. The beat came from infrastructure, where North American utility sales grew 34%. So yes, Valmont had a great quarter. It just didn't happen anywhere near irrigation, where Valmont may well have had a rougher few months than the smaller companies it competes with.
Deere breaks apart the same way. Construction and forestry sales rose 18% on data center and large infrastructure work, and the earthmoving order books are largely full for the rest of the year. Production and precision ag, the big green machines most people picture when they hear the name, fell 6%. And that Caterpillar quarter from the board meeting? Power generation retail sales were up 72%, and the CEO didn't dance around why. Generator sets and turbines for data centers.
The macro data has the same problem
July construction spending fell half a percent, to its lowest level since October 2023. Nonresidential spending technically went up, but according to ABC every bit of that increase came from data centers, and with data centers taken out, nonresidential was at its lowest since September 2023. AGC's chief economist, Ken Simonson, put it more bluntly: "Only three categories are propping up construction spending: data centers, power and highway projects."
Factory output shows a version of it too. The Fed reported in mid-September that manufacturing production fell 0.3% in August, the first drop after seven straight monthly gains, even with the ISM survey still in expansion territory. Both can be true at once. An index can climb for months because a few sectors are sprinting while plenty of others walk in place, and the version with the sprinters pulled out almost never makes the headline.
Somebody has to define reality in that room
Max De Pree wrote that the first responsibility of a leader is to define reality. I know, it sounds like something that belongs on an office poster next to a picture of a rowing crew. But when someone at the board table is holding a peer's earnings release, it turns into a pretty literal job description, and more often than not the job lands on the CFO.
So before the next meeting where a big peer's quarter might come up, do the split yourself. Pull the segment table from their release, find the line that actually overlaps with your business, and bring that number instead of the headline. The conversation moves from why you aren't growing like them to how your segment is doing across the industry and where you sit inside it. That's a much better conversation to be in, even in a quarter where your number isn't great.
Then do the same with the macro data. If your customers are farmers, net farm income will tell you more than the PMI ever will. If you sell into commercial construction, the series you want is nonresidential spending with data centers stripped out, which ABC is conveniently calculating for everyone at the moment. Pick the one or two series that really describe your customers and start putting them in the board pack every month, so the board already knows how to read them by the time it matters.
And if you're the one riding it
If you make something that ends up in a data center or a defense program, this year probably looks pretty good, and I'm not going to talk you out of enjoying it. It's still worth checking how much of your growth now depends on one kind of customer, because concentration tends to build quietly while things are going well. Hyperscaler capex gets set in a handful of boardrooms. Defense spending moves on whatever schedule Congress is in the mood for. I don't know anyone who could tell you with real confidence what either looks like in 2028, including, probably, the people doing the deciding.
CFOmarketIQ does this kind of read one company at a time, against your own segments and the competitors you actually run into. If you'd like to see where your company sits, the first briefing is free.