Bloom Energy makes the box that lets a hyperscaler put a 100 MW data center on a piece of land that does not yet have a grid connection. The product is the Bloom Energy Server, a solid oxide fuel cell that runs on natural gas, biogas, or hydrogen and converts it directly to electricity through a non…
Bloom Energy makes the box that lets a hyperscaler put a 100 MW data center on a piece of land that does not yet have a grid connection. The product is the Bloom Energy Server, a solid oxide fuel cell that runs on natural gas, biogas, or hydrogen and converts it directly to electricity through a non-combustion electrochemical reaction. Customers buy the boxes (or have them financed by infrastructure capital), Bloom installs them, then collects long-dated service revenue keeping them running for 15+ years.
The reason BE matters in 2026, after fifteen years of being a niche on-site power company with negative gross margins, is that grid interconnect queues for new US data centers run four to seven years, and AI capex cannot wait that long. Anyone building a gigawatt-scale AI factory now needs behind-the-meter generation that ships in months, not megawatts that arrive in 2031. Bloom is one of three companies (alongside reciprocating gas gensets and small modular reactors that do not yet exist commercially) that can credibly fill that gap. It is the only one operating at gigawatt scale today.
| Segment | What it is | Approx. % |
|---|---|---|
| Product | Energy Server hardware sales (the box itself) | ~74% |
| Service | Long-term O&M contracts (fixed-price, 15+ year terms) | ~14% |
| Installation | Site engineering and commissioning | ~10% |
| Electricity | Managed-services / PPA-style revenue | ~2% |
Source: Q4/FY2024 earnings release. Exact split for FY2025 ($2.02B revenue) not yet broken out at line-item level in the press release.
Business model: Hardware-plus-recurring-service. Each Energy Server sale drags a multi-decade service contract behind it. The economic logic (and historical pain point) is that Bloom records the service revenue but also bears the warranty replacement cost, so the service line was loss-making for years. Margin recovery has been the operational story since 2024.
Geographic mix: US-dominant. Korea is the second meaningful market, served exclusively through SK ecoplant. India, Japan, and Europe are smaller. Detailed split not disclosed at segment level.
Latest investor presentation: Q4 2025 earnings deck (Feb 6, 2026) and the July 2025 Investor Presentation.
Capacity guidance: management says the company is on track to double annual production capacity to 2 GW by end of 2026 (Utility Dive).
SK ecoplant (Korea — primary strategic partner since 2018). SK ecoplant is the environmental and energy-solutions arm of Korea’s SK Group, formerly SK E&C. They are Bloom’s exclusive distributor in South Korea and were a 10%+ shareholder. Cumulative equity investment hit ~$566M, including original $255M into zero-coupon non-voting redeemable convertible preferred at $25.50/share in 2021. The 2023 expanded offtake locked in 500 MW through 2027 worth roughly $1.5B in product revenue plus $3B in 20-year service revenue (Dec 2023 release). In July 2025 SK ecoplant trimmed its position by selling 10M shares via OTC block trade at $27.60 for ~$276M proceeds, but did not exit (Korea Times).
Oracle (April 2026 expansion). Master services agreement for up to 2.8 GW, with the initial 1.2 GW already contracted and being deployed. Oracle also received a warrant on April 9, 2026 to buy up to 3.53M shares at $113.28 (~$400M strike value), expiring October 9, 2026 (CNBC). This is the deal that has reset the BE narrative from “niche fuel cell stock” to “AI infrastructure pick-and-shovel.”
American Electric Power (AEP). Announced November 2024: framework deal for up to 1 GW, ~$2.65B over the term. As of January 2026, the first 100 MW order has been placed and AEP is calling it the world’s largest commercial fuel-cell procurement (Fuel Cells Works).
Brookfield (October 2025). Brookfield committed up to $5B in infrastructure capital to deploy Bloom systems across global AI factory sites (Bloom press release). Stock spiked 20% on the announcement.
