{
  "@context": "https://frontierpicks.com/schemas/dossier.v1.json",
  "ticker": "BE",
  "name": "Bloom Energy Corporation",
  "url": "https://frontierpicks.com/dossiers/BE/",
  "json_url": "https://frontierpicks.com/dossiers/BE.json",
  "status": "DORMANT",
  "current_conviction": "MEDIUM",
  "graded_conviction": null,
  "archetype": {
    "code": "a3",
    "n": 3
  },
  "current_thesis": "Q2 (2026-07-28) beat and raised — revenue $1.065B +165.5% YoY, FY26 guide lifted to $3.9–$4.2B — yet the 2026-08-14 close of $229.94 sits 33.5% below the 52wk high after July's 32% Hunterbrook-driven drawdown. Accelerating P&L into a deflating narrative premium: the $280–$290 shelf is lost, litigation runs to a 2026-09-28 deadline, and no company catalyst lands before the Q3 print.",
  "invalidation_trigger": "A weekly close below $200 negates the post-print recovery and returns price to the July breakdown range; secondary: FY2026 revenue guided under the $3.9B floor at the ~2026-10-28 Q3 print, or independent corroboration of the China-scandium sourcing allegations.",
  "catalyst_date": null,
  "outcome": "OPEN",
  "outcome_date": null,
  "invalidation_fired": null,
  "themes": [
    "industrial-power-grid",
    "ai-datacenter-infrastructure",
    "biofuels-low-carbon"
  ],
  "tags": [],
  "sources": [],
  "notes": [
    "Securities class action pending: class period 2025-02-27 to 2026-07-08, lead-plaintiff deadline 2026-09-28. The scandium-sourcing question stays open for several quarters.",
    "Brookfield's $25B (2026-06-30) and Oaktree/IDF's $1.7B (2026-07-16) are financing frameworks, not booked orders; revenue recognition follows physical deployment.",
    "High-beta name tied to the AI-power basket (VST/CEG/GEV/OKLO/NBIS) and to rates — the 2026-08-12 CPI-day move was macro, not company news.",
    "FY2026 guidance stands at $3.9-$4.2B revenue and $2.55-$2.85 non-GAAP EPS after the 2026-07-28 raise; the Q3 print is the first test of that path.",
    "Capital-intensive capacity ramp against $2.67B cash at 2026-06-30, with third-party partners carrying much of the deployment capex."
  ],
  "body_markdown": "## Current Thesis\nThe leg being bought is on-site solid-oxide fuel-cell (SOFC) power as the fastest behind-the-meter answer to datacenter interconnect queues. Since the last note that leg delivered on the numbers and lost on the tape. Q2 2026, reported 2026-07-28, put revenue at $1,065.4M, +165.5% versus $401.2M a year earlier, GAAP operating income $182.2M against a $3.5M operating loss in the prior-year quarter, non-GAAP EPS $0.78 versus the ~$0.39 consensus cited in coverage of the print, cash $2.67B at 2026-06-30, and FY2026 revenue guidance raised to $3.9–$4.2B from the prior $3.40–$3.80B range. Price went the other way: the stock fell 32% in July after Hunterbrook's 2026-07-08 report alleging undisclosed China sourcing of scandium, securities class actions followed, and the 2026-08-14 close of $229.94 sits 33.5% under the $345.85 52-week high. The $280–$290 shelf that framed the prior read is gone — that weekly-close condition triggered in July. What an investor is now underwriting is an accelerating P&L against a deflating narrative premium, with the legal overhang unresolved and no company-dated event until the Q3 print.\n\n## Bull Case\n- **Q2 2026 print (2026-07-28):** revenue $1,065.4M (+165.5% YoY), first quarter above $1B; GAAP gross margin 33.4% (26.7% prior year), non-GAAP gross margin 34.3%; non-GAAP operating income $239.6M; GAAP diluted EPS $0.62, non-GAAP $0.78.\n- **Guide raised, not defended (2026-07-28):** FY2026 revenue to $3.9–$4.2B, ~100% YoY growth at the midpoint; non-GAAP gross margin ~34%; non-GAAP operating income $800–$900M; non-GAAP EPS $2.55–$2.85. The July short case argued deliverability would force the guide down.\n- **Brookfield framework expanded 5x (2026-06-30):** AI-infrastructure financing commitment raised to $25B from $5B, drawn from Brookfield's AI Infrastructure Fund.\n- **Oaktree / IDF $1.7B project financing (2026-07-16):** a second third-party balance sheet funding deployment rather than Bloom's own.\n- **MiTAC expansion (2026-08-07):** islanded fuel-cell microgrid for an AI server manufacturing campus in Fremont, California — a repeatable template outside the two mega-contracts.\n- **Liquidity:** $2.67B cash at 2026-06-30 against a capital-intensive capacity ramp.\n- **Sell-side still above market:** consensus average target $273.51 across 29 analysts per S&P Global-compiled data cited 2026-08-14, versus the $229.94 close.\n\n## Bear Case\n- **A record print did not restore the high.** BE fell 13% on 2026-07-28 into the release, and the following sessions read as a two-way reset (2026-08-03 decline, 2026-08-04 rally). Beat-and-raise buying new highs is exactly what has stopped working.\n- **Litigation clock is running.** Complaints cover purchasers from 2025-02-27 through 2026-07-08 and allege Bloom understated reliance on China-sourced scandium obtained via intermediaries; lead-plaintiff deadline 2026-09-28 (Kaplan Fox, Robbins Geller notices, 2026-08-04).\n- **Durability claim is separate from sourcing.