GEV Q3 2026: Big Backlog, Early Cash

The 60-second read

Five things to know

1. Operations improved. Power margin reached 21.4%; this is current profit, not merely orders.
2. Cash arrives early. The $19.511B contract-liability inflow funds obligations still to be fulfilled.
3. Growth has two sources. Electrification grew 51% reported and 16% organically; Prolec explains much of the gap.
4. Wind is not fixed. A profitable quarter coexists with a $585M year-to-date segment loss.
5. Value needs discipline. Three routes yield roughly $774–$808; the $800 anchor excludes incremental Trench economics.

Three scenarios, one probability-weighted number

Twelve-month targets for late 2027; 2028 terminal values discounted one year at 10%. Distinct from current $800 fair value.

Signposts: what would change our mind

Company guidance is labeled; entry threshold and qualitative cash test are analyst rules. Future earnings date remains unconfirmed.

Price, expectation, and the decision

GE Vernova delivered a better operating quarter while the market lowered the price it was willing to pay. Our reference is the November 9, 2026 regular-session close of $829.34, verified against both the saved FMP quote and its daily close history. This is not a live quote. The October 28 earnings-day close was $878.03 versus $895.08 on October 27, a computed -1.90% decline. From that earnings-day close to our reference the stock declined another -5.55%. These are close-to-close calculations; vendor fields labeled change can describe a different interval and are not used here.

Our call is HOLD, conviction 3 out of 5, with high uncertainty and an $800 fair-value anchor. The decline has made the valuation less stretched, but the share price remains above our central value and well above our $640 accumulation threshold, which requires a 20% margin of safety. That threshold is a decision rule, not a prediction that the shares must trade there. An investor already holding the stock faces a different practical decision from an investor seeking an attractive entry. The holding case rests on backlog conversion, pricing and the service franchise; the case against adding rests on execution risk and cash that arrives before the work is done.

The market capitalization in the quote is approximately $219.5 billion. Our valuation uses 268 million diluted shares, rather than the 264.677 million period-end shares, to retain a modest dilution cushion without assuming future buybacks. At the reference price that creates a $222.3 billion modeled equity value. Neither share count is a forecast of the count after the announced repurchases. Keeping the denominator visible prevents a small apparent precision gain from hiding a material assumption.

The print: growth with a real margin gain

Revenue reached $12.642 billion, up 26.8% from $9.969 billion. Organic growth was 18%, a useful distinction because Prolec GE is now consolidated. Equipment revenue of $8.340 billion rose much faster than services revenue of $4.302 billion. The larger equipment contribution changes both profit timing and the cash cycle: customers may fund capacity slots well before shipments, while installation, warranties and service obligations remain ahead.

Gross profit was $3.107 billion and operating income was $1.394 billion. The computed operating margin was 11.0%, well above the prior-year quarter. Adjusted EBITDA of $1.943 billion and its 15.4% margin are non-GAAP measures, not substitutes for operating profit. The organic adjusted margin was 14.7%, up 690 basis points. This supports a real operating improvement after adjusting for acquisitions and currency, although company adjustments still need scrutiny.

The saved FMP earnings calendar carries a revenue estimate of $11.993 billion, putting reported revenue about 5.4% above that vendor estimate. Its $4.03 EPS estimate sits below the reported $4.81 GAAP EPS, but the estimate's accounting basis is not established. We therefore do not advertise a comparable adjusted EPS beat. The best evidence of the quarter is the internally consistent sequence of more equipment revenue, stronger Power pricing, higher margins and raised guidance. The weakest evidence would be a headline surprise that mixes different definitions.

Power, Electrification, and Wind

Power earned $1.297 billion of segment EBITDA on $6.067 billion of revenue, a 21.4% margin. Orders reached $21.780 billion. Favorable price and higher gas-equipment volume outweighed inflation. This is the core operating evidence for the thesis: scarce production capacity is producing better economics in current deliveries, not just a larger order tally. Services orders grew 12% organically, giving the installed-base argument another independent support.

Gas equipment commitments require careful vocabulary. The combined total is 122 gigawatts, comprising 64 gigawatts of equipment backlog and 58 gigawatts of slot reservation agreements. The quarter added 11 gigawatts of new contracts, converted 13 gigawatts of reservations into orders, and shipped 5 gigawatts. Conversion changes the mix but does not add to the combined total. It would be incorrect to count the conversion as new demand on top of the new contracts. Management expects at least 130 gigawatts under contract at year-end; that is a forward target, not achieved backlog.