Equinix. Decade-long customer that started in 2015 with a 1 MW pilot. February 2025 expansion took deployments past 100 MW across 19 IBX data centers, with ~75 MW operational and ~30 MW under construction.
| # | Customer | Ticker | Est. Revenue Share | Relationship |
|---|---|---|---|---|
| 1 | SK ecoplant | private (SK Group) | Historical 10%+ related-party customer | Exclusive Korea distributor + shareholder |
| 2 | Oracle | NYSE: ORCL | Largest forward backlog driver | Up to 2.8 GW frame, 1.2 GW contracted |
| 3 | American Electric Power | NASDAQ: AEP | Material once 1 GW frame draws | 1 GW frame, 100 MW initial order placed |
| 4 | Equinix | NASDAQ: EQIX | Reference customer, recurring | >100 MW across 19 IBX sites |
| 5 | Brookfield Asset Management / Infrastructure | NYSE: BAM, BIP | Capital partner not a direct revenue line | $5B AI deployment program |
Concentration risk: Historically high. SK ecoplant was a disclosed 10%+ customer in past 10-Ks. With the Oracle frame absorbing capacity through end-2027 per analyst commentary, single-customer concentration is rising not falling. The shape of the risk is changing too: from one strategic partner (SK ecoplant) to one hyperscaler (Oracle) to one infrastructure capital provider (Brookfield) all simultaneously. Top-1 and top-5 % from FY2024 10-K: not pulled in this session.
Dependency flags: Oracle has a $400M warrant struck below the current price, which means Oracle is also a future shareholder. That is unusual: a major customer with a warrant-based equity stake aligns interests but creates a dual relationship that needs disclosure scrutiny.
Bloom solves a problem that did not exist at scale until the AI capex cycle began: how to put 100 MW to 1 GW of clean firm power on a specific piece of land in 12 to 24 months when the local grid cannot deliver it for half a decade. Behind-the-meter SOFC is one of the few technologies that hits the speed and scale requirement today.
End-use applications (FY2026 narrative):
TAM framing. Goldman, JPMorgan and others now cite a US data center power TAM in the $200B–$500B range over the next decade depending on how AI compute demand scales. Bloom’s serviceable slice is the “off-grid or behind-the-meter, fast deploy” subset, which is smaller but where it has near-monopoly positioning at gigawatt scale. Specific company-cited TAM figures in the latest deck: not pulled in this session.
Secular tailwinds: 1. Grid interconnect timelines getting longer, not shorter 2. AI compute power demand outpacing utility-scale generation buildout 3. Hyperscalers willing to pay a premium for time-to-power 4. Natural gas as a transition fuel (US has it cheap and abundant) 5. Eventually, clean hydrogen optionality on the same hardware
The risk to the TAM thesis: if interconnect queues clear, or if SMRs commercialize on schedule, or if hyperscalers redesign for grid-only, Bloom loses its time-to-power moat. None of these are near-term threats but all three are real.
| Name | Title | Tenure | Background |
|---|---|---|---|
| KR Sridhar | Founder, Chairman & CEO | Founder (2001); CEO 25 years | Former director of NASA Ames Center for Space Power; the technology came out of his Mars-mission work on solid oxide cells. Has not stepped into Executive Chair role despite age and tenure. Holds combined Chair/CEO role. |
| Simon Edwards | CFO | April 13, 2026 (new) | Joined from Groq, where he was CFO and then briefly CEO. (Bloom press release) |
| Aman Joshi | Chief Commercial Officer | Multi-year | Owns the customer relationships including the Oracle and AEP wins. |
| Satish Chitoori | Chief Operating Officer | Multi-year | Manufacturing scale-up lead. |
| Maciej Kurzymski | Chief Accounting Officer | Multi-year | Served as interim CFO May 2025 through April 2026. |
| (Departed) Greg Cameron | Former CFO 2020–May 2024 | — | Now CFO at CF Industries. |
| (Departed) Daniel Berenbaum | Former CFO Apr 2024–May 2025 | — | Departed quickly, less than 14 months in role. |
The CFO churn is the loudest governance signal. Three CFOs in roughly two years (Cameron → Berenbaum → Kurzymski as interim → Edwards) is unusual for a company of this size in a complex capital-stack environment. It is not a smoking gun by itself but it warrants scrutiny on financial reporting controls and on whether the underlying issue is strategic disagreement, performance, or something else. Worth probing in the next /mgmt-dd pass.
