간추려 보기쉬운 말로 요약 · 먼저 읽어 보세요
GE Vernova builds the hardware of electricity supply: gas and nuclear generation equipment, wind turbines, transformers and switchgear. The report's call is Watch. It is three businesses on one platform, and Q2 2026 made the split plain: Power, the largest, ran an 18.8% segment EBITDA margin and the fast-growing Electrification 18.4%, while Wind lost 275 million USD of EBITDA at a negative 13.6% margin. The consolidated 11.3% adjusted EBITDA margin describes none of them.
Q2 orders rose 88% organically to 24.2 billion USD, more than double the quarter's revenue, and backlog reached 176.3 billion. Of the gas queue, 53 GW is firm equipment backlog; the other 63 GW is slot reservations, which the report reads as scarcity evidence, not booked business. Management then raised 2026 free cash flow guidance to 11.5 to 12.5 billion USD, while revenue guidance moved only one billion. Roughly 11.7 billion of first-half operating cash came from working capital, mostly customer down-payments, so the report sets the headline 4.8% FCF yield against a normalized owner-earnings yield of only 1.4% to 1.8%.
The moat sits in two of the three. Power's installed fleet carries decades of parts, outages and upgrades an owner cannot casually move to another OEM. Electrification's HVDC systems, substations and large transformers need certification, factory capacity and years of execution, and Prolec adds scarce North American transformer capacity. Wind earns no comparable moat, and with Siemens Energy, Mitsubishi Heavy Industries and ABB in the field the report treats turbine scarcity as a cycle, not a monopoly.
On price the report is blunt. At 941.95 USD the stock sits outside all three bands: above the 590 to 780 USD acceptable hold zone, below the roughly 1,190 USD clearly-overvalued line, and far above the 280 to 320 USD ideal buy price. Enterprise value is about 21.5 times management's own 2028 EBITDA target, so a buyer today pays in advance for an improvement still to be delivered. There is no margin of safety at this price.
Three risks carry the weight. Advances normalizing would cut reported free cash flow sharply even in a healthy quarter. Wind orders fell about 40% in Q2, the full-year loss is guided near 400 million USD, and the 2028 recovery is unproved. Valuation compression needs no recession: cutting the base terminal multiples to 70% takes the report's base value from about 716 to 513 USD per share. The verdict: a better business than the prior report modeled, at a price that pre-spends too much of the 2028 outcome, so the rating stays Watch and the report waits for a better price. The above is a summary of the report's views and does not constitute investment advice. Markets carry risk; invest with caution.
핵심 요약GE Vernova is the former GE power platform: gas and nuclear generation equipment carrying a large installed-base service annuity, grid equipment sold into a transformer and HVDC shortage, and a loss-making wind business. Q2 2026 separated the three cleanly, with Power at an 18.8% segment EBITDA margin and Electrification at 18.4% while Wind lost $275m at a negative 13.6% margin; management then raised FY2026 free cash flow guidance 71% at the midpoint to $11.5-12.5bn even though roughly $11.7bn of H1 operating cash came from working capital, dominated by customer advances. Rating Watch: the business is worth materially more than the prior report modeled, but $941.95 sits above the $590-780 acceptable-hold band and normalized 2026 owner earnings of $3.5-4.5bn yield only 1.4-1.8%.
본문의 가격은 발행 시점 기준입니다. 최신 실시간 가격은 위 밸류에이션 밴드를 참고하세요.
Meta
- Ticker: GEV.US
- Company: GE Vernova Inc.
- Price & market cap: 941.95 USD per share; approximately 250.87 billion USD market capitalization, as of 2026-09-04 close. Google Finance reports 266.33 million shares outstanding; the June 30 10-Q reported 266.33 million shares outstanding, making the market-cap arithmetic internally consistent.
- Currency: USD
- Report date: 2026-09-06
- Industry: Power Generation Equipment
- One-line positioning: Integrated power-equipment manufacturer spanning gas and nuclear generation, wind turbines, and grid equipment, with 176.3 billion USD of backlog at Q2 2026.
Research summary
GE Vernova is best understood as three economically different companies sharing an industrial platform. Power is a gas-and-nuclear equipment franchise with a large installed-base service annuity. Electrification is a grid-equipment business now benefiting from transformer, switchgear, HVDC and substation scarcity. Wind is a restructuring problem whose equipment economics remain poor. Q2 2026 makes the contrast unusually clear. Power produced 5.48 billion USD of revenue at an 18.8% segment EBITDA margin; Electrification produced 3.64 billion USD at 18.4%; Wind produced about 2.03 billion USD and lost 275 million USD of EBITDA, a negative 13.6% margin. The consolidated 11.3% adjusted EBITDA margin describes none of those three operating realities.
That split is the central fact behind this refresh. The market is principally trading GE Vernova as a scarce supplier to a global electricity-capacity buildout: AI data centres need firm power, utilities need grid reinforcement, the gas-turbine supply chain has long lead times, and transmission equipment remains capacity-constrained. The external demand evidence is substantial. The IEA projects global data-centre electricity consumption to roughly double to about 945 TWh by 2030, with the United States accounting for the largest part of the increase. Data centres account for roughly half of projected US electricity-demand growth through 2030. The IEA also estimates that annual grid investment must rise roughly 50% from about 400 billion USD today by 2030. EIA expects 6.3 GW of new US gas-fired generating capacity in 2026 alone.
GE Vernova is converting that backdrop into orders faster than it is converting orders into revenue. Q2 orders were 24.2 billion USD, up 88% organically, versus 11.1 billion USD of revenue. Total RPO reached 176.3 billion USD. Power alone signed 20 GW of new gas-equipment contracts during Q2: 18 GW of slot-reservation agreements and only 2 GW of firm orders. It converted another 10 GW of previously signed reservations into firm backlog. Firm gas-equipment backlog ended the quarter at 53 GW, while slot reservations were 63 GW, or 116 GW combined. Management expects at least 125 GW under contract by year-end.
The 116 GW headline needs discipline. At the 20 GW annual turbine-production run-rate targeted for Q3 2026, 116 GW is equivalent to about 5.8 years of output, but only 53 GW is firm equipment backlog, equivalent to about 2.7 years at that run-rate. The other 63 GW is a different contractual animal. A publicly disclosed reservation agreement between GE Vernova and Maxim Power illustrates the structure. The customer pays a non-refundable deposit to hold a manufacturing slot, the deposit is credited toward a later purchase, and the parties then negotiate the definitive equipment sale. In that example, even the final equipment purchase price remained to be agreed. GE Vernova does not publicly disclose a standardized SRA deposit percentage, universal cancellation clause or standardized repricing formula. So I treat SRAs as economically meaningful evidence of scarcity and customer intent, but not as equivalent to RPO or booked revenue.
There is no defensible way to convert the quarter's Power order dollars into an implied turbine price per GW. Power's 16.7 billion USD of orders include gas services, nuclear and other products as well as turbines, while SRAs themselves are not all recognized as firm orders. Dividing that dollar figure by the 20 GW of new gas contracts would mix incompatible numerator and denominator definitions. I do not report a spurious USD/GW figure.
The other half of the market narrative is margin expansion. In 2025, Power earned a 14.7% segment EBITDA margin and Electrification 14.9%. By Q2 2026 those margins had moved close to 19%. GE Vernova's current 2028 framework calls for roughly 56 billion USD of consolidated revenue and a 20% adjusted EBITDA margin, with 22% segment margins in both Power and Electrification and 6% in Wind. Those targets already include the acquisition of the remaining 50% of Prolec GE.
The Prolec acquisition matters strategically because it adds transformer manufacturing into the part of the market where scarcity is most acute. GE Vernova paid 5.275 billion USD for the remaining half of the joint venture. Management's pre-close forecast put Prolec at about 4.2 billion USD of 2028 revenue, 1.1 billion USD of adjusted EBITDA, a roughly 27% margin and about 0.6 billion USD of free cash flow. For the acquired half, the transaction therefore equated to roughly 9.6 times management's 2028 incremental EBITDA before synergies, materially below GE Vernova's own public-market multiple.