Electrification revenue rose 51% to $3.862 billion, but organic growth was 16%. Orders increased on a reported basis and fell 9% organically against large prior-year synchronous-condenser orders. Its approximately 1.4 book-to-bill and $42.2 billion equipment backlog remain constructive. Prolec contributes roughly $5 billion of that equipment backlog. Segment EBITDA reached $739 million, with a 19.1% margin. Acquired growth can be valuable without being organic growth, and investors should demand both labels.

Wind moved to a $72 million quarterly segment EBITDA profit, helped by lower offshore project costs, tariff recoveries and services. That does not erase a $585 million year-to-date loss. Full-year guidance now calls for a $400–$500 million loss, compared with approximately $400 million previously. The revenue outlook improved while the loss range worsened. Vineyard Wind claims have been commercially resolved, but remaining commissioning and offshore execution risks persist. A single profitable quarter is evidence of progress, not proof of a durable turnaround.

Profit bridge and acquisition accounting

The quarter's GAAP operating income of $1.394 billion becomes $1.782 billion of pretax income after $71 million of net financial income, $212 million of non-operating benefit income and $105 million of other income. A $493 million tax expense then leaves $1.289 billion of net income. Noncontrolling interests add $2 million to the amount attributable to GE Vernova because those interests bore a net loss. This is why the consolidated net-income figure and the EPS numerator differ.

The company's EBITDA reconciliation starts with consolidated net income and makes explicit additions and deductions. Depreciation and amortization adds $439 million; restructuring adds $6 million; separation costs add $31 million; business-interest gains subtract $36 million; non-operating benefit income subtracts $212 million; net financial income subtracts $71 million; and the reconciliation's tax adjustment adds $496 million. The last number differs from the income statement's $493 million because the company treats Financial Services tax within its operating measure. Rounded lines total $1.942 billion versus reported $1.943 billion; this disclosed rounding is not silently repaired.

Year-to-date GAAP net income of $6.687 billion is particularly misleading as a run-rate valuation input. The acquisition of the remaining Prolec stake caused a $3.992 billion remeasurement gain on the previously owned interest. The gain is accounting value recognition, not recurring customer profit. Prolec generated $944 million of quarterly revenue but a $5 million pretax loss after acquisition accounting and integration effects. The adjusted segment contribution and acquired GAAP profit are therefore different questions. We use neither the remeasurement gain nor the year-to-date effective tax rate as a permanent earnings engine.

Earnings quality: the cash is early

Nine-month operating cash flow was $16.737 billion and free cash flow was $15.307 billion. The net cash-flow increase in contract liabilities and deferred income was $19.511 billion; this is a net balance-change contribution to cash flow, not a gross customer-payment total. Across the five disclosed operating working-capital lines, the inflow was $16.969 billion. This is customer funding of future obligations, not evidence that the cash is fictitious. It can reduce financing needs and support capacity investment. It also means the current free-cash-flow figure cannot be capitalized indefinitely without modeling the deliveries and costs attached to it.

A simple stress calculation removes all of the year-to-date working-capital inflow from reported free cash flow and produces negative $1.662 billion. Removing only the contract-liability contribution produces negative $4.204 billion. Neither is our estimate of normalized annual cash generation: taxes, voluntary pension payments, growth inventory and timing all interact. These calculations isolate dependency on advances; they do not prove that the underlying business is cash negative. Our valuation instead uses forward cash-flow assumptions stated separately and a reserve against the balance-sheet surplus.

Receivable and inventory days are approximate balance-to-flow diagnostics, not company-reported operating KPIs. Using quarter-end current receivables of $11.587 billion divided by quarter revenue times 92 days gives about 84 days. Inventory of $13.045 billion divided by cost of revenue times 92 gives about 126 days. Acquisitions, supplier advances, deferred inventory and long-cycle equipment contracts reduce comparability. We do not interpret a mechanical days ratio as overdue customer balances.

The ending contract-liability and deferred-income balance was $45.747 billion, of which $33.167 billion belonged to the Power segment. This is a segment balance concentration, not a single-customer concentration or an allocation of the $19.511 billion cash-flow change.

A missing disclosure is not a zero. FMP returns zero for current stock compensation, but the reviewed current documents do not isolate a comparable quarterly SBC expense. The scorecard marks that check unavailable. Annual SEC facts show allocated SBC expense of $257 million in 2025; that cannot be substituted for the current quarter. Likewise, the company has not supplied a clean GAAP forward reconciliation for all guidance measures. High-quality analysis preserves those boundaries.