| Name | Role | Independent? | Background |
|---|---|---|---|
| KR Sridhar | Chair (combined with CEO) | No | Founder |
| Jeffrey Immelt | Lead Independent Director (since May 2020) | Yes | Former Chairman/CEO of GE; brings industrial scale-up credibility |
| Jim Snabe | Director (since August 2025) | Yes | Chairman of Siemens AG; appointment signals strategic positioning toward European industrial customers |
| John T. Chambers | Director | Yes | Former Chairman/CEO of Cisco; brings hyperscaler/networking customer Rolodex |
| Mary K. Bush | Director | Yes | Long-tenured |
| Michael Boskin | Director | Yes | Former Chair of US Council of Economic Advisers |
| Cynthia Warner | Director | Yes |
10-member board total. Specific committee chairs (Audit, Comp, Nom/Gov) not pulled from latest DEF 14A (filed April 2, 2025) in this session.
Direct competitors in on-site / behind-the-meter power for AI data centers:
| Competitor | Technology | Status |
|---|---|---|
| Caterpillar (CAT) / Cummins (CMI) | Reciprocating gas gensets | Off-the-shelf, fast deploy, but burn-cycle emissions and lower efficiency |
| GE Vernova (GEV) | Aeroderivative turbines, hydrogen-ready | Larger scale, longer lead times, more grid-integrated |
| Mitsubishi Heavy / Siemens Energy | Gas turbines | Same |
| NuScale, Oklo, X-energy | Small modular reactors | Not commercial yet; relevant 2030+ |
| PowerSecure / CoreWeave-affiliates | Engineering-procurement-construction firms bundling generation | Emerging |
| Plug Power (PLUG), FuelCell Energy (FCEL) | Other fuel cell types (PEM, MCFC) | Smaller scale, weaker economics, distressed balance sheets |
Bloom’s moat: 1. Operating fleet at gigawatt scale — 1.5+ GW deployed, 15+ years of reliability data. Hyperscaler procurement teams risk-rank vendors on operating history; nobody else in the SOFC space is close. 2. Vertical integration — designs and manufactures the cells, the stacks, the system, and the service model. Most peers outsource one or more layers. 3. Customer reference list — Equinix, Oracle, AEP, SK. That is a reference set that takes a decade to build. 4. Time-to-power — the structural feature that matters most right now. SOFC deploys in months and produces firm power on natural gas immediately.
Competitive vulnerabilities: - Fuel cells run on natural gas in 2026 reality, not hydrogen. Long-term decarbonization story depends on H2 economics that are not yet there. - Reciprocating gensets are cheaper per MW upfront. Bloom’s pitch is lifecycle cost, not nameplate. - Service-segment economics are still the weakest line in the model.
Porter’s Five Forces snapshot: - Buyer power: Moderate-high. Hyperscalers are concentrated and price-sensitive but also speed-sensitive; right now speed is winning. - Supplier power: Moderate. Yttria-stabilized zirconia and other cell materials have multiple suppliers; nickel and ceramics processing is the bottleneck. - New entrants: Low near-term. SOFC manufacturing requires fifteen years of process knowledge. Korean and Chinese state-backed entrants are real long-term threats. - Substitutes: Reciprocating gensets (now), SMRs (later), grid-attached generation (always). - Rivalry: Moderate. The on-site SOFC space is a near-monopoly. The broader on-site power space is more competitive.