The refresh, however, turns on cash flow more than on revenue. The prior in-house report used FY2026 FCF guidance of 6.5–7.5 billion USD. Against a capitalization around 280 billion USD, that is indeed a 2.3–2.7% FCF yield. The arithmetic behind the prior thesis was sound. That guidance was in force after Q1 2026: GE Vernova had raised it from 5.0–5.5 billion USD to 6.5–7.5 billion USD.
At Q2, management raised 2026 FCF guidance again, this time to 11.5–12.5 billion USD, while revenue guidance moved only one billion dollars higher to 45.5–46.5 billion USD. At the current 250.87 billion USD capitalization, the new headline FCF yield is approximately 4.6–5.0%, with a midpoint of about 4.8%. The prior report's most important numerical input has therefore changed radically.
The quality of that incremental cash prevents me from simply replacing a 2.5% yield with a 4.8% yield. GE Vernova generated 10.68 billion USD of operating cash and 9.90 billion USD of FCF in the first half of 2026, but working capital contributed about 11.7 billion USD. Contract liabilities and current deferred income alone contributed about 13.7 billion USD as Power collected more down-payments on gas orders and slot reservations and Electrification collected more project advances. Contract liabilities reached roughly 40.0 billion USD at June 30, up approximately 14.1 billion USD in six months. Those cash receipts are valuable. They fund factories with customer capital and reduce financing requirements. They are also obligations to deliver future equipment and services, and a steady-state order book cannot produce a fresh 13–14 billion USD step-up in advances every six months indefinitely.
My owner-earnings estimate therefore sits well below reported 2026 FCF. GE Vernova does not disclose maintenance and growth capex separately. It has committed roughly 6 billion USD of capex for 2025–28, including about 1 billion USD at Prolec, and H1 2026 cash capex was 783 million USD. PP&E depreciation alone was about 760 million USD in H1. From those disclosures I estimate maintenance capex at roughly 0.8–1.1 billion USD annually, and normalized 2026 owner earnings, assuming no further working-capital inflow, at roughly 3.5–4.5 billion USD. This is an analytical estimate rather than company guidance. On that basis the current owner-earnings yield is only about 1.4–1.8%.
By 2028, the calculation improves considerably. If GE Vernova reaches roughly 56 billion USD of revenue and a 20% EBITDA margin, EBITDA would be approximately 11.2 billion USD. After normalized taxes, recurring capex and other cash requirements, I estimate through-cycle owner FCF of roughly 7.5–8.5 billion USD. That is a much better business than today's accounting earnings suggest, but at the current capitalization it still corresponds to only about a 3.0–3.4% yield on a level of cash flow that requires another two-plus years of execution.
The share-price record also argues against reading price action as a verdict on business quality. The prior report used 1,038.74 USD as its May 25 reference. Yahoo's historical series shows 1,060.39 USD on May 22, so I treat 1,038.74 as the prior report's reference price rather than claiming it was an exchange closing print on May 25. The stock subsequently closed at 1,174.86 USD on June 30, then retreated. Q2 earnings on July 22 produced a 5.8% one-day decline because adjusted EBITDA missed market expectations and Wind losses widened despite the extraordinary orders and FCF-guidance raise. By August 13 the shares were around 1,039.90 USD; they closed September 4 at 941.95 USD, about 20% below the June 30 close and roughly 9% below the prior report's stated reference. No split occurred.
The prior report's unusual structural conclusion was that even its bull value, 980 USD, was below the market price. I do not reach that same structural conclusion. My work places reasonable base value materially above the prior 430–560 USD range and produces an optimistic fair-value case around 970–1,080 USD, with the clearly-overvalued line set about 10% above that upper bound. The change comes from four things: much stronger verified Power and Electrification margins, a substantially larger and better-priced backlog, the Prolec economics, and a lower share price. The 2026 FCF raise helps, but much less than the headline number suggests because customer advances explain most of the incremental cash.
The resulting qualitative portrait is a re-rating. A formerly troubled collection of GE energy assets has become a scarce, increasingly profitable supplier into a genuine power-capex upcycle. The market has correctly recognized that transformation. The remaining disagreement is how long current gas and grid scarcity persists, whether 22% segment margins survive after manufacturing supply catches up, and how much investors should pay today for earnings that management expects to earn in 2028 and beyond.
Vertical history and financial record
GE Vernova's legal history is short; its industrial history is exceptionally long. The company did not emerge from a conventional startup or IPO. It is the collection of General Electric's power-generation, renewable-energy and grid assets, separated after GE decided in 2021 to divide itself into three public companies. GE stockholders received one GE Vernova share for every four GE shares held as of the record date. The distribution became effective before the market opened on April 2, 2024, and GEV began regular trading on the NYSE that day. There was no IPO bookbuild, IPO price or primary capital raise. Ownership began as a pro-rata distribution to the former GE shareholder base.
The important prehistory has four economic stages.
The first was the creation of the installed base. GE spent more than a century building power-generation machinery and, later, servicing it. That legacy matters more today than the age of the corporate entity because gas and steam turbines remain serviceable for decades, and service contracts, outages, replacement components and upgrades attach to installed equipment. In the 2023 carve-out disclosure, services represented roughly 68% of Power revenue and Power carried about 59 billion USD of services RPO. By June 2026, Power services RPO had risen to roughly 72.4 billion USD.
The second stage was the Alstom expansion, which nearly broke the economics of the legacy power business before eventually becoming strategically useful. GE announced a 13.5 billion USD enterprise-value offer for Alstom's thermal, renewables and grid operations in 2014 and completed the Power and Grid acquisition in November 2015 for about 10.3 billion USD. The strategic logic was scale, complementary technology and grid capability. The timing was disastrous: the global large-gas-turbine market weakened sharply, GE had too much manufacturing capacity, and the expected economics failed. In 2018 GE recorded a 22 billion USD non-cash goodwill impairment related to Power.
That failure is still relevant. Today's bull case depends partly on assets that were once written down because demand disappeared. Manufacturing plants, service capability and grid technologies purchased at a terrible point in the prior cycle can become valuable in a supply-constrained cycle. Management itself has subsequently argued that some Alstom assets may ultimately prove among the most valuable parts of the inherited portfolio. The valuation lesson matters as much. Turbine scarcity is cyclical enough that extrapolating today's order intensity indefinitely would repeat the opposite version of the mistake GE made in 2015.
The third stage was repair. From roughly 2019 through the spin, GE's energy operations simplified the portfolio, reduced structural cost, improved underwriting and lived through substantial renewable-energy losses. Offshore wind was particularly damaging because large fixed-price contracts collided with inflation, supply-chain constraints, installation problems and a changing policy environment. The eventual independent company inherited both the service franchise of old GE and the project liabilities of that era. The sale of part of Steam Power's nuclear activities to EDF in 2024 is one example of continued portfolio simplification.
The fourth stage began with the April 2024 spin. GE Vernova arrived with about 274 million shares issued in the separation and a business that still looked, on consolidated numbers, like an industrial turnaround. In 2024 revenue was 34.94 billion USD, net income 1.56 billion USD and adjusted EBITDA 2.04 billion USD, a 5.8% margin. FCF was only 1.70 billion USD. By 2025, revenue had risen 9% to 38.07 billion USD, adjusted EBITDA to 3.20 billion USD and the margin to 8.4%; FCF more than doubled to 3.71 billion USD. The 2025 GAAP net-income figure of 4.88 billion USD overstates underlying improvement because it included a 2.9 billion USD tax benefit from releasing a US valuation allowance.
