Want to Make a Fortune From the AI Power Boom? Buy These 2 Industrial Stocks

Goldman Sachs estimates U.S. data-center power demand could approach 108GW by 2030, underpinning growth opportunities for utility American Electric Power and data-center infrastructure supplier Vertiv. AEP has 69GW of contracted load additions through 2030, a $78B five-year capital plan, and raised 2026 operating EPS guidance to $6.25-$6.55 while targeting more than 9% annual operating-earnings growth through 2030. Vertiv guided to 2026 sales of $14B, representing roughly 31% organic growth, supported by a $15B backlog, though its approximately 43x forward P/E leaves the shares exposed if data-center spending or order timing weakens.

Analysis

The investable distinction is regulated-load monetization versus equipment-cycle monetization. AEP's incremental data-center load should expand its rate base, but equity upside depends on timely regulatory recovery, allowed ROE, and whether large-load contracts shift interconnection and stranded-asset risk to customers. The near-term risk is that accelerating capex outstrips internally generated cash flow, raising financing needs and making the stock more rate-sensitive than the AI-demand narrative implies; the more durable beneficiaries may be transmission-component suppliers such as Eaton (ETN) and Hubbell (HUBB), which avoid a utility's regulatory lag.

VRT has the superior operating leverage to higher rack density because liquid cooling and power-distribution content rises faster than facility megawatts. That also makes it the most exposed name if hyperscalers move from broad capacity reservation to phased deployment: backlog is not equivalent to revenue until projects clear power availability, construction, and customer acceptance. At a premium earnings multiple, even intact long-run demand can produce a sharp 1-3 month de-rating if bookings, book-to-bill, or margin guidance merely decelerates.

Consensus is treating power demand as a single bottleneck, but grid connection timing may become the binding constraint. Delayed energization shifts demand from near-term electrical/cooling equipment toward later delivery dates, while increasing the relative value of on-site generation, storage, and power-management vendors. Over 6-18 months, the best risk-adjusted expression may therefore be suppliers with data-center content growth but diversified end markets, rather than a concentrated bet on either utility rate-base execution or VRT's elevated growth expectations.

Key Decisions for Investors

• Maintain a 6-12 month long VRT / short ETN relative-value position only after confirming quarterly orders and book-to-bill remain positive; VRT offers greater AI-density upside, while ETN hedges a broad data-center electrical-infrastructure slowdown. Exit if VRT guides to a material sequential orders decline or reduces margin expectations; target 15-20% relative outperformance, with a 8-10% relative stop.
• Accumulate AEP on rate-driven weakness rather than chase AI-related strength, with a 12-24 month horizon. The underwriting trigger is evidence that large-load agreements include minimum-demand, contribution-in-aid-of-construction, or cost-recovery protections; avoid adding if financing guidance implies persistent equity issuance that dilutes per-share rate-base growth.
• Use a basket long ETN, HUBB, and MOD versus a modest short VRT as a 1-3 month catalyst hedge if hyperscaler capex or power-connection schedules soften. These suppliers retain broader utility, industrial, and retrofit demand, whereas VRT's valuation is more vulnerable to a growth-duration reset; cover the short if VRT reports accelerating orders or raises full-year sales guidance.
• Set an event alert around AEP's relevant state commission and ERCOT interconnection decisions. A favorable cost-allocation framework supports a rerating; delays, customer withdrawals, or a lower allowed ROE would falsify the thesis even if aggregate data-center power demand remains strong.