Valuation (as of April 27, 2026 close, $234.68)
| Metric | Value |
|---|---|
| Market cap | $66.7B |
| Enterprise value | $66.4B |
| P/E (TTM) | N/A (negative EPS) |
| Forward P/E | 75.2x |
| Price / Book | 85.5x |
| Price / Sales (TTM) | 33.0x |
| EV / Revenue (TTM) | 32.8x |
| EV / EBITDA (TTM) | 477.6x |
| FCF yield | ~0.09% (FCF $57M on $66.7B mkt cap) |
| Dividend yield | 0% |
| 52-week range | $16.05 – $242.20 |
| Beta | 3.19 |
The valuation is the single most important fact about this stock. BE trades at multiples that price in flawless execution against the mid-to-high analyst growth case for 3+ years. Forward EV/Revenue of ~12x on FY2027 consensus revenue ($5.2B) is more reasonable, but you are paying for the consensus to be right.
Income statement & margins
| Metric ($M) | FY2022 | FY2023 | FY2024 | FY2025 | FY2026E | FY2027E |
|---|---|---|---|---|---|---|
| Revenue | 1,201 | 1,335 | 1,474 | 2,024 | 3,227 | 5,237 |
| Revenue growth (YoY) | +23% | +11% | +10.5% | +37.3% | +59% | +62% |
| Gross margin | ~13% | ~15% | ~27.5% | ~29.0% | ~32% guide | n/a |
| Operating income | (261) | (209) | 23 | 73 | 425–475 (guide) | n/a |
| Operating margin | -22% | -16% | +1.6% | +3.6% | ~14% guide | n/a |
| Net income | (301) | (302) | (29) | (88) | n/a | n/a |
| Diluted EPS | ($1.62) | ($1.42) | ($0.13) | ($0.37) | $1.40 | $3.12 |
The story in this table is the operating-income inflection. After years of bleeding, Bloom flipped to GAAP operating profit in FY2024 and expanded materially in FY2025. Net income remains negative because of interest expense on the convertible stack and unusual items, but the underlying operating engine has turned.
Cash flow & balance sheet
| Metric ($M) | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|
| Operating cash flow | (192) | (373) | 92 | 114 |
| Capex | (117) | (84) | (59) | (57) |
| Free cash flow | (309) | (456) | 33 | 57 |
| Cash & equivalents | 348 | 665 | 803 | 2,454 |
| Total debt | 1,021 | 1,455 | 1,530 | 2,992 |
| Net debt | 63 | 182 | 326 | 164 |
| Stockholders equity | 341 | 502 | 562 | 769 |
| Shares outstanding (M) | 205.7 | 224.7 | 229.1 | 280.0 |
The FY2025 balance sheet jumped because of the upsized $2.2B convertible senior notes priced late 2025, plus an October 2025 debt-for-equity exchange that retired ~$443M of old converts in exchange for ~$449M cash and 18.1M new shares. Cash position more than tripled, gross debt nearly doubled, and share count grew 22% in one year. Net debt fell despite the gross debt buildup because cash grew faster.
Capital structure detail: - $350M green convertible senior notes due 2029, 3.00% coupon (May 2024 issuance) - ~$2.2B convertible senior notes priced late 2025 - ~280M shares outstanding plus 3.53M Oracle warrants (struck $113.28, exp Oct 9 2026) - SK ecoplant preferred stock previously converted; SK still holds residual common stake
What is fueling growth today: 1. Oracle / AI data center pull — 1.2 GW already contracted under the master agreement, frame extends to 2.8 GW. This is the single biggest revenue driver in 2026–2027. 2. AEP utility procurement — 100 MW initial order placed, frame for up to 1 GW. Sets precedent for other regulated utilities to procure SOFC behind-the-meter. 3. Backlog buildout — reportedly $20B post-Oracle expansion vs. ~$6B at end of FY2025. Provides 4–5 years of revenue visibility at FY2027 consensus. 4. Capacity doubling — 2 GW annual nameplate by end of 2026 is the rate-limiter on revenue conversion. 5. Brookfield financing partnership — $5B AI deployment vehicle removes the customer-financing constraint for non-investment-grade buyers.