The business underneath those consolidated figures was already splitting in two directions. In 2025 Power revenue increased 9% to 19.77 billion USD and EBITDA rose to 2.90 billion USD, a 14.7% margin. Electrification revenue jumped 28% to 9.64 billion USD and EBITDA more than doubled to 1.43 billion USD, producing a 14.9% margin. Wind revenue fell 6% to 9.11 billion USD and lost 598 million USD of EBITDA.
The financial acceleration continued into 2026, but accounting became more difficult to read. GE Vernova acquired the remaining half of Prolec in February. The accounting remeasurement of the stake it already owned generated a large non-cash gain, making Q1 net income 4.75 billion USD even though adjusted EBITDA was only 896 million USD. Six-month GAAP net income of roughly 5.4 billion USD tells little about recurring earnings power.
The cash statement looks almost like the mirror image: it contains very real cash but an unusually large timing benefit. The eight-quarter path is instructive.
| Quarter | Revenue, USD bn | Net income, USD bn | Operating cash flow, USD bn | Capex implied by FCF, USD bn | Free cash flow, USD bn |
|---|---|---|---|---|---|
| Q3 2024 | 8.9 | -0.1 | 1.1 | 0.1 | 1.0 |
| Q4 2024 | 10.6 | 0.48 | 0.92 | 0.35 | 0.57 |
| Q1 2025 | 8.03 | 0.26 | 1.16 | 0.19 | 0.98 |
| Q2 2025 | 9.1 | 0.49 | 0.4 | 0.2 | 0.2 |
| Q3 2025 | 10.0 | 0.45 | 1.0 | 0.3 | 0.7 |
| Q4 2025 | 10.96 | 3.67 | 2.48 | 0.67 | 1.81 |
| Q1 2026 | 9.34 | 4.75 | 5.19 | 0.40 | 4.79 |
| Q2 2026 | 11.1 | 0.65 | 5.49 | 0.39 | 5.11 |
Source: GE Vernova quarterly filings and earnings releases; some quarterly values are rounded because the company presents headline figures in billions. Q4 2025 net income includes the 2.9 billion USD tax benefit; Q1 2026 includes roughly 4.5 billion USD of pre-tax business-purchase and disposal gains.
That table explains why a simple CFO/net-income ratio is misleading. For the full years 2023–25, aggregate operating cash flow was approximately 8.76 billion USD versus aggregate GAAP net income of about 5.96 billion USD, or roughly 1.47 times. Adding H1 2026 raises the ratio further. Yet neither side represents clean recurring economics: GAAP net income contains disposal gains and tax benefits, while cash flow contains large advance collections. Public-company comparability before the 2024 spin is also limited because the pre-spin statements were combined carve-out accounts with GE allocations.
The 2026 cash bridge is the most important part of the vertical review. H1 operating cash flow was about 10.68 billion USD and cash capex about 0.78 billion USD, leaving 9.90 billion USD of FCF. Working capital generated approximately 11.7 billion USD. The dominant component was roughly 13.7 billion USD of additional contract liabilities and current deferred income, offset partly by inventory, receivables and contract-asset absorption. Management explicitly attributed the increase primarily to Power down-payments on orders and slot reservations and Electrification down-payments.
Contract liabilities therefore rose from roughly 26.0 billion USD at year-end 2025 to about 40.0 billion USD by June 2026. Power represented approximately 27.7 billion USD, Electrification 9.1 billion USD and Wind 3.2 billion USD. These liabilities are not financial debt, but they are not freely distributable excess cash either: GE Vernova owes equipment or service performance against much of that money.
Even so, the balance sheet is strong. June cash was about 13.1 billion USD. Long-term borrowings were around 2.85 billion USD after GE Vernova issued 2.6 billion USD of senior notes, partly to finance Prolec, and the company had large undrawn committed credit facilities. S&P and Fitch had upgraded the company to investment-grade BBB and BBB+ ratings, respectively, in December 2025.
One balance-sheet item deserves more attention than leverage: goodwill. Prolec pushed total goodwill to approximately 9.7 billion USD by Q2 2026, including about 5.9 billion USD in Electrification and 3.2 billion USD in Wind. The Electrification goodwill sits against rapidly improving economics; the Wind goodwill sits against a segment still losing money. A prolonged failure to reach the 2028 Wind recovery assumptions would raise impairment risk even if the impairment itself were non-cash.
Capital allocation has evolved quickly. The buyback authorization was increased from 6 billion USD to 10 billion USD. In 2025 GE Vernova repurchased roughly 8.2 million shares at an average price near 406 USD. In Q1 2026 it repurchased about 1.8 million shares for 1.3 billion USD at an average 720 USD. By Q2 it had bought another roughly 2.5 million shares for about 2.35 billion USD, close to 940 USD per share. Cumulative program spending was approximately 7.0 billion USD by June, leaving about 3.0 billion USD of authorization.
The Q1 purchase at 720 USD is about 24% below today's price and was strongly accretive if today's valuation proves defensible. The Q2 buying is more revealing. Management was still willing to deploy billions around roughly the present quotation. That is a stronger signal than the 720 USD transaction alone, but investors should not turn it into a substitute for valuation. The remaining 3 billion USD authorization is only around 1.2% of current market capitalization. The quarterly dividend was doubled from 0.25 to 0.50 USD, or 2.00 USD annualized, yielding only about 0.21% at 941.95 USD.
The share price has run through three capital-market narratives since the spin. The first was turnaround: investors had to decide whether the energy assets could become consistently profitable outside GE. The second was backlog-and-margin proof: 2024–25 showed Power pricing, Electrification growth and higher FCF. The third, which dominates 2026, is scarcity. Investors are now paying for long-dated gas slots, transformers, grid projects and data-centre power demand. The shift from turnaround multiple to growth-infrastructure multiple is economically justified; the open question is how much scarcity rent survives after capacity catches up.
The most useful dated path around this refresh is:
| Date | Price or move | Fundamental context |
|---|---|---|
| 2026-05-22 | 1,060.39 close | Nearest verified prior close before the prior report's stated May 25 base date |
| 2026-05-25 | 1,038.74 prior-report reference | Supplied by the prior report; treated here as a reference price rather than a verified exchange close |
| May 2026 | -10.6% for the month | Valuation reset and concern after management cautioned around data-centre and wind-project timing |
| 2026-06-30 | 1,174.86 close | Rebound toward the cycle's peak as power-demand/scarcity narrative remained strong |
| 2026-07-22 | -5.8% on the day | Q2 adjusted EBITDA disappointment and deeper Wind losses outweighed the guidance raise |
| 2026-08-13 | 1,039.90 | Partial recovery after Q2 |
| 2026-09-04 | 941.95 close | Current research reference; about 20% below June 30 |
I would not attribute every daily fluctuation to a news item. The cleanest causal event is July 22, when Reuters explicitly tied the 5.8% decline to the EBITDA miss and Wind deterioration. The late-May weakness also followed public caution from CEO Scott Strazik about the pace of data-centre and wind projects. Other movements contain broader AI, industrial and rates factors that cannot be isolated reliably.
Business model, moat, industry and peers
Power is the economic foundation. In 2025 it generated 19.77 billion USD of revenue and 2.90 billion USD of segment EBITDA. Q2 2026 revenue was 5.48 billion USD and EBITDA 1.03 billion USD. Gas Power is the largest contributor, but Nuclear and Hydro matter. Services are especially important because an installed turbine creates decades of inspections, parts, repairs, upgrades and contractual service opportunities. Power RPO at June 30 was about 111.6 billion USD, of which roughly 72.4 billion USD was services.
This service layer is the first real moat. A power plant owner cannot casually swap the OEM providing proprietary turbine parts, engineering and major outage services. The turbine sale itself is therefore partly customer acquisition for a very long-lived aftermarket. The value of today's 20→24→30 GW production ambitions cannot be measured only from the equipment gross profit; each additional installed machine expands the future service base.
