Back to News
Market Impact: 0.35

Big Tech Is on Pace to Spend $735 Billion on AI Data Centers in 2026. These 3 Industrial Stocks Collect the Checks.

+2
Artificial IntelligenceCompany FundamentalsCapital Returns (Dividends / Buybacks)Corporate EarningsCorporate Guidance & OutlookEnergy Markets & PricesCredit & Bond Markets
Big Tech Is on Pace to Spend $735 Billion on AI Data Centers in 2026. These 3 Industrial Stocks Collect the Checks.

AI data center build-outs are driving strong results in industrials: Caterpillar’s Power & Energy revenue rose 17% YoY to $8.2B in Q2, while Eaton posted record Q2 revenue of $8.5B (+21% YoY) with data-center revenue up 65% and electrical backlog up 43%. GE Vernova reported 22% revenue growth in Q2 and said data center orders exceeded $5B YTD (more than double 2025), leading it to raise 2026 guidance for total revenue to $45.5B–$46.5B and free cash flow to $11.5B–$12.5B. Overall, the article frames industrial power-generation, grid, and equipment as key beneficiaries of ongoing Big Tech AI capex.

Analysis

The market is still underpricing the fact that AI capex is a transmission story, not just a compute story. The immediate beneficiaries are the firms selling bottlenecked, high-spec infrastructure with long lead times and pricing power: electrical gear, switchgear, backup generation, gas turbines, and service. That means the earnings translation is likely cleaner than semis in the next 2-4 quarters because backlog converts into revenue with less demand volatility and less direct exposure to GPU supply cycles.

The second-order effect is a relative value rotation within industrials. Names with data-center exposure plus long-cycle backlog should keep gaining share at the expense of more generic heavy equipment franchises, which have lower direct AI sensitivity and more cyclical end markets. The real supply-chain choke points are transformers, turbine components, and grid interconnect equipment; if those remain tight, margins can stay elevated even if unit growth slows, but they also create execution risk if shipment schedules slip.

The contrarian view is that consensus may be chasing the right theme but the wrong duration: the first leg of upside may already be in the stocks, while the larger move comes from sustained backlog conversion over 6-18 months. What could break the thesis is not AI spend slowing abruptly, but hyperscalers deferring build schedules due to power constraints, permitting delays, or a capex digestion phase after the current surge. That would hit revenue timing before it hits order books, which is why the market may pay up too early for headline growth.

More News