BE — Bloom Energy Corporation

As of 2026-10-04 the tape and the sell side point the same way: the stock is +33.1% over one month and +113.2% over six months, the consensus rating is Buy (17 buy / 12 hold / 3 sell), and the consensus 12-month target of $289.87 sits almost exactly on the $289.15 price, with RBC at $335 Outperform and Jefferies just raised to $264 (Hold). The narrative is AI-infrastructure power: Bloom's new 800V DC-native architecture and 'Power Connect' format aimed at data centers, a 158,000 sq ft Fremont manufacturing expansion (announced late September 2026) that doubles its local footprint, an expanded MiTAC partnership, and Virginia's 2026 Energy Plan carving out a 'non-combustion gas resource' category targeting up to 5GW. Investors are paying for future capacity and order flow, not for current earnings.

Revenue is accelerating, not decelerating. Q2 2026 (period ended 2026-06-30) revenue of $1.065B was +166% vs. Q2 2025's $401M, and Q1 2026's $751M was +130% vs. Q1 2025's $326M — on top of Q4 2025's $778M (+36% YoY) and Q3 2025's $519M (+57% YoY). TTM revenue is $3.11B. Profitability is following: Q2 2026 operating income of $182M (17.1% margin) and net income of $196M ($0.62 diluted) versus a $42.6M net loss in Q2 2025; TTM operating margin is 11.2% and TTM net margin 7.9%. Two caveats keep this from being a clean line. First, results are extremely lumpy and back-half weighted — Q3 2025 operating margin was just 1.5%, Q1 2025 was -5.8%, and Q3 2023 posted negative gross profit. Second, the deterministic score's own notes flag structural distortion: revenue stepped +73% from Q3 2024 to Q4 2024, stockholders' equity jumped +75% ($921M to $1.612B) between 2026-03-31 and 2026-06-30, and total assets ballooned from $2.64B (Q3 2025) to $4.40B (Q4 2025). Direction of travel is unambiguously up; the multi-year curve is not purely organic.

Expensive on every measure, including forward. As of 2026-10-04 the trailing P/E is 346x on $0.853 of TTM EPS; even on the next-FY consensus EPS of $2.70 the forward P/E is still ~107x. Price/sales is 27.4x TTM versus a 5-year average of 4.6x — roughly six times its own history — and price/book is 51.5x on a 1.9% ROIC. EV/EBITDA per the ratio screener is ~268x. The PEG ratio of -0.016 is meaningless (negative growth base distorts it), so the usual 'growth justifies the multiple' check fails on the input side. The deterministic score's valuation context label, 'Rich vs. its own history,' is exactly right: this multiple is not merely high in absolute terms, it is unprecedented for this company.

The stock is discounting earnings that do not exist yet. TTM EPS is $0.853 and TTM net margin is 7.9%, while the $289.15 price implies the market is underwriting the ~$2.70 next-FY consensus — which in turn implies roughly a 20% net margin on $4.11B of revenue, about 2.5x today's margin. That is the entire bet: revenue scale plus fixed-cost absorption converting 33% gross margins into 20% net margins. If Bloom hits the revenue line but holds a ~10% net margin, EPS is closer to $1.30-1.40 and the multiple is 200x+. Practically, that means revenue beats and, more importantly, order/backlog disclosures tied to the Brookfield and AEP frameworks and AI data-center wins matter far more than the EPS line. A single quarter where revenue growth decelerates below ~50% would likely be read as the narrative breaking rather than as a soft quarter.

As of 2026-10-04, RSI(14) stood at 58.71 — firm but not overbought, so momentum alone is not a sell signal. Price was +6.2% above its 20-day SMA (~$272), +22.4% above its 50-day (~$236) and +39.4% above its 200-day (~$207): a textbook strong-uptrend configuration, but also a stretched one, with the 1-month return at +33.1% and 6-month at +113.2% (the 1-week return of +0.16% suggests the run has just stalled). The 52-week range of $75.70-$351.28 leaves the stock ~18% below its high. Implication: this is a chase-risky entry; a pullback into the $236-250 zone (the 50-day area) would offer materially better risk/reward, with the ~$207 200-day as deeper support.