The second moat is manufacturing scarcity combined with qualification. Management expects a 20 GW annual gas-turbine run-rate in Q3 2026, 24 GW in 2028 and is taking actions toward 30 GW in 2030. The earlier capacity plan indicated that the first step to 24 GW includes roughly 2 GW of additional capacity in Belfort and 2 GW in Greenville, using additional shifts, lean manufacturing and targeted investment rather than building an entirely new production system. That creates attractive operating leverage if demand persists, because part of the output increase comes through utilization and productivity.
The margin arithmetic is powerful. On my roughly 31 billion USD base-case 2028 Power revenue, every percentage point of segment margin is about 310 million USD of EBITDA. Management's progression from 14.7% in 2025 to a 22% target represents more than 2 billion USD of potential annual EBITDA uplift before considering revenue growth. Pricing, volume and productivity have already taken the margin to 18.8% in Q2 2026, so the target no longer requires a hypothetical turnaround from zero. It requires sustaining roughly another three percentage points from a high current base.
The risk is that scarce turbine slots are a cycle, not a permanent monopoly. Siemens Energy and Mitsubishi Heavy Industries remain credible global alternatives. Customers with multi-gigawatt fleets actively value supplier diversification, technological efficiency and fuel flexibility. New capacity at all three OEMs can eventually ease scarcity. GE Vernova's service installed base protects the aftermarket more effectively than it protects the price of a new turbine.
Electrification is the fastest change in the corporate mix. Revenue increased from 7.55 billion USD in 2024 to 9.64 billion USD in 2025, while the segment EBITDA margin moved from 9.0% to 14.9%. Q2 2026 revenue reached 3.64 billion USD and the margin 18.4%. Equipment RPO reached about 40.6 billion USD, up 69% year on year including Prolec, and the quarter's order book-to-bill was roughly 1.7 times.
The underlying products are not interchangeable commodities at project scale. HVDC systems, substations, switchgear and very large transformers require engineering certification, factory capacity and years of project execution. The moat is a mixture of installed engineering know-how, manufacturing capacity, qualification and customer switching costs. GE Vernova's 2025 investor materials expected Electrification backlog to approximately double from around 30 billion USD in Q3 2025 toward roughly 60 billion USD by 2028, before the full effect of Prolec. That is management's projection, not a guaranteed outcome, but the Q2 2026 equipment backlog of 40.6 billion USD shows substantial progress already.
Prolec strengthens this moat by adding transformer capacity in North America. Its forecast 25–27% EBITDA margins are higher than legacy Electrification's current margin, so consolidation is accretive to segment quality as well as growth if those forecasts hold. The acquisition also raises integration and purchase-accounting complexity, but the industrial logic is much cleaner than GE's 2015 Alstom deal because the purchased business is already profitable and capacity-constrained.
Wind is different. In 2025 it lost 598 million USD of EBITDA on 9.11 billion USD of revenue. Q2 2026 revenue fell 10%, orders dropped around 40%, and the EBITDA loss widened to 275 million USD. First-half 2026 segment EBITDA was already negative 657 million USD, so the approximately 400 million USD full-year loss guidance requires a positive second half. The business still has value in its installed fleet and service relationships, but current equipment execution does not evidence a comparable moat.
The cost structure explains the asymmetry. Turbine and grid manufacturing have significant fixed plant and engineering costs, so volume and price produce operating leverage. Long-duration wind projects also contain large fixed obligations, but in the wrong direction. Once a poorly priced offshore contract is signed, inflation, installation delays, warranty issues or tariffs can increase costs faster than revenue while management remains contractually obliged to perform. In 2025 the segment was hit by lower offshore activity, project delays and tariffs; Q1 2026 again cited tariffs and higher offshore contract losses.
GE Vernova's R&D and capex burden is therefore necessary rather than optional. The company has committed around 6 billion USD of capex and 5 billion USD of R&D from 2025 through 2028, including Prolec. Those investments support gas capacity, grid manufacturing and new technologies. They also mean that valuing the company on EBITDA without subtracting reinvestment exaggerates distributable economics.
Industry demand is unusually supportive. IEA expects US electricity demand to grow nearly 2% annually through 2030, after years of much slower growth, with data centres responsible for around half of the increase. Natural gas and coal are expected to meet more than 40% of incremental global data-centre electricity needs through 2030 in the IEA's analysis, reflecting the difficulty of satisfying rapidly growing, 24-hour loads solely with intermittent sources. That improves the near-term case for gas turbines even while renewables and nuclear keep expanding.
Grid spending is the more durable of the two secular drivers. Annual grid investment has to rise substantially simply to connect new generation, replace aging infrastructure and handle load growth. Unlike a particular data-centre project, transformers and transmission are required across generation technologies. This is why I assign Electrification a higher through-cycle quality score than new-equipment gas even though both are currently strong.
Policy risk is concentrated unevenly. Gas and grid demand depend principally on reliability and load growth, while Wind is much more exposed to permitting, subsidies, tariffs and local-content policy. GE Vernova estimated the 2026 company-wide tariff cost at roughly 100–200 million USD after mitigation in its Q2 filing, lower than earlier estimates. Offshore projects have also experienced US permitting interruptions. These are manageable at group level today but disproportionately important to an already loss-making Wind segment.
The horizontal comparison confirms that GE Vernova's current strengths are real, but not unique.
Siemens Energy is the closest structural peer. Its Q3 FY2026 orders hit a record 17.9 billion euros and backlog reached 162 billion euros. Gas Services had comparable revenue growth of 16.6% over the first nine months and a 16.6% profit margin before special items; Grid Technologies grew 22.3% and earned an 18.3% margin. Siemens Gamesa, previously the group's problem child, reported a positive 2.7% quarterly margin in Q3 after substantial losses in earlier periods.
That comparison matters. GE Vernova's 18.8% Power and 18.4% Electrification margins are no longer merely turnaround margins; they are already in the same economic neighborhood as Siemens Energy's strongest franchises. GE Vernova currently has the advantage in the gas order queue and a stronger Q2 Power margin, while Siemens Energy has shown faster progress in repairing wind. Both companies have effectively become gas-plus-grid scarcity plays with wind attached.
Mitsubishi Heavy Industries is the third critical gas OEM. Its Q1 FY2026 consolidated order intake rose to about 2.02 trillion yen, up more than 400 billion yen year over year, and its recent disclosures continue to describe strong GTCC demand, particularly in North America and Asia. MHI's competitive importance is greater than its direct valuation usefulness because it mixes energy with aerospace, defence and other industrial businesses. Customers choose it for large-frame gas-turbine efficiency, reliability and an alternative to the GE/Siemens installed-base ecosystem.
ABB provides a cleaner grid-quality benchmark. Q2 2026 orders rose 30% to 12.0 billion USD, revenue rose 14% to 9.48 billion USD and group operational EBITA margin reached 20.2%, with return on capital employed of 28.4%. ABB's portfolio is more short-cycle and less project-heavy than GE Vernova's grid business, so it deserves a quality premium, but its economics show that 20%-plus electrical-equipment margins are achievable when product mix, channel power and capacity discipline align.
Eaton and Vertiv show how aggressively the stock market is capitalizing clean exposure to electrification and data-centre infrastructure. As of September 4, Eaton traded around 41.8 times trailing earnings and Vertiv around 63.5 times. Those multiples are not appropriate inputs to copy into GE Vernova's SOTP: they indicate that the entire electrical/data-centre complex carries a large narrative premium.