Balance sheet and capital claims

The headline cash balance includes roughly $400 million of restricted cash. We start with $16.764 billion, remove that restricted amount, then deduct $2.826 billion of borrowings and finance leases, including the current finance-lease portion. The resulting approximate net cash is $13.538 billion. We also deduct $1.156 billion of noncontrolling interests as a book-value proxy because consolidated operating cash flows include economics not wholly owned by common shareholders. Operating lease costs remain within operating cash flow; their $852 million liability is not separately deducted again in our chosen convention.

That leaves $12.382 billion before an analyst execution and capital-deployment reserve of $7.382 billion. We credit only $5 billion of surplus cash in the valuation. The reserve is a deliberately visible judgment, not a reported liability or a forensic estimate of advance-payment runoff. It recognizes customer-funded obligations, factory expansion and acquisition commitments. Releasing the entire reserve would add about $27.54 per diluted share, so cash optimism alone cannot bridge the gap to the highest Street targets.

Trench is a pending transaction, not a completed earnings contribution. The agreed price is approximately €8 billion, described in the release as $9 billion using a 1.12 euro-dollar rate. Management expects roughly equal cash and debt funding and a late-2027 close subject to regulatory approvals. A cash-and-debt split changes liquidity and leverage; it does not make half the purchase free. Our standalone model excludes both Trench operating benefits and an explicit acquired enterprise value. The reserve is not a precise pro-forma acquisition adjustment. Once financing and operating details are firm, the model needs a full sources-and-uses and acquired-cash-flow update.

The board increased the repurchase authorization to $20 billion, with $10.1 billion remaining as of the release. Authorization is permission, not an obligation or guaranteed reduction in shares. The company spent approximately $1.6 billion repurchasing 1.7 million shares in the quarter, and another $1.3 billion on 1.4 million shares through October 23. Buying shares can create value below intrinsic value and destroy it above intrinsic value. A large authorization is therefore not an independent reason to raise our target.

Three routes and what the price assumes

Our published $800 fair value is a rounded central anchor supported by three explicit routes, not a claim that intrinsic value is measurable to the dollar. All routes use 268 million diluted shares and $5 billion of credited surplus. The models are standalone GE Vernova estimates excluding incremental Trench economics. They use analyst assumptions, not management's long-range guidance. High uncertainty calls for a substantial entry discount; our $640 threshold is 80% of the central value.

Route one discounts unlevered cash flows of $6.0, $8.0, $10.0, $12.5 and $15.0 billion for 2027 through 2031. The discount rate is 9.5% and terminal growth is 3.5%. Explicit-period present value is $37.99 billion and terminal present value is $164.37 billion; adding $5 billion and dividing by 268 million gives $773.72. The terminal component supplies most of enterprise value, which makes this route sensitive to both discount rate and durable cash conversion. Cash flows represent post-tax operating cash after reinvestment, with no repeat of the extraordinary advance-payment windfall assumed. They are forecasts, not a subtraction from this quarter's reported cash flow.

Route two uses $12 billion of 2028 adjusted EBITDA at a 21-times enterprise multiple, discounted two annual periods at 10%. That produces $795.76 per share after the same surplus-cash bridge. The two-year convention is an explicit rounded model horizon, not a precise day-count target. This route requires substantial growth from the $6.58 billion implied midpoint of 2026 EBITDA guidance. A high multiple rewards scarcity and services, while discounting prevents future earnings from being treated as already earned.

Route three assigns 2028 Power EBITDA of $8 billion a 22-times multiple and Electrification EBITDA of $4 billion a 23-times multiple. Wind receives no positive value; $800 million of assumed corporate cost is capitalized at 15 times and deducted. The resulting $256 billion future enterprise value, discounted at 10% for two periods, yields $808.10 per share after the bridge. These are underwriting choices, not observed peer averages. The close agreement among routes is useful, but shared assumptions mean they are not independent proofs.