** Hunterbrook's work, cited in 2026-08-05 coverage, reported 37 metered New York systems falling below efficiency benchmarks in a median of 20 months and missing a 95% output target. Neither the Brookfield nor the Oaktree facility addresses field degradation.\n- **Jefferies raised its target to $229 while maintaining Hold (2026-08-14)** — a target set at the market, which is sell-side marking to price rather than leading it.\n- **Valuation still assumes persistence:** FY2026 non-GAAP EPS guidance of $2.55–$2.85 against a $229.94 close, with the bear scenario in the same S&P-compiled target set at $187.24 and the bull at $321.44.\n- **The bid is partly macro.** BE appeared in the 2026-08-12 CPI-day rally lists alongside Sandisk and Nebius; a rate-driven bounce is not company-specific confirmation.\n\n## Setup & Price Structure\n- Reference close 2026-08-14: $229.94. 52-week high $345.85, -33.5% below it. Three-month return -16.7%. RSI(14) 63.5.\n- The prior structure broke: the $280–$290 pre-June breakout shelf was lost during the July decline, and the weekly-close condition published on 2026-07-26 at $280 triggered.\n- Price is now working the $230–$250 zone previously identified as deeper pre-Nebius support. RSI 63.5 while a third below the high describes a bounce inside a broken structure; the base is not confirmed until price holds this zone through a weekly close and then reclaims $290.\n- **The narrative is maturing.** Operationally the narrative is still working — 2026-07-28 delivered the largest quarter in company history and a raised guide. Flow is what has moderated: the print produced no new high, coverage has shifted to daily-mover and options-flow list pieces (whale-activity lists 2026-08-04, 2026-08-07, 2026-08-11; mover posts 2026-08-05, 2026-08-06, 2026-08-12, 2026-08-14), and the marginal sell-side action is a Hold pinned at spot. That is a well-known story still delivering, with a thinner incremental bid than in June.\n- **Crowding / positioning observables:** retail-facing coverage clusters on at least six sessions in the first two weeks of August; target dispersion runs $187.24 to $321.44 around a $273.51 average; Jefferies sits at $229 with a Hold; there is no scheduled company catalyst inside the next 30 days to force resolution; the filings feed shows no new insider transactions in the post-print window through 2026-08-14.\n\n## Catalyst Calendar (next 30 days)\n- **2026-08-16 → 2026-09-15:** no confirmed company-scheduled event. The window is empty of forcing functions, which leaves price to macro and headline flow.\n- **2026-09-28** — lead-plaintiff deadline in the securities class action (class period 2025-02-27 to 2026-07-08). Consolidation of counsel and the amended complaint that follows are the first look at what the plaintiffs actually have.\n- **~2026-10-28 (est.)** — Q3 2026 print. First test of the raised $3.9–$4.2B revenue and $2.55–$2.85 non-GAAP EPS path, and the first quarter where MiTAC-type awards can show in orders.\n\n## What Would Change Our Mind\nThe structural break already happened in July: $280–$290 gave way and has not been reclaimed, so the burden now sits on the $230–$250 zone holding. A weekly close below $200 would negate the post-print recovery and put price back inside the July breakdown range, at which point the accelerating-fundamentals read is being rejected by the market on two consecutive quarters of good news. On the fundamental side, FY2026 revenue guided below the $3.9B floor at the Q3 print, or non-GAAP gross margin guided under ~34%, would break the raise that answered the July bear case. Independent corroboration of the China-scandium allegations — a surviving amended complaint, a disclosed regulatory inquiry, or a supply-chain revision in a 10-Q — would re-open the question the financing announcements did not settle. In the other direction, a weekly close back above $290 with the Q3 guide intact would restore the pre-July trend and the frame that broke in July.\n\n## Correlation Notes\n- BE trades with the AI-power complex (VST, CEG, GEV, OKLO, NBIS, IREN) and with rates. The 2026-08-12 CPI reaction lifted it alongside Sandisk and Nebius; the 2026-08-03 commentary treating the Situational Awareness episode as a clearing event tied the same names together.\n- Fuel-cell peers moved as a group on 2026-07-28: BE -13%, FuelCell -7%, Plug -4% into Bloom's own report, so a share of the July de-rating was sector-level rather than purely idiosyncratic.\n- It offers no diversification against other AI-infrastructure exposure; in a capex-expectations unwind the correlation to that basket rises toward one.\n- Financing partners are now a shared dependency: Brookfield (2026-06-30) and Oaktree/IDF (2026-07-16) both intermediate deployment capital, so a credit-spread widening event touches the deployment schedule and the equity at the same time.",
  "first_seen": "2026-05-22",
  "last_analyzed": "2026-08-16T14:28:14+00:00",
  "last_synthesized": "2026-08-16",
  "last_update_source": "watchlist_research",
  "license": "Content © FrontierPicks. Cite the canonical URL."
}