Pipeline / R&D: - R&D spend not separately broken out in headline release; historically ~6–8% of revenue. - Hydrogen electrolyzer line (Newark) at 2 GW capacity. Commercial scale-up pending policy and offtake clarity (US 45V tax credit dynamics, EU Green Deal pacing). - Marine application (Samsung Heavy pilot) — current materiality not verified. - M&A activity: none material disclosed in last 12 months; capital is being deployed into capacity not acquisitions.
| Contract / Program | Counterparty | Value | Status (April 2026) | What’s Left | Revenue Impact |
|---|---|---|---|---|---|
| Oracle Master Services Agreement | Oracle (ORCL) | Up to 2.8 GW; 1.2 GW already contracted; warrant for 3.53M shares at $113.28 | Expanded April 13, 2026; warrant issued April 9, 2026; deployment underway | Capacity build-out at Fremont/Newark; site permits at hyperscaler facilities | Material 2026 and 2027 revenue contributor; underpins the FY2027 consensus jump to $5.2B |
| AEP Frame Agreement | American Electric Power (AEP) | ~$2.65B over up to 1 GW | Frame agreement announced Nov 2024; 100 MW initial firm order Jan 2026 | Subsequent draw orders against the frame; state PSC approvals in service territories | First 100 MW ships through 2026; cadence of follow-on orders is the swing factor for FY2027–2028 |
| Brookfield Infrastructure Partnership | Brookfield Asset Mgmt (BAM) / Brookfield Infrastructure (BIP) | Up to $5B | Announced October 2025; structure is project-finance, not direct purchase | Specific project sites announced over time | Brookfield is the financing layer. Revenue still flows from end-customers (hyperscalers); Brookfield underwrites them. |
| SK ecoplant 500 MW Offtake (2023 expansion) | SK ecoplant | ~$1.5B product + ~$3B service over 20 years | Active, running through 2027 | Order cadence | Steady ~$200M/year product run-rate plus growing service tail |
| Equinix 19-Site Expansion | Equinix (EQIX) | >100 MW deployed across 19 IBX sites | Active, ~75 MW operational + ~30 MW under construction (Feb 2025) | Site commissioning | Reference customer; demonstrates uptime track record more than it drives the P&L |
The growth story is fundamentally contract-driven. If Oracle and AEP draw against their frames at the pace currently implied by consensus, FY2027 revenue lands at $5.2B and the stock can grow into its multiple. If draws slip by 12 months, the consensus path slips and the multiple looks indefensible.
| Risk | Likelihood | Existing Mitigants | Mgmt De-risk Plan | Can It Be Closed? |
|---|---|---|---|---|
| 1. Customer concentration (Oracle as new top customer) | High | Diversifying customer base (AEP, Brookfield, Equinix); Oracle warrant aligns interests | Brookfield financing partnership broadens customer pool to non-IG buyers; hyperscaler outreach beyond Oracle | Partly. Concentration shifts shape (from SK to Oracle) but cannot be eliminated at this growth rate. Will improve naturally as new hyperscalers convert. |
| 2. Capital-stack expansion / dilution | High | $2.45B cash buffer; FCF positive 2 years running; no near-term debt maturities | Self-funding signal in FY2026 guide; convertible structure delays dilution until conversion | Partly. Bloom can stop issuing new converts now that operations fund themselves. The Oracle warrant (3.53M shares) is the next dilution event before October 2026. |
| 3. CFO turnover (3 in 2 years) / governance | Medium | New CFO Edwards has Groq pedigree; CAO Kurzymski stayed through transition | Edwards onboarding Q2 2026; need clean Q1 print on April 28 | Yes if Edwards stays 24+ months and earnings calls show financial-reporting maturity. Closes by 2027 if no further turnover. |
| 4. Natural gas price exposure (customer ROI sensitivity) | Medium | Customers pay for time-to-power, not lowest-cost-of-energy | Hydrogen optionality; long-term contracts hedge customer side | No. Structural. Can only be hedged not eliminated. Mitigates as gas/power price spread widens. |
| 5. Service segment economics | Medium | Margin trajectory improving; Sridhar called out service margin gains on FY2025 call | Predictive maintenance on installed fleet; warranty redesign on newer SKUs | Yes if Service crosses to GAAP positive contribution by FY2026. Watch line-item breakout. |
| 6. Permitting / interconnect risk for AEP-style deals | Medium | Behind-the-meter siting avoids ISO interconnect; gas-line tie-in well-understood | State-by-state engagement; AEP partnership de-risks regulated utility path | Partly. State PSC processes still gate volume. |
