2 Industrial Stocks I'd Buy Every Day Before RTX

Vertiv (VRT) reports Q2 revenue of $3.27B (+24% YoY) and adjusted EPS of $1.52 (+60% YoY), then raised full-year guidance to $14B revenue (+37% YoY) and adjusted EPS guidance to $6.70 (midpoint, +60%), while the stock is up 122% over 52 weeks. Quanta (PWR) saw Q2 revenue rise 41% to $9.6B, adjusted EPS +71% to $4.24, and posted a record $53.4B backlog with management increasing 2026 expectations across metrics. The article argues RTX (RTX) is stronger on valuation already (39x P/E) but sees Vertiv and Quanta as benefiting more directly from an AI-driven multiyear power/cooling and grid upgrade cycle.

Analysis

The real tradeable signal is not “AI is good” but that the bottleneck is shifting from compute to power-delivery and thermal infrastructure. That usually favors the names closest to the constraint set — grid interconnect, transformers, cooling, switchgear, and field labor — because they can reprice faster than hyperscaler cloud budgets can be reallocated. The second-order effect is that capital is likely to keep rotating up the AI supply chain, while mature industrial/defense compounders like RTX become lower-beta sources of funds rather than incremental destination capital.

The risk is that both VRT and PWR are now duration trades on sustained AI capex, not just good quarterly prints. If hyperscalers defer buildouts, if utility interconnection queues worsen, or if labor/material inflation squeezes project margins, these stocks can de-rate quickly even while revenues remain strong. The first place to look for reversal is management commentary on backlog conversion, order intake, and 2026 margin cadence over the next 1-2 quarters.

Consensus may be underpricing how much execution is already embedded in the multiples. For PWR especially, backlog is only valuable if permitting, labor, and equipment availability allow conversion on schedule; otherwise backlog becomes a timing asset, not an earnings asset. RTX is not a short on fundamentals, but relative to the AI infrastructure names it looks like a crowded, lower-upside place to hide inside industrials; the more interesting expression is to own the secular bottleneck, not the already-recognized beneficiary.

Key Decisions for Investors

• Initiate a 3-6 month long VRT position on any 5-8% pullback; use a 10-15% drawdown stop and watch for any slowdown in data-center order growth or FY guidance revision as the falsifier.
• Put on a relative-value pair: long PWR / short RTX for the next 1-2 earnings cycles. Thesis is that AI grid buildout has a longer runway than defense/aerospace rerating, while RTX has less multiple expansion left; target 8-12% spread capture if the AI infrastructure theme stays intact.
• If already long RTX from defense exposure, trim into strength or sell covered calls against the position rather than adding here; the risk/reward is compressed near highs unless there is a new budget/geopolitical catalyst.
• Watch hyperscaler capex updates and utility interconnection data over the next 1-3 months; a pause there would be an early warning that VRT/PWR multiples can compress before reported revenue inflects.