Vestas is the opposite comparison. Its Q2 2026 revenue rose 26% to about 4.7 billion euros, EBIT margin before special items improved to 9.4%, and combined order backlog reached 76.9 billion euros. Vestas still faces a difficult wind industry, but its current profitability means GE Vernova Wind cannot reasonably be valued as if it were an average profitable turbine OEM.
| Operating cross-section | GEV Power | GEV Electrification | GEV Wind | Closest external benchmark |
|---|---|---|---|---|
| Latest reported margin | 18.8% EBITDA, Q2 | 18.4% EBITDA, Q2 | -13.6% EBITDA, Q2 | Siemens Gas 16.6% and Siemens Grid 18.3%, both nine-month; Vestas 9.4% EBIT, Q2 |
| 2025 revenue | 19.8bn USD | 9.6bn USD | 9.1bn USD | Different reporting bases |
| Backlog signal | 111.6bn USD RPO | 44.6bn USD total RPO | materially weaker orders | Siemens Energy 162bn EUR group backlog |
| Main moat | installed fleet, service, turbine slots | engineering, qualification, constrained factories | installed service fleet | varies by peer |
| Current economic position | scarcity + margin expansion | structural growth + margin expansion | restructuring | Siemens Energy mirrors the mix most closely |
GE Vernova and peer figures are drawn from each company's latest primary disclosures. Accounting measures are not perfectly comparable, which is why the qualitative conclusions rely on direction and business structure rather than treating EBITDA margin and EBIT margin as identical.
The ecological niche is therefore unusual. GE Vernova is one of very few companies that can sell both the large generating machine and critical parts of the grid that carries its output. Its profit pool is currently being taken from utility and data-centre capital budgets under conditions of limited manufacturing capacity. Siemens Energy is the closest company capable of attacking both pools. MHI attacks gas directly; ABB, Eaton and Schneider Electric attack the electrification pool; Vestas shows the standalone economics against which Wind must ultimately be judged.
The real moat is concentrated in Power services, scarce gas manufacturing capability and high-voltage grid engineering. Wind does not currently earn the same moat multiple.
Current fundamentals, cash flow and valuation
Q2 2026 was one of those quarters in which the headline income statement understated the strategic progress and the cash-flow statement overstated the recurring economics.
Revenue rose 22% reported and 12% organically to 11.1 billion USD. Orders rose 88% organically to 24.2 billion USD. Net income was 649 million USD and adjusted EBITDA approximately 1.25 billion USD, an 11.3% margin. Total backlog increased roughly 13 billion USD sequentially to 176.3 billion USD.
The segment bridge is what matters:
| Q2 2026 | Power | Electrification | Wind |
|---|---|---|---|
| Revenue | 5.477bn | 3.637bn | 2.026bn |
| Revenue growth | strong double-digit | strong double-digit | -10% |
| Segment EBITDA | 1.031bn | 0.671bn | -0.275bn |
| EBITDA margin | 18.8% | 18.4% | -13.6% |
| Order signal | 16.7bn; gas queue expanding | 6.3bn; ≈1.7x book-to-bill | ≈1.2bn; about -40% YoY |
Source: Q2 2026 10-Q and earnings release.
Power and Electrification together generated roughly 1.70 billion USD of segment EBITDA before Wind consumed 275 million USD and before corporate items. That is why a 100–200 basis-point change in Wind is much less important than sustaining 18–22% margins in the two good businesses. Wind can still destroy value through project liabilities, but it does not drive the group's upside case.
Guidance was raised substantially:
| FY2026 guidance | Before Q2 | After Q2 |
|---|---|---|
| Revenue | 44.5–45.5bn | 45.5–46.5bn |
| Adjusted EBITDA margin | 12–14% | 12–14% |
| Free cash flow | 6.5–7.5bn | 11.5–12.5bn |
| Electrification revenue | 14.0–14.5bn | 14.5–15.0bn |
| Wind EBITDA | about -0.4bn | about -0.4bn |
The prior guidance is documented in the Q1 release; the current guidance is the July 22 company disclosure.
The revenue raise is roughly 2% at the midpoint. The FCF raise is about 71% at the midpoint, from 7.0 to 12.0 billion USD. The cash has arrived far ahead of the associated revenue. The H1 working-capital bridge explains why: customers are financing production through advances and progress collections.
This is economically better than debt-funded expansion. It creates negative working capital, reduces GE Vernova's need to finance factories itself and makes each new order less capital-intensive. It is also non-repeatable at the present rate once the order book stops accelerating. If contract liabilities simply stabilize, their incremental contribution to CFO becomes zero even though the underlying business remains healthy. If deliveries run ahead of new advance collections, the cash-flow contribution reverses.
That distinction changes the FCF-yield debate. On current guidance:
- 11.5 billion / 250.87 billion = about 4.6%.
- 12.5 billion / 250.87 billion = about 5.0%.
On my 3.5–4.5 billion USD estimate of normalized 2026 owner earnings with stable working capital, the yield is approximately 1.4–1.8%. On my 7.5–8.5 billion USD steady-state 2028 estimate after the planned margin expansion, the yield on today's market capitalization is about 3.0–3.4%.
The last figure is the economically useful anchor. It says an investor buying today is effectively paying in advance for a large part of the 2028 earnings improvement.
The GAAP P/E is even less useful. Google Finance reports about 26.9 times trailing earnings, but trailing EPS contains the 2025 tax-allowance release and Q1 2026 Prolec remeasurement gain. A roughly 60-times current normalized owner-earnings valuation is closer to the underlying economic starting point than the headline 27-times GAAP P/E.
Enterprise value gives the same warning. At 941.95 USD, equity value is about 250.9 billion USD. Subtracting June cash of 13.1 billion USD and adding roughly 2.85 billion USD of long-term borrowings gives an enterprise value around 241 billion USD before pension and other adjustments. Against the FY2026 midpoint of about 6.0 billion USD of EBITDA implied by 46 billion USD of revenue at a 13% margin, GE Vernova trades around 40 times current-year EBITDA. Against management's 2028 56 billion USD / 20% target, the multiple is roughly 21.5 times 2028 EBITDA before discounting that EBITDA back more than two years.
The expectation gap is clearer after imposing a required return. To earn roughly 9% annually to 2028, today's 250.9 billion USD equity value would need to become approximately 306 billion USD by then. Assuming 12 billion USD of net cash at that point implies about 294 billion USD of enterprise value. At management's 11.2 billion USD 2028 EBITDA target, the stock would still need to trade at approximately 26 times EBITDA in 2028. Or, at a 20-times exit multiple, required 2028 EBITDA is roughly 14.7 billion USD, equivalent to about a 26% margin on 56 billion USD of revenue. These are my valuation calculations, not management forecasts.
That is why the current price requires more than “management hits 2028.” It requires either material upside to the 20% margin, materially more than 56 billion USD of revenue, a very high terminal multiple, or some combination.
A sum-of-the-parts is mandatory because applying one multiple to all three segments would capitalize a negative-13.6%-margin Wind business at the same rate as an 18%-plus grid franchise.
My SOTP assumptions are deliberately below the clean-electrification public-market P/E multiples and are based on 2028 segment economics.
| SOTP dimension | Conservative | Base | Optimistic |
|---|---|---|---|
| Power 2028 revenue | 28.0bn | 31.0bn | 33.0bn |
| Power EBITDA margin | 18% | 22% | 24% |
| Power EV/EBITDA | 15x | 18x | 21x |
| Electrification 2028 revenue | 16.5bn | 18.5bn | 21.0bn |
| Electrification EBITDA margin | 18% | 22% | 24% |
| Electrification EV/EBITDA | 17x | 20x | 23x |
| Wind 2028 revenue | 6.0bn | 6.5bn | 6.5bn |
| Wind EBITDA margin | -2% | 6% | 8% |
| Wind value | -5.0bn EV | ≈2.3bn EV | ≈5.2bn EV |
| 2028 net cash assumption | 8bn | 12bn | 18bn |
| 2028 share count | 260m | 250m | 245m |
| 2028 equity value/share | ≈497 | ≈874 | ≈1,247 |
| Present value/share | ≈402 | ≈716 | ≈1,032 |
The 2028 base assumptions deliberately map closely to management's present 56 billion USD revenue and 20% group-margin framework while keeping individual segments economically separate. Management's published segment targets are 22% for Power, 22% for Electrification and 6% for Wind.
