The On-Site Power Duel: Bloom Energy vs. GE Vernova

Bloom Energy’s Q3 revenue rose 138.1% year over year to $1.24B and product gross margin reached 38.4%, but it used $181M of free cash flow; management said year-end deliveries will determine revenue placement within its $4.40B-$4.60B guidance range. GE Vernova booked $29.3B of orders, up 96% organically, reported $5.41B of free cash flow and 21.4% Power EBITDA margin, and agreed to acquire Trench Group for $9B, with closing expected in late 2027. Over the past month Bloom shares fell 40.53% and GE Vernova shares fell 22.32%; the article highlights Bloom’s cash-conversion and delivery risks versus GE Vernova’s larger, cash-generative backlog, while noting expected 2026 Wind EBITDA losses of $400M-$500M.

Analysis

Bloom’s key risk is not whether customers value fast power, but whether installations convert into cash before delivery schedules slip. A few delayed projects can hit reported revenue while contract assets and working capital absorb cash; rising demand alone does not resolve that execution risk. On-site systems also remain exposed to gas availability, permitting, and customer-site infrastructure, so they route around parts of the grid rather than eliminate energy constraints.

GE Vernova offers a different risk: its backlog can support multi-year visibility, but bookings are not equivalent to realized margins or cash. Watch order quality, cancellations, and conversion into deliveries; the Trench acquisition adds integration and closing risk before its expected late-2027 completion. Utility-scale generation and grid equipment may complement on-site power, making a winner-take-all framing misleading. A data-center capex pause would pressure both, but Bloom’s project-level timing and cash conversion make it more vulnerable near term. Conversely, if utilities adopt Bloom systems broadly, that could expand its addressable market while also creating a channel that competes with direct customer sales.

The recent relative drawdown may already price in near-term disappointment for Bloom, so avoid treating weakness alone as a fresh short signal. Over 1–3 months, focus on Bloom’s delivery timing and operating cash flow, and GEV’s backlog quality and margin progression. Over 6–18 months, the structural question is whether grid investment catches up enough to reduce customers’ urgency for behind-the-meter capacity. The thesis weakens if Bloom demonstrates repeatable cash conversion and utility orders, or if GEV’s backlog fails to translate into profitable execution.

Key Decisions for Investors

• Prefer GEV over BE as a conditional relative-value position, not a broad power-demand bet. Consider entry only after reviewing Bloom’s year-end delivery and cash-conversion update; size for GEV’s wind losses and acquisition execution risk.
• Do not add to a BE short solely on the reported selloff. Reassess if deliveries slip, operating cash flow remains weak, or contract assets fail to convert; a signed utility order plus positive cash generation would be a thesis-falsifying signal.
• Track GEV order cancellations, backlog conversion, and power-margin guidance over the next 1–3 months. Deterioration in any of these would challenge the assumption that headline backlog provides durable earnings visibility.
• Monitor data-center project delays and utility interconnection timelines across the next 6–18 months. Faster grid buildout could reduce the urgency premium for on-site systems; persistent delays could support both models, with near-term execution still favoring the better-funded supplier.