| 7. Valuation / momentum risk | High | n/a (this is a market risk, not a company risk) | n/a | No. The multiple compresses on any execution miss; this is a structural feature of the current setup. |
Last earnings (Q4/FY2025, reported February 6, 2026): - Revenue: $777.7M for Q4 (+35.9% YoY); FY2025 $2.024B (+37.3%) - FY2025 operating income: $73M (vs. $23M FY2024) - Operating cash flow FY2025: $114M; FCF $57M - FY2026 guide: revenue $3.1–3.3B, non-GAAP operating income $425–475M, gross margin ~32% non-GAAP
Next earnings (Q1 2026): TODAY, April 28, 2026, after market close. - Consensus revenue: $540M (+66% YoY) per Yahoo aggregation of 19 analyst estimates - Consensus EPS: $0.13 (vs. $0.03 LY, +328%) - High estimate $674M / low $410M (wide dispersion)
Material news in last 90 days:
| Date | Event |
|---|---|
| Feb 6, 2026 | Q4/FY2025 print, FY2026 guide |
| Mar 2026 | Stock down ~13% on weak US jobs report; Jefferies maintained Underperform with $97 PT |
| Apr 9, 2026 | Oracle warrant issued (3.53M shares at $113.28, $400M strike value) |
| Apr 13–14, 2026 | Oracle frame expanded to 2.8 GW; stock +15% on session |
| Apr 14, 2026 | Jefferies upgraded Underperform → Hold, PT $97 → $187 |
| Apr 14, 2026 | JP Morgan raised PT $166 → $231 (Overweight) |
| Apr 21, 2026 | UBS raised PT $170 → $251 (Buy) — street high; thesis is “800 VDC data center revolution” |
| Apr 21, 2026 | Citigroup raised PT $162 → $229 (Neutral) |
| Apr 22, 2026 | Baird raised PT $172 → $242 (Outperform) |
| Apr 27, 2026 | Barclays raised PT $153 → $177 (Equal-Weight) |
| Apr 28, 2026 | Q1 2026 earnings (after close) |
The April analyst-revision wave is the key context for tonight’s print. Most desks have raised prints and PTs in the last two weeks specifically pricing in Oracle. Q1 needs to confirm the order intake is real.
Note on data: Pulled from yfinance (which aggregates 13F + Form 4 filings) as of latest reporting dates. Cross-check against latest DEF 14A and EDGAR Form 4s for material recent moves. SK ecoplant’s residual common-stock holding is filed as 13G/13D-type and shows up in insider roster as “Beneficial Owner of more than 10% of a Class of Security.”
Major holders breakdown: - Insider ownership: 5.82% - Institutional ownership: 86.0% (institutional float ownership: 91.3%) - Number of institutional holders: 1,106 - Short interest: 23.2M shares = 9.34% of float; 2.17 days to cover
| Holder | Type | Who They Are | Shares | % | Notable |
|---|---|---|---|---|---|
| Ameriprise Financial | Active asset manager | Wealth-management parent including Columbia Threadneedle’s Seligman Tech franchise | 28.65M | 10.14% | Largest active holder; Seligman Tech is a sector-specialist holding ~17M of these shares |
| BlackRock | Passive index | World’s largest asset manager; mostly index inclusion | 24.16M | 8.55% | +16.9% QoQ — index-driven buying as BE entered/grew weight in indices |
| Vanguard Group | Passive index | Same as BlackRock structurally | 20.85M | 7.38% | Stable position |
| Situational Awareness LP | Active sector specialist | Leopold Aschenbrenner’s AI infrastructure-focused fund (formed 2024) | 10.08M | 3.57% | New position; 100% increase in Q4 2025. The thesis-driven bet in the holder list. |
| Morgan Stanley | Multi-strategy | IB / asset mgmt holdings | 9.35M | 3.31% | +6.3% |
| D.E. Shaw | Quant | Systematic / multi-strategy hedge fund | 8.80M | 3.12% | -25.6% (trimming) |
| State Street | Passive index | SPDR / institutional indexer | 6.03M | 2.14% | +13.4% |
| Geode Capital | Passive index | Fidelity sub-advisor | 5.28M | 1.87% | +5.4% |
| Jane Street | Market maker / quant | Systematic liquidity provider | 4.83M | 1.71% | +71.9% (likely options-related hedging buildup) |
| Graticule Asia Macro Advisors | Macro | Adam Levinson’s Asia-focused macro fund | 4.76M | 1.69% | Stable |
Read on the holder mix: Passive (Vanguard, BlackRock, State Street, Geode) accounts for roughly 20% of shares outstanding, which is normal for an S&P-eligible name. The interesting holders are Ameriprise/Seligman (active conviction position) and Situational Awareness (thesis-driven AI infrastructure specialist). The 71.9% jump in Jane Street’s position likely reflects derivatives-related hedging activity around the post-Oracle option expansion, not a directional view.
