2 AI Data Center Infra Giants Looking More Promising After Q1 Results

AI infrastructure demand remains exceptionally strong, with McKinsey estimating global AI data center capex could reach about $7 trillion by 2030 and hyperscaler spending rising to $650 billion in 2026, up 71.1% year over year. Vertiv raised 2026 guidance to $13.5 billion-$14.0 billion in net sales and $6.30-$6.40 in adjusted EPS, while Comfort Systems reported backlog of $12.45 billion, up 80.8% from a year ago. The article is broadly bullish for AI infrastructure and specialized HVAC suppliers, though it is largely a sector commentary piece rather than a single company catalyst.

Analysis

The setup favors the infrastructure layer over the model layer for the next 6-18 months: when hyperscalers accelerate capex, the first monetization usually accrues to power, cooling, and electrical bottlenecks before software names re-rate. VRT is the cleaner expression because it sits directly on the constraint that matters most as rack densities rise—power distribution and thermal management—so its order book should have better visibility than most AI beneficiaries. FIX is more of a second-derivative winner: not pure-play data center exposure, but a backlog compounding story that can surprise if AI-related HVAC demand keeps forcing premium mix and faster price realization.

The more interesting second-order effect is that this spend cycle likely tightens capacity across the supply chain, not just demand. Lead times for switchgear, transformers, liquid cooling, and skilled install labor should stay extended, which supports pricing power and reduces near-term downside unless project delays become widespread. That also means competitors with less scale or weaker procurement can get squeezed even if end-demand stays strong; the risk is not demand collapse but execution bottlenecks, which can push revenue recognition rightward while still improving backlog quality.

The market may be underestimating how much of the upside is already in the stocks’ absolute moves, but not necessarily in earnings revisions. The key contrarian issue is that consensus could be extrapolating the current growth rate too far into 2027-2028, while the real surge in AI infrastructure spending likely remains concentrated in a narrower window. If hyperscaler budget discipline or GPU deployment timing slips, these names can de-rate quickly because expectations are now anchored to sustained hypergrowth rather than cyclic construction activity.

Key Decisions for Investors

• Long VRT on 3-6 month horizon; prefer pullbacks into post-earnings consolidation. Risk/reward remains attractive while estimates are still being revised upward, but trim if the stock starts pricing 2027 growth too early.
• Initiate a pair trade: long VRT / short a basket of lower-quality electrical or industrial contractors with weaker AI exposure. Thesis is that the AI bottleneck accrues to the most vertically integrated power/thermal platforms, not generic project installers.
• Buy FIX only as a relative-value laggard, not a core conviction long. Timeframe 3-9 months; best entry is on backlog-driven dips if same-store backlog continues to compound faster than revenue, signaling pricing power and future margin support.
• Consider a call spread in VRT for 6-9 months rather than outright equity. The name is already extended, so defined-risk upside exposure is preferable if hyperscaler capex headlines remain supportive but volatility stays elevated.
• Fade NVDA relative to the infra suppliers tactically over 1-3 months if the market continues to reward the picks-and-shovels trade. The article’s real incremental signal is not GPU demand, but the infrastructure spend required to support it.