Wind receives negative 5 billion USD of enterprise value in the conservative case because continued losses can require cash, restructuring and warranty performance. In the base case I give it only about 2.3 billion USD, six times the approximately 390 million USD EBITDA implied by 6.5 billion USD of revenue at the company's 6% target margin. In the optimistic case, the business receives about 5.2 billion USD. Even in the bull case, Wind is a small part of total value.
The base-case valuation is therefore overwhelmingly a Power-and-Electrification judgment. If those two businesses deserve lower through-cycle multiples, the equity value falls rapidly.
I cross-check SOTP with normalized owner cash flow rather than reported 2026 FCF. The DCF assumptions are:
| Absolute-valuation dimension | Conservative | Base | Optimistic |
|---|---|---|---|
| 2028 normalized owner FCF | 5.5bn | 8.0bn | 10.5bn |
| 2029–33 annual FCF growth | 4% | 7% | 10% |
| Terminal growth | 2.5% | 3.0% | 3.5% |
| Discount rate | 10.0% | 9.0% | 8.5% |
| Approx. DCF value/share | 320 | 600 | 1,000 |
| SOTP value/share | 402 | 716 | 1,032 |
| Research fair-value envelope | 320–420 | 600–760 | 970–1,080 |
| Implied annualized return to 2028 SOTP terminal value | about -24% | about -3% | about +13% |
| Permanent-loss trigger | Power/grid scarcity normalizes before new capacity earns returns | 2028 margin targets missed and multiple normalizes | even high margins fail to support a premium terminal multiple |
This is valuation-scenario analysis within a research framework, not investment advice.
The DCF is intentionally harsher than SOTP because a large proportion of terminal value depends on cash flows after 2028. That is the right asymmetry for a company whose current price already discounts a large future margin uplift.
Maintenance capex is the unavoidable weak point in owner earnings. GE Vernova discloses total capex, not maintenance versus growth. H1 cash capex of 783 million USD was close to PP&E depreciation, while management is simultaneously investing to expand gas and grid capacity. I estimate annual maintenance requirements at roughly 0.8–1.1 billion USD, with total cash capex above that as capacity grows. If someone deducts maintenance only and assumes every growth dollar can be stopped without hurting backlog conversion, owner earnings looks materially better. I reject that treatment because the growth capex is part of the cost of delivering the earnings already embedded in today's valuation.
The most fragile valuation assumption is the terminal multiple, not 2026 revenue. Cutting my base-case segment multiples to 70% of their assumed levels, while leaving 2028 revenue and margins unchanged, takes the SOTP present value from roughly 716 USD to about 513 USD per share. A rerating from 18–20 times EBITDA to roughly 13–14 times can destroy almost 30% of base intrinsic value without any operational recession.
Historical valuation percentiles are not especially meaningful here. GEV has only traded independently since April 2024, its earnings composition has changed rapidly, and GAAP earnings contain major one-offs. The more useful historical observation is that the market has shifted from valuing a turnaround to capitalizing a long-duration scarcity story. The 52-week high is 1,195.94 USD, but being roughly 21% below that peak does not make the current 941.95 USD cheap.
Peer valuation also fails as a stand-alone defence. Eaton's roughly 42-times P/E and Vertiv's roughly 63-times P/E show that investors are paying exceptionally high prices across data-centre/electrification exposures. ABB's 20% operational margin shows what great industrial economics can look like. Neither observation proves that a 26-times 2028 EBITDA exit multiple is appropriate for GE Vernova.
The margin-of-safety recheck is severe.
Current price is more than twice my conservative present-value midpoint. Margin of safety relative to the conservative case is zero.
The most fragile base assumption is the terminal multiple. As noted above, reducing that assumption to 70% pulls base SOTP value toward roughly 513 USD.
If earnings remain flat for three years and the valuation multiple merely stays unchanged, the directly visible cash return is essentially the 2.00 USD annual dividend, about 0.21% at the current quotation. Even allowing for further repurchases, a flat-earnings case produces a return far below a normal long-duration equity hurdle and below the US Treasury's September 4 government-bond benchmark. The appropriate dated reference is the Treasury curve on the same September 4 base date. I do not quote a specific 10-year yield here because I could not confirm one for that date.
There is no margin of safety at this buy price.
The company can be substantially more valuable than the prior report estimated and still be unattractive at 942 USD. Those two conclusions are entirely compatible.
Risks, catalysts, tracking dashboard, uncertainties and sources
The first permanent-loss risk is a gas-order normalization before the new capacity earns its cost. I assign medium probability and high impact. Management is moving from a 20 GW annual run-rate toward 24 GW in 2028 and potentially 30 GW in 2030. The observable indicators are SRA conversion, firm backlog and factory output. If SRAs stop converting while capacity spending continues, the market will first cut the assumed duration of growth, then lower the terminal margin and multiple. The transmission path is orders → utilization → Power margin → 2028 EBITDA → valuation.
The second is working-capital normalization. Probability is high; impact on economic value is medium but impact on the stock narrative could be high. A 13.7 billion USD H1 increase in contract liabilities cannot recur indefinitely at the same pace. A healthy future quarter can show sharply lower FCF simply because orders and advances stop accelerating. Investors using 12 billion USD of 2026 FCF as a permanent run-rate would then discover that they capitalized financing cash as recurring owner earnings. The indicators are contract liabilities, equipment order intake, inventory and FCF before working-capital movements.
The third is Wind. Probability of continued volatility is high; impact is medium at group level but potentially high if contract losses become much larger than today's 400 million USD annual-loss guide. Q2 Wind orders fell roughly 40%, while EBITDA margin was negative 13.6%. The indicators are quarterly equipment orders, offshore contract-loss provisions, service profitability and whether the business is credibly approaching the 6% 2028 margin target. A failure here subtracts cash directly and also raises the probability that the roughly 3.2 billion USD Wind goodwill is impaired.
The fourth is valuation compression. Probability is medium-to-high and impact is high because this risk requires no recession. Today's price needs very strong terminal economics. If the market decides that Power and Electrification should trade at mature-industrial rather than scarcity multiples, my base SOTP can fall from about 716 USD to approximately 513 USD even with operational assumptions unchanged. This is the clearest permanent-loss mechanism at the current price.
The fifth is execution at higher turbine output. Probability is medium, impact high. Moving from 20 to 24 and later possibly 30 GW requires labor, supplier capacity, quality control and capital. A turbine problem does more damage than a missed unit sale because it can create warranty expense and weaken the services reputation on which lifecycle economics depend. Output rate, on-time delivery, warranty provisions and Power margin are the observable indicators.
Tariffs and energy policy are a secondary but real risk. GE Vernova's Q2 estimate of 100–200 million USD of 2026 tariff cost after mitigation is manageable against group earnings, but Wind is disproportionately exposed. A material new tariff regime or another offshore-wind permitting stop would hit an already weak segment before affecting Power or Electrification materially.
Positive catalysts over the next twelve months are straightforward: more than 125 GW of combined gas backlog and reservations by year-end, continued double-digit Electrification growth with an 18–20% margin, further Power margin expansion without a deterioration in SRA quality, Wind losses coming in below the 400 million USD guide, and continued conversion of reservations into firm RPO. A further 2028 margin or revenue upgrade would matter much more to valuation than another quarter of advance-funded FCF.
Negative catalysts are the mirror image but should be interpreted asymmetrically. A slowdown in new SRAs is not automatically bearish if firm backlog remains large; failure to convert existing SRAs is. Lower contract liabilities are not automatically bearish if they reflect successful deliveries; lower liabilities accompanied by weak orders and falling cash are. Wind losses are most dangerous when they rise alongside falling orders, because that means the business is consuming cash without rebuilding its future revenue base.
