Bloom Energy: hyperscalers turned a fuel-cell also-ran into the default behind-the-meter power vendor

Stevie AI on Bloom Energy Corporation Class A (BE-USA | bloomenergyc)

10/27/2026

Summary

Bloom Energy makes solid-oxide fuel cell Energy Servers that let data centers and industrial customers generate power on-site, bypassing grid interconnection queues that now stretch years in many US markets. The structural insight is simple: hyperscalers need gigawatts of power faster than utilities can deliver it, and Bloom's modular systems ship in 2-3 months versus multi-year timelines for gas turbines or grid upgrades. That has turned a company long dismissed as a subscale, cash-burning fuel cell story into the de facto standard for rapid on-site power, with guidance raised three consecutive times in 2026 alone. The historical numbers show the inflection starting late: FY2024 revenue was $1.5B with essentially break-even net income and -$0.13 EPS; FY2025 revenue grew to $2.0B but net income stayed negative at -$0.1B and EPS worsened to -$0.37 as the company absorbed costs ahead of the hyperscale ramp. FY2026 is where the model flips, with revenue guided to $4.4-4.6B (+122% YoY), non-GAAP gross margin raised to ~34%, and non-GAAP EPS guided to $3.10-3.40 - a swing from persistent losses to a $0.4B net income, $1.57 GAAP EPS year. Free cash flow, however, stays negative in 2026 (-$0.1B) despite GAAP profitability, reflecting working-capital intensity (receivables, inventory, contract assets) that management frames as self-funding but which has produced negative operating and free cash flow in recent quarters. Applying a 35x multiple to FY2027 EPS of $3.48 - chosen to reflect a premium growth-industrial name decelerating from triple-digit to 35-40% growth with expanding margins, but discounted versus pure-software hyperscale plays given capital intensity and working-capital drag - yields a FY2027 price target of roughly $122, discounted back to approximately $111 for FY2026. Rolling forward, FY2028 EPS of $5.17 and FY2029 EPS of $6.90 at the same multiple imply targets of roughly $181 and $241 respectively. Against the current $203.92 price, the stock is already discounting much of the 2027-2028 earnings power, leaving the risk/reward balanced rather than compelling at these levels - hence a HOLD rather than BUY, with the FY2029 target offering upside only if execution and margin trajectory hold without disruption.

Thesis

1. Category leadership in a newly created market. Bloom is not competing on levelized cost of electricity against gensets or turbines; it is selling speed and site-level certainty in a market where grid interconnection queues are the binding constraint for hyperscale data center buildout. The strategic assessment of a 7.1/10 blue-ocean score reflects that Bloom has effectively created a new purchasing category - non-combustion, modular, 2-3 month deployment on-site power - where it faces limited direct competition from Plug Power or FuelCell Energy at scale, and where traditional genset makers like Caterpillar, Generac and Cummins do not match the permitting and community-acceptance profile. 2. Financial inflection is real and already visible in guidance. The jump from -$0.37 EPS in FY2025 to a guided $3.10-3.40 non-GAAP EPS in FY2026 is not a model assumption - it is management's own raised guidance, now issued three times upward in a single year. Revenue guidance of $4.4-4.6B represents 122% growth at the midpoint, and non-GAAP operating margin guidance has moved from 14% at the February outlook to 22% at the midpoint of current guidance, which is a genuine structural re-rating of the business's unit economics, not merely a volume story. 3. Margin expansion is structural, not just mix. Non-GAAP gross margin guidance has moved from 30% in FY2025 to approximately 34% in FY2026, and product gross margin in Q3 reached 38.4% non-GAAP, up 250bps YoY. This is consistent with a manufacturing business climbing the utilization curve at Fremont and Newark as ramp projects mature - the kind of margin trajectory that, if sustained through FY2029's modeled low-30s% GAAP margin, supports the EPS path from $1.57 in FY2026 to $6.90 in FY2029. 4. The market is mispricing the shape of the growth curve, not its existence. Consensus narratives on Bloom have oscillated between 'fuel cell cash incinerator' and 'AI power play,' but the real debate now is about the deceleration profile: growth moderates from 122% in 2026 to 35-40% in 2027 and then to 20-25% by 2029 as the hyperscaler buildout matures. At $203.92, the stock already prices a meaningful chunk of this deceleration-adjusted earnings power, which is why upside from here requires the multiple to hold rather than compress as growth normalizes. 5. Capital discipline is being tested in real time. Capex of $231 million year-to-date - roughly 4x all of 2025's spending - is being funded from operations per management's own framing, and backlog is reported to be growing faster than revenue, though without specific disclosed figures. This self-funding capex model, if it holds, differentiates Bloom from prior-generation fuel cell companies that relied on dilutive capital raises, and is a key reason the market has re-rated the stock over the past year. 6. Catalysts are concrete and dated. Specific customer handoffs (Wyoming, New Jersey) will serve as near-term proof points for the $4.4-4.6B FY2026 revenue range, and management has committed to providing FY2027 guidance and backlog disclosure in upcoming reports. Positive delivery of these catalysts, especially quantified backlog, could support further multiple expansion; failure to hit them given the guided range's sensitivity to 'a small number of end-of-year deliveries' is a binary near-term risk.

Risks

1. Working capital intensity threatens the self-funding narrative. Despite guided non-GAAP profitability, Bloom posted negative Q3 operating cash flow (-$27M) and negative free cash flow (-$181M), driven by contract asset growth, inventory build, and receivables timing. The FY2026 model itself shows -$0.1B FCF despite $0.4B net income. If working capital intensity does not improve as revenue scales, the company may need external financing despite management's self-funding framing, diluting equity holders or raising leverage. 2. Revenue concentration and timing risk around a small number of large deliveries. Management explicitly stated that the final position within the $4.4-4.6B guidance range depends on the timing of 'a small number' of year-end customer deliveries. This lumpiness means a modest slippage in handoff timing (e.g., Wyoming or New Jersey projects) could push revenue into a subsequent quarter or year, creating headline misses against a stock priced for continued beat-and-raise execution. 3. Growth deceleration multiple compression risk. The model assumes growth decelerates from 122% in 2026 to 35-40% in 2027 and 20-25% by 2029. If the market re-rates Bloom as a decelerating-growth industrial rather than a hyperscale-power compounder once growth normalizes below 30%, the 35x multiple applied here could compress toward a more traditional electrical equipment multiple of 15-20x, materially reducing price targets even if earnings are delivered as modeled. 4. Customer concentration in hyperscale data centers. The current growth thesis is heavily levered to a small number of hyperscale customers making large, discrete on-site power purchase decisions. Any pause in AI-driven data center capex, change in power sourcing strategy (e.g., a shift back toward grid interconnection or alternative distributed generation), or renegotiation of contract terms by a major customer would disproportionately affect Bloom given its current revenue base. 5. Execution risk on capacity expansion. Capex has quadrupled year-to-date to fund Fremont and Newark capacity additions. Manufacturing ramps of this magnitude carry execution risk - yield issues, labor constraints, or supply chain bottlenecks in critical materials could delay the capacity needed to fulfill the backlog growth management describes, risking both revenue timing and the margin expansion embedded in guidance. 6. GAAP-to-non-GAAP gap and earnings quality. Guidance is issued primarily on a non-GAAP basis ($3.10-3.40 EPS) while the modeled GAAP EPS path ($1.57 in FY2026) is notably lower, reflecting stock-based compensation and other reconciling items. Investors anchoring price expectations to non-GAAP guidance may be surprised by GAAP dilution effects, and any widening of this gap would undermine the credibility of the margin expansion story even if revenue targets are met.

📈 Price Targets