Not Chips. Not Memory. Data Center Cooling Is the Unsung AI Infrastructure Play Nobody's Watching

Vertiv reported Q2 revenue growth of 24% year over year to $3.27 billion and adjusted EPS growth of 60% to $1.52, while raising full-year guidance to $14 billion of revenue (+37%) and $6.70 of EPS (+60%). Demand for AI data-center cooling and power infrastructure remains strong, supported by a $15 billion backlog and a 2.9 book-to-bill ratio in Q4 2025. Although the shares have fallen 35% from their May 14 high after rising 50% in 2026, the article argues that 28x forward earnings offers an attractive entry point, with a modeled 2030 price target of $537 implying roughly 120% upside.

Analysis

VRT’s investable edge is not simply AI cooling demand; it is the transition from air-cooled, standardized deployments to liquid-cooling architectures that raise system complexity, attach rates for power distribution/monitoring, and switching costs. The key second-order beneficiary is electrical infrastructure: higher rack density requires integrated power-and-thermal design, favoring VRT against more component-oriented peers such as Modine (MOD) and AAON (AAON). Conversely, hyperscalers’ push toward proprietary rack designs could reduce vendor capture over time if cooling packages become increasingly standardized.

The near-term setup remains sensitive to order quality rather than backlog size. A large backlog can support estimates, but long lead times expose VRT to customer scheduling changes, mix shifts, and execution risk; investors should focus on book-to-bill normalization, backlog conversion, gross-margin progression, and working-capital consumption over the next 1-3 quarters. A weaker-than-expected capex outlook from MSFT, AMZN, GOOGL, META, or Oracle would likely compress the entire AI-infrastructure multiple before it materially affects reported revenue.

The pullback creates a more credible entry point only if earnings expectations stop falling. At roughly 28x forward earnings, VRT is not priced for a conventional industrial cycle; sustaining the multiple requires evidence that liquid cooling expands margins and that AI revenue is durable beyond a concentrated set of hyperscale projects. The contrarian risk is that cooling is treated as a pure bottleneck play while data-center power availability—not thermal equipment—is the actual gating constraint, delaying shipments and revenue recognition despite robust demand.

Over 6-18 months, VRT can outperform semiconductor AI beneficiaries if physical deployment becomes the binding constraint and customers prioritize time-to-commissioning over equipment price. However, the bull case’s long-duration earnings assumptions should be discounted: competitive capacity additions from Schneider Electric, Eaton (ETN), Johnson Controls (JCI), and private liquid-cooling specialists could cap incremental margins as the market matures.

Key Decisions for Investors

• Initiate a staged long VRT position over the next 2-6 weeks, beginning at one-third size and adding only after the next earnings report confirms backlog conversion and stable-to-higher gross margin. Target 20-30% upside over 6-12 months if estimates continue rising; exit/reassess on a material guidance cut or book-to-bill below 1.2x.
• Prefer a relative-value pair: long VRT / short JCI over a 3-6 month horizon. VRT has greater exposure to high-density AI deployments, while JCI retains more conventional commercial-building HVAC exposure; size the short to neutralize broad industrial and rate sensitivity.
• Use VRT call spreads rather than outright long-dated calls if implied volatility remains elevated: buy 6-9 month at-the-money calls and sell strikes 20-25% higher. The structure expresses continued estimate revisions while limiting premium paid for an already well-publicized AI narrative.
• Monitor hyperscaler capex guidance and data-center power interconnection disclosures as leading indicators. Defer additional VRT exposure if major customers signal power-delivery delays, since that would shift revenue rather than eliminate demand and can still drive a 1-3 quarter valuation reset.
• Avoid chasing MOD or AAON solely as read-through trades until liquid-cooling revenue mix, capacity plans, and valuation are verified; their exposure to AI thermal demand is less direct and may not offset cyclicality in legacy HVAC end markets.