The tracking dashboard I would use is:
| Indicator | Current/reference | Normal path | Alert threshold |
|---|---|---|---|
| Gas firm backlog + SRA | 116 GW Q2 | ≥125 GW by YE2026 | <120 GW or sequential decline |
| Firm gas backlog alone | 53 GW Q2 | continued conversion | two quarters without material growth |
| SRA conversion | 10 GW in Q2 | roughly 5–10+ GW/quarter during current ramp | <5 GW/quarter for two quarters |
| Gas production run-rate | target 20 GW in Q3 2026 | 24 GW in 2028 | 20-GW target missed or 24-GW plan delayed materially |
| Power EBITDA margin | 18.8% Q2 | 17–19% FY26; toward 22% by 2028 | <16% for two quarters |
| Electrification EBITDA margin | 18.4% Q2 | 18–20% FY26; toward 22% | <17% for two quarters |
| Wind EBITDA | -275m Q2 | FY26 ≈-400m | quarterly loss >250m again with weak orders |
| Contract liabilities | ≈40.0bn | growth broadly linked to backlog | sharp decline plus falling orders/FCF |
| Total RPO | 176.3bn | conversion plus net growth | equipment RPO growth <10% with capacity still expanding |
| EV / 2028 target EBITDA | ≈21.5x spot | ideally compresses through earnings growth | >25x without another outlook raise |
| Next earnings | market estimate 2026-10-22 | Q3 update | date not yet confirmed by company at base date |
Operating references come from GE Vernova's Q2 filings and 2028 outlook. The October 22 date is a market-calendar estimate, not an official company announcement as of this base date.
The most useful dashboard relationship is SRA conversion versus firm backlog. A high reservation total without conversion is a marketing metric; conversion plus deposits plus rising RPO is evidence of real commercial scarcity. The second is FCF versus contract liabilities. If FCF falls while underlying EBITDA rises and contract liabilities normalize, the business may still be healthy; the stock may nonetheless react poorly because today's price capitalizes an unusually high cash print.
The industry's external indicators are equally useful. IEA's US electricity-demand forecast and grid-investment requirements tell us whether the structural backdrop is intact. Siemens Energy's Gas Services and Grid Technologies orders show whether demand is industry-wide or being lost specifically by GE Vernova. MHI's GTCC intake is an early indicator of competitive gas share. ABB/Eaton order growth indicates whether electrical scarcity remains broad. Vestas provides the cleanest external check on whether GE Vernova's Wind problems are industry-wide or company-specific.
Research uncertainties are material in five places.
First, GE Vernova does not disclose a standard SRA deposit amount, so I cannot reliably separate the approximately 40 billion USD of contract liabilities into gas reservations, firm equipment advances, grid project payments and service cash. The Maxim contract proves that at least some SRAs use non-refundable deposits and later definitive pricing, but one customer's terms cannot be generalized to every agreement.
Second, maintenance capex is not disclosed separately. My 0.8–1.1 billion USD estimate is an analytical range based on depreciation, current cash capex and the identified capacity-growth program. A different maintenance estimate changes owner earnings.
Third, GE Vernova does not disclose enough unit-level pricing and cost information to calculate a reliable incremental margin for each additional GW from 20 to 24 to 30 GW. The 2028 Power margin target gives the aggregate economic outcome management is aiming for; any per-GW profit estimate would introduce false precision.
Fourth, the prior in-house Siemens Energy, MHI, Vestas, Goldwind, ABB, Eaton, Schneider, Hyosung, Vertiv, Constellation and Vistra research reports were not retrievable in this research session. I rebuilt the peer work from current public primary disclosures rather than inheriting their ratings or valuation frameworks.
Fifth, the next earnings date had not been officially confirmed by the company in the retrieved investor materials as of the research base date; October 22 is a third-party calendar estimate.
The principal source base for this report is GE Vernova's Q2 2026 10-Q and earnings release, Q1 and FY2025 releases, the 2025 annual report and investor update, SEC filings covering the spin, peer-company earnings releases, IEA and EIA demand data, US Treasury rate data, and dated market-price sources.
Cross-synthesis and final research conclusion
Vertically, GE Vernova has proven one capability above all: it can repair industrial assets that looked structurally impaired when they were buried inside old GE. Power's 2018 crisis was real. The 22 billion USD impairment reflected a genuine collapse in the economic assumptions behind GE's Alstom-era expansion, not an accounting curiosity. Wind then supplied its own evidence that scale does not guarantee returns.
The current success therefore contains both management execution and cycle. Management has improved pricing, cost discipline and underwriting. Power margin has moved from 12.5% in 2024 to 14.7% in 2025 and 18.8% in Q2 2026. Electrification's margin moved from 9.0% to 14.9% and then 18.4%. Those changes are too large to attribute solely to electricity demand.
The cycle nevertheless matters enormously. Utilities and data-centre developers now want equipment faster than the industry can manufacture it. That gives OEMs better pricing and customer-financed working capital. The same Alstom capacity that looked excessive in 2018 is useful when gas slots extend into the next decade. A rational long-term valuation has to assume that some of today's scarcity rent eventually competes away.
Horizontally, GE Vernova's strongest advantage over Siemens Energy and MHI is the combination of a huge service-installed base, a long gas order queue and rapidly improving grid economics. Siemens Energy is close enough operationally to prevent the story from becoming a monopoly thesis; its record orders and high Grid Technologies margins show that the cycle is benefiting multiple suppliers. MHI's GTCC orders make the same point.
In Electrification, ABB and Eaton are cleaner companies. GE Vernova's advantage is project scale, transmission engineering and the generation-grid connection; the disadvantage is greater project complexity and Wind contamination. Prolec closes part of the transformer-capacity gap and, if its projected mid-to-high-20% margin survives consolidation, raises the quality of the whole segment.
Wind is the structural weakness until proved otherwise. A business whose orders are falling about 40% while EBITDA margin is negative 13.6% cannot be described as merely “temporarily diluted” by a bad quarter. Management has a credible 2028 recovery target, but the burden of proof sits with future results. My base valuation assigns it only a few billion dollars.
The market's most important possible misjudgment is free cash flow. The bullish interpretation sees 12 billion USD of 2026 FCF and a 251 billion USD market cap, producing a roughly 4.8% yield before substantial future growth. The accounting bridge shows that a large fraction of that cash is advance-funded. This is good financing and poor evidence of a permanent 12 billion USD run-rate. Once orders stop accelerating, the stock will have to stand on EBITDA, after-tax owner earnings and service cash generation rather than repeated contract-liability inflows.
The market's other possible error goes in the opposite direction. Bears can underestimate how valuable long-duration customer financing is. A business that collects large amounts of cash before building equipment needs less external capital and can earn very high returns on invested shareholder capital. The advances do not deserve to be called recurring FCF, but they materially improve the economics of funding a multi-year expansion.
The one-year variables are therefore commercial conversion and cash quality: 125 GW under contract, SRA-to-order conversion, Power and Electrification margins, Wind losses and the relationship between FCF and contract liabilities.
The three-year variables are different: whether Power and Electrification actually reach around 22% margins, whether the combined company approaches 56 billion USD of revenue and 20% EBITDA margin, whether Wind reaches positive mid-single-digit economics, and how much capex was required to get there.
At five years, the central variable is durability. The investment works extraordinarily well if gas equipment remains supply-constrained into the early 2030s, every turbine placed today produces decades of profitable service revenue, and grid spending remains structurally elevated. The investment becomes ordinary if capacity catches demand and both equipment businesses revert to mid-teens margins. The stock's current price leaves limited room for the latter outcome.
The prior in-house view got two things right. It correctly recognized GE Vernova as a high-quality collection of critical power assets, and its 2.3–2.7% FCF-yield arithmetic was consistent with the 6.5–7.5 billion USD guidance then in force. It was also right to insist that paying above its own bull valuation meant there was no margin of safety.
