Oracle's Data Center Spending Is Exploding. Here's the Power Stock That Actually Benefits.

Bloom Energy reported Q2 record revenue of $1.065B (+166% YoY) and turned to profitability with EPS of $0.62 vs a year-ago loss of $0.18. The company raised full-year revenue guidance to $3.9B–$4.2B (+~100% vs 2025) and expanded gross margins by 668 bps to 33.4%, while generating strong operating cash flow to fund manufacturing expansion. A major demand tailwind comes from a large Oracle deal scaling up to 2.8GW of on-site fuel cell power, though the stock’s valuation is cited as a key risk (up 146% YTD; ~290x trailing and ~78x forward earnings).

Analysis

Bloom’s real edge is not “fuel cells” per se; it is monetizing the time-to-power bottleneck that AI data centers cannot afford. That makes the company a second-order beneficiary of hyperscaler capex, but also a quasi-substitute for utility buildouts, which means the trade is really about how long grid interconnect delays persist. In the near term, that supports BE and, more modestly, EQIX as uptime and tenant activation remain scarce assets.

The market is likely underpricing how concentrated this growth is. A handful of very large deployments can make the revenue curve look software-like for a quarter or two, but industrial economics eventually reassert themselves: gas input costs, manufacturing yield, and field-install execution determine whether gross margin stays in the low-30s or reverts. If Henry Hub moves up or Oracle/Brookfield cadence slows, BE’s multiple can compress fast because the stock is pricing both speed and certainty.

Contrarian view: consensus is treating this as a durable secular winner when part of the upside may simply be a temporary backlog flush caused by grid scarcity. That means the thesis is strongest over the next 1-3 quarters, weaker over 6-18 months unless management proves repeatable deployment and conversion rates. The biggest structural losers are regulated utilities and slower behind-the-meter alternatives if BE keeps winning on installation speed; the biggest hidden risk is that batteries, gas turbines, or faster permitting narrow the urgency premium before BE scales into its valuation.

Key Decisions for Investors

• Long BE only on pullbacks, not strength: use a 10-15% retracement after the next earnings/contract update to establish a position, because the current multiple already discounts near-flawless execution.
• Pair trade: long BE / short XLU over the next 1-3 months to express the thesis that behind-the-meter power displaces utility load growth; cover the short if utilities show accelerating load-interconnect wins or if rates fall sharply.
• Use BE call spreads instead of stock for upside exposure into the next Oracle/Brookfield milestone cycle; the payoff is best if new GW bookings continue, but the premium protects against a de-rating on any execution slip.
• Watch-item, not a trade yet: if forward revenue growth drops below ~70% y/y or gross margin falls back under the low-30s in the next print, the growth-compounder narrative breaks and the stock likely de-rates 20%+.