At the reference price, the diluted-share equity value less credited surplus is about $217.3 billion. Holding our cash-flow shape, discount rate and terminal growth fixed requires roughly $16.1 billion of 2031 free cash flow, versus our $15 billion case. That reverse calculation scales the entire path, not merely its terminal year. The stock is not pricing an impossible future, but it offers little room if capacity ramps or durable conversion disappoint. Our scenario distribution is 25% bear, 50% base and 25% bull, producing a weighted twelve-month target of $902.50. These scenarios target late 2027: the 2028 terminal enterprise values are discounted one annual period at 10%, rather than the two periods used for current value. The $5 billion surplus and 268 million shares are held constant at the target date; interim dividends are excluded from this price-only target. The rounded bear, base and bull targets are $570, $875 and $1,290. This future target is distinct from the current $800 fair-value anchor and its $520–$1,175 current-value uncertainty band. The distribution is analytical, not a calibrated probability forecast.

Wall Street: fresh marks and mixed vintages

The saved FMP consensus target is $1,240.56, with a median of $1,250 and a range of $949 to $1,450. The ratings snapshot contains 23 buys, 6 holds and 1 sell. These vendor snapshots do not establish that every constituent is updated for this quarter, nor do they disclose the full model behind each recommendation. We report them as context and avoid presenting the mean as a primary-source valuation.

The retained target-news records do not specify the external analysts’ target horizons. We therefore label those horizons unverified; only our $875 late-2027 base target is explicitly a twelve-month target. This comparison shows dated price levels rather than claiming identical forecast periods. Named October 29 target reports give a more timely comparison: BMO at $1,200, RBC at $1,230, Jefferies at $1,100 and Susquehanna at $1,250. BMO, Jefferies and Susquehanna lowered targets in those records; RBC raised its target modestly. Earlier entries are separately dated, including Evercore at $1,350 on October 1 and Mizuho at $949 in July. Fresh targets can still be optimistic, and stale targets can still be correct; timing matters because it shows which information the analyst could have incorporated.

Our disagreement is concentrated in cash conversion, long-range earnings and the premium investors should pay. FMP's annual estimate snapshot places 2028 revenue near $62.0 billion and EBITDA near $12.54 billion. The vendor accounting basis and acquisition inclusion are not fully documented, so the figure is a directional cross-check, not an exact comparison to our $12 billion adjusted EBITDA assumption. The stronger Street targets can be reconciled with a larger cash-flow base or a higher multiple; neither follows automatically from the backlog total. We want to see deliveries and margin realization validate the optimism.

Management: delivery over promises

The raised full-year revenue range is $46.5–$47.5 billion, adjusted EBITDA margin 13.5–14.5%, and free cash flow $16–$17 billion. At the midpoints, revenue and margin imply $6.58 billion of adjusted EBITDA. That is a calculation, not a separately issued profit target. The earlier revenue midpoint was $46 billion and margin midpoint 13%, implying $5.98 billion. The midpoint improvement is meaningful even after acknowledging the strong quarter already reported.

Power's full-year organic growth guide moved to 19–21%, with segment EBITDA margin 19–21%. Electrification's $14.5–$15.0 billion revenue guide includes approximately $3.1 billion from Prolec; its 18–20% margin is now trending toward the upper end. Wind's revenue outlook improved to a mid-single-digit organic decline, but its loss guidance deteriorated at the high end. A management scorecard should preserve this mixed evidence instead of assigning one uniformly favorable label.

Our questions for management are specific. How much of the cash advance pool is tied to slots rather than firm equipment orders? What proportion of the higher-price backlog converts within the next two years? How much factory investment is required before the next margin step? What contractual protections remain if customers delay projects? And what return on invested capital supports the Trench price after acquisition accounting and financing? These questions aim at the conversion from commercial demand to common-shareholder value, where the forecast is most vulnerable.

Risk ranking and the strongest counterarguments

The bull case is stronger than a generic artificial-intelligence spending narrative. GE Vernova has actual equipment orders, a large service base, customer payments and expanding current margins. Power's performance indicates that scarcity is monetizable. If production expands without sacrificing quality and new installed equipment creates durable service work, our forward cash flows could be conservative. The bull scenario allows $15 billion of 2028 EBITDA and a 25-times multiple, yielding a twelve-month late-2027 target of approximately $1,290 per share after one year of discounting and the cash bridge; the same terminal assumptions discounted two years produce approximately $1,175 of current value.

The bear case does not require electricity demand to collapse. Delivery bottlenecks, defects, unfavorable contractual estimates or slower customer projects can defer profit while cash already collected must fund execution. A smaller order inflow would reduce the financing benefit before the large backlog is exhausted. Wind can consume resources even while the other segments grow, and acquisitions can create amortization, debt and integration costs that make adjusted growth look cleaner than common-shareholder returns.