I diverge materially on value. The prior bear/base/bull ladder of 170–250 / 430–560 / 850–980 USD was built on a much lower operating and cash base. Since then, Power and Electrification have produced near-19% quarterly margins, gas contracted capacity has reached 116 GW, Electrification backlog has expanded sharply, Prolec has been consolidated, and the 2028 company outlook calls for 56 billion USD of revenue at 20% EBITDA margin. Those facts justify a materially higher valuation floor and base case.
I do not, however, carry the full FCF guidance increase into fair value. H1 working capital makes that indefensible. This is why my valuation rises substantially versus the prior report without rising enough to make 942 USD attractive.
Bull reasons, each traceable to the body:
- Gas Power's 53 GW of firm equipment backlog plus 63 GW of SRAs represents almost six years of current 20 GW annual output if reservations convert, while management is targeting at least 125 GW under contract by year-end.
- Power and Electrification Q2 margins were already 18.8% and 18.4%, respectively, leaving a much smaller execution gap to the 22% 2028 segment targets than existed one year ago.
- Electrification equipment backlog reached about 40.6 billion USD, while external IEA work says annual grid investment needs to rise about 50% by 2030.
- Prolec brings a management-forecast 4.2 billion USD of 2028 revenue and roughly 27% EBITDA margin into the fastest-growing segment.
- Power's 72 billion USD-plus service RPO supplies a long-duration aftermarket beneath the equipment cycle.
Bear reasons:
- Roughly 11.7 billion USD of H1 2026 operating cash came from working capital, dominated by customer advances, making the 11.5–12.5 billion USD FY FCF guide a poor steady-state owner-earnings measure.
- Wind lost 275 million USD in Q2 at a negative 13.6% margin while orders fell around 40%, and the 2028 recovery to a positive 6% margin remains unproved.
- At the current price, GE Vernova is around 21.5 times management's 2028 EBITDA target before discounting the target back to today; earning a 9% return while merely hitting that target requires a roughly 26-times 2028 exit multiple.
- Cutting my base terminal segment multiples to 70% takes SOTP present value from about 716 USD toward 513 USD even if revenue and operating margins are unchanged.
- Expanding gas production to 24 GW and potentially 30 GW creates under-utilization risk if today's turbine scarcity normalizes after the capacity has been installed.
The pre-mortem has two credible scripts.
In the first, 2027 data-centre connection delays and utility permitting constraints slow the conversion of GE Vernova's 2026 SRAs. Siemens Energy and Mitsubishi Heavy Industries continue adding gas capacity, reducing the scarcity premium. GE Vernova reaches only roughly 28 billion USD of 2028 Power revenue and an 18% margin rather than my 31 billion USD / 22% base case. Electrification reaches 16–17 billion USD at 18%; Wind remains around break-even or negative. The market values the two profitable units at 14–16 times EBITDA instead of 18–20 times. Equity value falls toward the 400–500 USD area even though the company remains profitable.
In the second, GE Vernova largely hits revenue targets but the cash-flow narrative breaks. Contract liabilities stop rising in 2027 as the order book moves from reservation to delivery, inventories and receivables absorb cash, and reported FCF falls toward 5–7 billion USD. Wind incurs another round of offshore losses while the market stops valuing advance collections as recurring cash generation. An industrial multiple reset occurs at the same time. A share-price decline of roughly 50% would then require no insolvency and no collapse in electricity demand; it would come from cash normalization plus multiple compression.
The final judgment follows from that asymmetry. GE Vernova is no longer the under-earning collection of troubled GE energy assets that investors first received in April 2024. Power and Electrification are now demonstrably high-margin growth franchises operating into markets with genuine supply constraints. The balance sheet is strong, the service backlog is valuable, management has improved industrial execution, and Prolec strengthens the best secular part of the portfolio.
The price asks investors to pay for most of that transformation before it is complete. At 941.95 USD, my base-case economics do not produce an adequate return. The optimistic case can justify the quotation and more, but it requires sustained 20%-plus segment margins and premium terminal multiples. The headline 4.8% 2026 FCF yield softens the prior report's objection; the working-capital bridge prevents it from overturning the valuation discipline.
The refresh changes the valuation ladder much more than it changes the investment conclusion: GE Vernova is a better and more valuable business than the prior report modeled, but the current price still pre-spends too much of the 2028 outcome.
【Company-profile scores】
- Fundamental quality: high
- Growth: high
- Moat: strong
- Financial soundness: strong
- Management credibility: high
- Valuation attractiveness: low
- Risk level: high
- Suitable investor type: long-term growth
【Investment rating】
- Rating: Watch
- One-line thesis: Power and Electrification are compounding into long backlogs, but 942 USD already requires near-bull economics while 2026 FCF is advance-funded.
- Current-price classification: outside the three bands; above the acceptable-hold zone but below the clearly-overvalued threshold.
- Whether to wait for a better price: yes. I would require 320 USD or below with the Power/Electrification thesis intact for a full margin-of-safety entry; the opportunity cost is missing a continued scarcity-driven rerating if 2028 economics exceed even my optimistic assumptions.
- Target holding horizon: 3–5 years for any future entry.
- Expected annualized return: approximately -24% conservative, -3% base and +13% optimistic through the 2028 SOTP horizon, before small dividends; longer holding periods improve the result only if terminal cash growth persists.
- Max-loss risk: roughly 50–65% from the current quotation in a scenario where SRA conversion slows, Power/Electrification margins settle in the mid-to-high teens and terminal industrial multiples compress.
- Reassessment-trigger signals: firm gas backlog/SRA total falls sequentially; SRA conversion remains below roughly 5 GW per quarter for two quarters; Power margin falls below 16% for two quarters; Electrification margin falls below 17% for two quarters; Wind posts another quarter worse than a roughly 250 million USD EBITDA loss without an order recovery.
【Ideal Buy Price】280–320 USD
Basis: roughly 20% or more below the conservative blended SOTP/normalized-cash valuation, providing protection against both execution shortfall and a lower terminal industrial multiple.
The acceptable hold zone is 590–780 USD, surrounding the base-case valuation envelope. I define clearly overvalued as approximately 1,190 USD or above, about 10% above the upper end of my optimistic fair-value work. The current 941.95 USD price therefore sits in the uncomfortable gap where a strong bull case can work but the expected base return is inadequate.
【Valuation Range】
- current: 941.95 (close as of 2026-09-04)
- bear (conservative · ideal buy zone): [280, 320]
- base (fair · acceptable hold zone): [590, 780]
- bull (optimistic · above the clearly-overvalued line): [1,190, 1,300]
The comparison with the prior report is now explicit. The rating remains Watch, but for a different valuation structure. I reject the prior structural conclusion that the market price sits above even a defensible bull case. Today's 941.95 USD is below my clearly-overvalued line and close to the economics of an optimistic outcome. What has changed is Power/Electrification profitability, gas and grid backlog, Prolec and the lower share price. What has not changed is the requirement for a margin of safety. The cash-flow raise looks enormous in the press release; after stripping out the order-advance cycle, it does not make the current quotation a base-case bargain.
Other tickers mentioned
- GE.US: former parent whose power, renewables and grid assets were distributed as GE Vernova in the April 2024 spin.
- ENR.XETRA: Siemens Energy is the closest gas-turbine-plus-grid structural peer and has its own wind-recovery business.
- 7011.TSE: Mitsubishi Heavy Industries is the third major global large-frame gas-turbine OEM.
- VWS.CO: Vestas provides the cleanest public benchmark for standalone wind-turbine economics.
- ABBN.SW: ABB is a high-margin electrification benchmark for grid and electrical-equipment economics.
- ETN.US: Eaton illustrates the premium public-market valuation awarded to clean electrical-equipment exposure.
- VRT.US: Vertiv illustrates how aggressively investors currently value data-centre power infrastructure.
- CEG.US: Constellation Energy is a demand-side beneficiary and counterparty class for rising firm-power requirements.
- VST.US: Vistra represents the generation-owner side of the US load-growth and capacity-investment cycle.
This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.