| Insider | Position | Shares Held |
|---|---|---|
| KR Sridhar | CEO / Chair | ~4.72M (stock award Feb 27, 2026) |
| John T. Chambers | Director | 431K (5/14/2025 award) |
| Daniel Berenbaum | Former CFO | 300K (legacy holdings) |
| Shawn Soderberg | Officer | 517K (sold Apr 15 2026) |
| Jeffrey Immelt | Lead Independent Director | 230K (Mar 31 2026 award) |
| Satish Chitoori | COO | 212K (sold Apr 14 2026) |
| Aman Joshi | CCO | 181K (sold Apr 1 2026) |
| Mary K. Bush | Director | 134K |
| Michael Boskin | Director | 102K |
| Cynthia Warner | Director | (recent grant) |
| SK ecoplant Co., Ltd | 10%+ Beneficial Owner | 13.49M (most recent filing reflects July 2025 sale, not residual after sale) |
Recent insider activity (last 6 months, summary): 19 buys for 513K shares, 26 sells for 611K shares, net ~98K shares sold. April 2026 is concentrated selling at $135–$225 by Soderberg, Chitoori, Joshi after the Oracle pop. Sridhar himself acquired 300K via PSU vesting in February, has not been a recent open-market seller.
Activist filers: None active as a 13D in this session. SK ecoplant historically files as 13G strategic. Worth running an EDGAR 13D search before earnings if Pink wants confirmation.
| Metric | Value |
|---|---|
| Number of analysts | 24 |
| Buy / Hold / Sell breakdown | 13 Buy (incl. 4 Strong Buy) / 12 Hold / 2 Sell |
| Consensus rating | Buy / Moderate Buy |
| Mean PT | $166.96 |
| Median PT | $174.00 |
| High PT | $251 (UBS) |
| Low PT | $55 |
| Current price (Apr 27, 2026) | $234.68 |
The current price sits above the mean target and roughly at the 90th percentile of the analyst range. Two read-throughs: 1. The street has not fully caught up to the post-Oracle reality. Several recent revisions (Baird $242, UBS $251, Citi $229, JPM $231) cluster above $225, suggesting the mean is being dragged down by stale low-side targets. 2. Or the rally has overshot and the mean target is the right anchor.
The wide dispersion ($55 to $251) is itself diagnostic: this is a bimodal stock where bears think the multiple cannot hold and bulls think the AI infrastructure thesis is just getting started. Q1 earnings tonight is the first inflection-test of the bull case.
A full /filings BE pass should pull: - Q1 2026 10-Q
(after tonight’s print) - 10-K FY2024 detail on segment revenue split,
customer concentration disclosures, related-party SK ecoplant terms -
DEF 14A April 2025 for committee chairs, exec compensation, succession
provisions - Form 4s for Sridhar, Edwards (new CFO), and recent officer
sales - Any 8-K events in last 90 days (Oracle warrant filing was
8-K)
EDGAR PDF retrieval was blocked in this session via the standard fetcher; manual download or alternate fetch needed.