Top US Workforce Services Stocks for the AI Infrastructure Boom: Barclays

The underappreciated beneficiaries of AI capex are not the chip names already in every model, but the constraint providers that translate capex into physical deployment. Workforce housing and camp operators have a cleaner near-term revenue path than most infrastructure adjacencies because every data center build needs labor, beds, food, logistics, and site support before the compute stack ever goes live. That makes CVEO the better quality way to express the thesis: existing room inventory can reprice quickly, and incremental occupancy should flow through with high operating leverage if data center construction ramps as expected.

TH is more of a trading vehicle than a compounder right now. The stock has a self-inflicted overhang from the secondary and a recent earnings miss, so any AI-driven upside likely gets partially offset by skepticism around whether the contract wins are repeatable and how much of the announced revenue is already reflected in guidance. If hyperscaler spend is real, the first second-order effect is probably not higher multiples for every “pick and shovel” name, but tighter labor accommodation capacity and better pricing power for the few operators with scalable campuses and room inventory.

The key risk is that the $1T capex narrative stays a 2028 story while investors try to front-run 2026-27 bookings today. If AI training spend decelerates faster than expected, or sovereign/China capex disappoints, these names can de-rate hard because their valuation support depends more on duration than on current earnings quality. The contrarian point is that the market may still be underestimating how much of the spend leaks into non-sexy support services; the better trade may be to own the scarce physical bottlenecks, not the obvious semiconductor beta.