• AI data-center power demand versus grid interconnection delays — chronic utility queue times are pushing hyperscalers toward behind-the-meter, fast-deploy onsite generation, which is Bloom's core demand driver.
• State regulatory tailwinds — Virginia's 2026 Energy Plan creates a 'non-combustion gas resource' category targeting up to 5GW of deployment between 2029 and 2035, a template other states may copy.
• Rates and cost of capital — with a beta of 3.811, BE is extremely rate-sensitive, and large multi-year corporate infrastructure deployments are financing-dependent; rate volatility can slow signings even when demand is intact.
• Federal clean-energy policy and incentive uncertainty — fuel-cell deployment economics remain tied to tax credits and to the levelized-cost comparison against natural gas and grid power.
• Electrification capex cycle and supply-chain scaling — ramping new manufacturing (Fremont) and securing stack inputs alongside turbine/transformer shortages that are themselves part of Bloom's sales pitch.

• Q3 2026 earnings on 2026-10-27 (estimated) — the first test of whether the Q2 2026 $1.07B revenue run-rate holds through the historically lumpy September quarter.
• Execution and conversion updates on large framework partnerships such as Brookfield and AEP.
• Buildout and utilization of the newly acquired 158,000 sq ft Fremont facility, which doubles local advanced manufacturing capacity.
• New order wins for the 800V DC-native data-center architecture and the 'Power Connect' format (claimed >40% shorter installation timelines, no transformers required).
• Further state-level adoption of clean-firm distributed generation policy modeled on the Virginia framework.
• Sell-side target revisions — the price-target range is already unusually wide ($176 low to $351 high), so consensus can move sharply on a single announcement.

• Valuation is the dominant risk: 27.4x TTM sales, 51.5x book and ~107x forward earnings leave essentially no cushion, and a 3.811 beta means the de-rating would be violent on any growth stumble.
• Profitability quality is still thin: a 1.9% ROIC and 7.9% TTM net margin mean today's earnings do not yet justify the capital employed, and the score's 100/100 growth leg is flattered by step changes.
• Cash flow is working-capital driven and seasonal: operating cash flow swung from -$213M in Q2 2025 to +$226M in Q2 2026 and FCF from -$220M to +$175M — smooth quarterly FCF should not be assumed.
• Balance-sheet opacity and dilution: total assets went from $2.64B (Q3 2025) to $4.40B (Q4 2025), LT debt rose to $2.47B against $2.67B of cash, and equity rose 75% in a single quarter — the financing structure and any future conversion/dilution are genuine unknowns.
• Customer and backlog concentration in a handful of very large framework deals; one contract delay or push-out moves the entire model.
• Competition from gas turbines, grid-supplied power and other distributed-generation providers, plus execution risk in scaling manufacturing fast enough to meet the AI-driven demand spike.

Bloom Energy is a genuinely improving business at a genuinely demanding price. The operational inflection is real — revenue up 166% YoY in Q2 2026, free cash flow positive, $2.67B of cash against $2.47B of long-term debt, and an AI-data-center power crunch that gives it a rare, hard-to-substitute product. But at 27x trailing sales (versus a 4.6x five-year average), 51x book and roughly 107x forward earnings, the shares already price in years of flawless execution and a margin structure (~20% net) the company has never demonstrated. I would not buy at $289.15 as of 2026-10-04; I would want either a pullback toward the $235-250 area or hard evidence that 33% gross margins are converting into durable net margin.

Blended multiple approach. (1) 80x forward P/E on next-FY consensus EPS of $2.70 = $216 (roughly 1.2-1.3x the ~60-70x forward multiple AI-power peers command, crediting Bloom's faster growth but penalising the 1.9% ROIC). (2) 21x forward EV/Sales on next-FY revenue of $4.11B = $86.3B EV, plus $0.2B net cash, divided by ~294.5M shares = ~$294. Weighting 60/40 toward the earnings method gives ~$247; I apply a further ~5% haircut for the seasonality and structural-distortion flags in the quarterly series, landing at $235 — about 19% below the 2026-10-04 price of $289.15 and below the $260.27 recent-average analyst target.