Legacy separation obligations add another risk. Note 22 describes $1.008 billion of indemnification commitments and a $671 million recorded liability, including restricted funds associated with legacy GE matters. An IRS audit of historical GE returns could create additional obligations allocated to GE Vernova that are not reasonably estimable and have no accrual. We do not turn an unquantified risk into a made-up liability; it belongs in the uncertainty assessment.

Our central case accepts the operating opportunity and refuses to treat deposits as perpetual owner earnings. The greatest model risk is that the same long-range optimism appears in EBITDA growth, terminal cash flow and the exit multiple simultaneously. That creates correlated downside. We therefore keep a broad $520–$1,175 current-value uncertainty band and high uncertainty. The separately dated twelve-month targets span $570–$1,290. The range is a set of conditional outcomes, not a claim that losses outside it are impossible.

Dated checkpoints

The next useful catalyst is evidence that the 2026 guide converts into results, not merely another order announcement. A year-end gas commitment total of at least 130 gigawatts and backlog above $200 billion would validate management's commercial targets. Margin delivery and cash composition then determine whether that demand is becoming durable earnings. The company has not confirmed the next earnings date within this source packet; the vendor calendar's January 27, 2027 date is tentative and must be checked before publication-time scheduling.

Trench's expected late-2027 closing is a separate catalyst with regulatory and funding conditions. It should not be pulled into the current quarter or treated as guaranteed. Prolec's purchase-price allocation remains preliminary, with completion expected no later than the first quarter of 2027. That can alter accounting allocations without changing the commercial demand thesis. Investors should distinguish accounting finalization, financing news and operating performance because they answer different questions.

Data, method, and source limitations

All amounts are US dollars unless indicated. Tables in millions use $M; share prices use ordinary dollars. Current-period issuer filings control whenever a vendor mapping or rounded arithmetic differs. Historical rows are FMP reported data and retain the provider's definitions; they are not presented as a fully restated segment time series. A business-unit reporting realignment became effective January 1, 2026, so the current filing's prior-year segment comparison is the reliable basis for the segment charts here.

The nine-quarter table supplies revenue, operating profit, attributable income, diluted EPS and cash flow. Full provider income, balance-sheet and cash-flow records are retained in the source-bound dataset. Historical detailed segment profits, quarterly non-GAAP reconciliations, employee counts and like-for-like quarter SBC were not all separately verified. They are unavailable for this analysis and are not filled with estimates. Three fiscal years of cash flow show the pre-advance scale of the business, but 2023 and part of 2024 precede the separation and are not clean standalone comparisons.

GE Vernova is a US common-stock issuer, not an ADS structure; no ADS conversion is used. The filing identifies a Delaware company and NYSE listing. The quarter's disclosure controls were evaluated as effective. Prolec's integration controls remain in progress and its operations are excluded from the 2026 annual control assessment under the disclosed SEC guidance. This is not described as a finding of ineffective controls. The current Trench agreement is an ordinary transaction risk, not a change in the covered listed security. The 2025 annual report identifies Deloitte & Touche LLP as auditor and approximately 75,000 employees at that year-end. Those are dated annual facts, not current headcount. The September 2026 filing reports $138 million of investments in unconsolidated variable-interest entities and $130 million of investment commitments; these are investment exposures, not an ADS ownership structure. Voting-control details, customer concentration percentages and historical employee series were not independently established for this packet.

Method glossary: RPO means remaining performance obligations, a contracted-revenue measure with timing and modification risk. Slot reservation agreements reserve production capacity and are distinct from equipment backlog. EBITDA excludes depreciation and amortization; adjusted EBITDA additionally follows company-specific exclusions. FCFF is our modeled cash available to all capital providers after operating taxes and reinvestment. Enterprise value is the value of operating claims before the equity bridge. Our price targets are conditional estimates, not investment guarantees.

Source discipline also governs negative findings. The vendor's zero SBC is rejected; its current-assets and receivables mapping is not used for issuer balance-sheet ratios; its EBITDA is not labeled company adjusted EBITDA; and its EPS estimate is not claimed to be comparable adjusted earnings. Historical print reactions require verified release timing plus prior-close and next-close data. Only the current before-open release reaction is used here; unavailable earlier reaction studies are not synthesized. The previous Q2 episode's $800 anchor is retained as historical context, but its old source assumptions are not recycled into this quarter. There was no prior research-packet signpost table available to grade.