BE · Bloom Energy Corporation · Stock research

Current Thesis

The demand leg an investor is buying on-site solid-oxide fuel-cell (SOFC) power as the fastest behind-the-meter fix for the grid-interconnect bottleneck starving AI datacenters is still ACCELERATING and cluster-confirmed. It re-rated hard through June: the FERC large-load interconnection order (2026-06-17/18) and the steel/aluminum tariff cut (effective 2026-06-08) drove a record $328.91 close on 2026-06-18 (+15.41%, intraday ATH $329.51), stacking on Oracle's 2.8 GW Project Jupiter (2026-04-27), the Nebius $2.6B/328 MW master agreement (2026-05-20), and the Q1 turn to GAAP profit (2026-04-28). What changed in July is the other side of the trade: two short reports Hunterbrook Media (2026-07-08) attacking the "China-free" supply claim and financial structure, and Crossroads Capital (2026-07-09) alleging a "supply wall" and hidden costs put an active bear campaign over the tape days ahead of the ~2026-07-30 Q2 print. Management rejected both (2026-07-09) and shares rallied on the rebuttal, holding the 200-day line into 2026-07-10. The demand story is intact; the setup is not clean. The asymmetric entry is a based, post-print structure with the short thesis refuted not the contested high with a binary event 19 days out.

Bullish and bearish views on Bloom Energy Corporation

The model's bull view on Bloom Energy Corporation (BE), in brief: FERC large-load order (2026-06-17/18): FERC directed RTOs/ISOs to make >20 MW interconnections faster and non-discriminatory, spotlighting the multi-year grid queue that pushes hyperscalers to self-generate. The bear view: Active short campaign (2026-07-08/09): Hunterbrook Media questioned the China-free supply claim and financial structure; Crossroads Capital alleged a "supply wall" physical inability to deliver the 2.8 GW Oracle + 328 MW Nebius backlog on schedule plus hidden costs. Both cases follow in full.

Bull Case

• FERC large-load order (2026-06-17/18): FERC directed RTOs/ISOs to make >20 MW interconnections faster and non-discriminatory, spotlighting the multi-year grid queue that pushes hyperscalers to self-generate. Bloom's own survey: 61% of developers would "bring their own power" if the grid falls short. Stock closed +15.41% at $328.91.
• Tariff reset (effective 2026-06-08, through 2027-12-31): steel/aluminum derivative duties cut to 15% from 25%, plus a 10% lane for capital equipment ≥85% U.S. "melted and poured" a direct margin tailwind on steel/aluminum-heavy SOFC hardware.
• Oracle Project Jupiter (2026-04-27): up to 2.8 GW SOFC (1.2 GW already in progress) for a Doña Ana County, NM AI campus; BE +25.7% on the news. The larger of the two marquee deals.
• Nebius master agreement (2026-05-20): 10-year, up to $2.6B, 328 MW (≈250 MW guaranteed), first project live in 2026; Daiwa moved Hold→Outperform, $324, same day.
• Q1 2026 print (2026-04-28): revenue $751.1M (+130.4% YoY), first GAAP net income $70.65M, adj EPS $0.44, non-GAAP gross margin 31.5%, adj EBITDA $143M, operating cash flow +$73.6M.
• Guidance raised: FY2026 revenue $3.40–$3.80B (midpoint growth ~80% YoY from ~60%), non-GAAP operating income $600–$750M, non-GAAP EPS $1.85–$2.25; backlog ~$20B as of Q1.
• Street still adding through the short attack: Baird reiterated Outperform, $310 (2026-07-09); Susquehanna raised its target to $298 (2026-07-10); even Hold-rated Jefferies lifted to $246 (2026-07-06). RBC $335 remains the bull high mark.

