AI Data Center Boom Lifts Industrial Stocks Including Coatings and Asphalt Suppliers
N.R. Finch
Five hyperscalers are set to spend over $750 billion on AI this year, and the money is spilling past chips and servers — paint, cable and asphalt companies are flagging data-center demand in earnings calls, sending their shares sharply higher.
Which companies are cashing in on this unexpected windfall?
Sherwin-Williams jumped 8.3% in a single session — its biggest one-day gain in over four years. CEO Heidi Petz said AI data-center construction pushed its protective-coatings sales growth to the "mid-teens", far above the company's overall 7.5% quarterly pace.
Cable distributor WESCO International beat on sales and earnings; its data-center solutions unit posted a 45% year-on-year revenue jump. This means → data centers are not just buying chips — even cable procurement is growing at close to double.
3M disclosed that Microsoft is using its patented fiber-optic connectors in Azure data centers, triggering its largest one-day gain of the year. Asphalt and aggregates firms Vulcan Materials and Martin Marietta also flagged data-center demand in this week's results.
Why are paint and asphalt companies riding the AI wave?
Building a data center is, at its core, a large-scale construction project: foundations, structural steel, fireproofing coatings, insulation, dielectric coatings — a specialty layer that prevents signal interference between circuits — and massive amounts of cable. It is not fundamentally different from building a factory.
In plain terms = chips are the data center's brain, but the brain has to sit inside a building — and the money to build that building ends up in paint cans and asphalt trucks.
Sherwin-Williams' CEO stressed that "speed matters" — hyperscalers racing to bring capacity online are willing to pay a premium for one-stop coating solutions. This reflects a shift in bargaining power toward industrial suppliers that can deliver fast.
How big is the $750 billion capex pool?
Amazon, Alphabet, Meta, Microsoft and Oracle are expected to spend a combined $750 billion-plus on AI this year. This means → even if paint and asphalt capture only a sliver of that total, the absolute dollar amount is large enough to move the growth needle for traditional industrial firms.
Information-processing equipment and software investment contributed nearly 0.5 percentage points to the U.S. economy's 1.5% second-quarter GDP growth — still well above pre-AI historical norms.
But GDP accounting only covers the most expensive components — servers and chips. Paint, asphalt and cable purchases are not fully captured. This means → AI construction's real pull on the physical economy is likely understated.
What risks come with being an "accidental winner"?
These industrial firms' earnings are now tightly linked to the AI capex cycle: they rise when hyperscalers spend and shrink when spending slows.
In plain terms = paint companies used to ride real-estate and infrastructure cycles. Now they have an AI cycle too — one more growth engine, but also one more source of volatility.
This reflects an AI supply chain whose radius of influence is stretching far beyond the market's original expectations: from chips → servers → cloud platforms, all the way to paint cans and asphalt trucks.
Content is for reference only, not financial advice.