Bear Case

• Active short campaign (2026-07-08/09): Hunterbrook Media questioned the China-free supply claim and financial structure; Crossroads Capital alleged a "supply wall" physical inability to deliver the 2.8 GW Oracle + 328 MW Nebius backlog on schedule plus hidden costs. Both target the two load-bearing pillars: the tariff-tailwind narrative (if the supply chain is not China-free, the margin story weakens) and deliverability of the ~$20B backlog. A rebuttal is not a resolution.
• Price above most of the PT cluster: the $328.91 record (2026-06-18) prints above Daiwa $324, UBS $322, Baird/Morgan Stanley $310, Susquehanna $298, Bernstein $276 (Market Perform, 2026-06-16) and Jefferies $246 only RBC $335 sits above. Consensus is a headwind, not a floor.
• The June breakout was a +15.41% single-session move into a fresh ATH the classic chase candle. The 2026-07-10 "shorter-term pullback" confirms the spike is digesting.
• Binary earnings ~2026-07-30: a ~120x forward P/E on a beta-~3.75 name printing Q2 into an open short thesis is a two-sided event. A guide that fails to firmly defend the $3.40–$3.80B FY range hands the shorts the tape.
• Insider distribution: director Mary Bush filed to sell 25,000 shares into strength a soft supply signal alongside the short reports.
• Peer wobble: FuelCell Energy traded lower pre-market 2026-07-08, a reminder the fuel-cell cohort can move together on sentiment, not just fundamentals.

Setup & Price Structure

The name is off its $329.51 intraday ATH (2026-06-18) after the short reports interrupted a Monday 2026-07-06 surge, but it held the 200-day moving average through 2026-07-10 the structure is digesting, not broken. The relevant map: the $328–$335 supply zone (June ATH + RBC target) caps upside; the $310–$322 area is prior resistance-turned-pivot; the $280–$290 shelf is the pre-June-breakout base built after the Oracle/Nebius deals; and the 200-day line is the last line of demand defense. A fresh long at the contested high pays up for a name with an unresolved short thesis and a binary print 19 days out poor asymmetry. The cleaner structure is a higher-low retest into the $280–$310 band that holds the 200-day and the breakout base, ideally after the Q2 print de-risks the supply-chain and guidance questions. Chasing back toward $329 ahead of earnings is the trap this playbook exists to avoid.

Catalyst Calendar (next 30 days)

• ~2026-07-30 (est.): Q2 2026 earnings the binary. Watch whether management defends the FY2026 $3.40–$3.80B revenue range and directly answers the supply-chain/"supply wall" claims. Blackout window ~2026-07-27 (three trading days prior).
• Rolling (days): short-report rebuttal cycle potential Hunterbrook/Crossroads follow-ups, a third-party supply-chain audit, or customer confirmation that either corroborates or kills the "China-free" and "supply wall" allegations.
• Rolling: Oracle Project Jupiter / Nebius installation-milestone disclosures any timeline slip is a thesis dent; any milestone hit is a rebuttal.

Elapsed catalysts

• Rolling: analyst revisions clustering (Susquehanna $298 on 2026-07-10, Jefferies $246 on 2026-07-06, Baird $310 on 2026-07-09) watch for any post-short-report downgrade. _(passed 9d ago)_
• Ongoing: rate path small-cap debt costs hit a 6-year high (2026-07-07); a hawkish shift pressures high-beta capex names like BE independent of company news. _(passed 12d ago)_

What Would Change Our Mind

• Bull confirmation (would re-open a fresh entry): a weekly close reclaiming and holding above the $310–$322 pivot on a Q2 print that defends the FY range and forensically rebuts the supply-chain claims momentum re-establishing above the PT cluster with the short overhang removed.
• Bear confirmation (kills the leg): a weekly close below the $280–$290 shelf that loses the 200-day; the Hunterbrook/Crossroads supply-chain or "supply wall" claims gaining independent corroboration; or a Q2 guide that trims FY2026 revenue below the $3.40B floor. Any Oracle/Nebius timeline slip belongs in the same bucket.
• Theme flip: on-site-power-for-AI moving from ACCELERATING to SATURATED hyperscalers publicly favoring gas turbines or grid-scale nuclear over SOFC would remove the demand premium regardless of the short-report outcome.

Correlation Notes

Beta is roughly 3.75 size for 2–4x SPY daily moves. BE is not a diversifier against other AI-infrastructure longs; it correlates toward ~1 with the industrial-power-for-AI basket (VST, CEG, GE Vernova, OKLO, IREN, Nebius) and toward the fuel-cell cohort (FuelCell Energy traded lower 2026-07-08) in any capex-unwind or sentiment reversal. The July short campaign adds idiosyncratic single-name risk on top of that beta: a short report that lands would decorrelate BE lower even if the basket holds. Rate sensitivity is live the 6-year-high small-cap debt cost print (2026-07-07) means a hawkish macro turn hits this name harder than the index. Treat any concurrent industrial-power-AI exposure as one correlated position, not several.