Built by Grok 4.6 High

Reports Wednesday, October 28, 2026. Prepared Monday, October 26, 2026. This is a preview of the unreported third quarter. GE Vernova reports before the open, with the webcast at 7:30 a.m. ET. CEO Scott Strazik and CFO Ken Parks are both still on it. Two calendar days remain. The last reported quarter is Q2 2026, released before the open on July 22, 2026. The fiscal year ends December 31, so 2026Q3 is both the fiscal and the calendar quarter. Daloopa’s latest reported quarter is still 2026Q2. The homepage calendar stores October 28 as 2026Q4. That label is one quarter ahead of the quarter being reported. This page is 2026Q3.

GEV | Earnings Preview

GE Vernova is the gas-turbine and grid franchise in a power market that is still taking equipment years ahead of the factory. That leadership is not what Wednesday decides. Wednesday decides whether a guide that was raised on July 22, and then left alone at Morgan Stanley Laguna on September 16, is still the right bar. There is no consolidated Q3 revenue or EPS guide. The checkable lines are Power organic growth of 17–19% at a 17–18% EBITDA margin, Electrification revenue of $3.8 billion–$4.0 billion, Wind revenue down a low-double-digit rate with EBITDA about breakeven, and free cash flow that is only required to be positive. Q4 is the quarter management has already called the high quarter for revenue and EBITDA.

Growth is still accelerating on the reported line, and the organic line is slower. Revenue growth went from +3.8% in Q4 2025, on $10,956 million versus $10,559 million, to +16.3% in Q1 2026 and +21.9% in Q2 2026, $11,104 million versus $9,111 million. Company organic growth in Q2 was only 12%. The gap is Prolec, which closed February 2, 2026 and is not in the 2025 base. Power organic growth is the clean acceleration: +5% in Q4 2025, +10% in Q1 2026, +14% in Q2 2026, and a Q3 guide of 17%–19%. Electrification organic growth has not accelerated. It was 29% in Q2, down from 32% in Q3 2025. A Q3 Electrification print inside $3.8–$4.0 billion is still about +46% to +54% versus reported Q3 2025 revenue of $2,601 million, and that rate is the acquisition, not a faster organic franchise. Adjusted EBITDA margin has expanded year over year in every comparable quarter in the window and reached 11.3% in Q2, +280 bps versus 8.5%. The full-year band is still 12%–14%. They raised revenue and cash in July and refused to tighten the margin.

The street is underwriting the July guide, not another order acceleration. FMP, pulled October 26, is the only consensus on this page. Visible Alpha and Bloomberg are not connected. Q3 revenue of $11.993 billion is +20.3% versus $9,969 million. EPS of $3.99 is +143% versus $1.64. The earnings-calendar row rounds those to $11.993 billion and $4.01, with both actuals null. Applying the organic Power rate to the organic base of $4,789 million, the Electrification midpoint, a low-double-digit Wind decline on the organic base of $2,619 million, and Q3 2025 eliminations of −$119 million lands near $11.76 billion, about 2% under the street. The high end of that same stack is about $11.94 billion. Matching $11.99 billion needs the high end of Power and Electrification and a Wind decline no worse than about 10%. Applying the organic rates to reported dollars instead, which is what a looser stack does, puts the street near the middle of an $11.7–$12.1 billion range. Either way the street is not below the guide. FY2026 revenue consensus of $46.329 billion sits inside $45.5 billion–$46.5 billion, 0.7% above the $46.0 billion midpoint. FMP’s EBITDA averages are a different definition. Do not grade the 12–14% adjusted-margin band against the $10.34 billion FY2026 EBITDA average.

Orders are the argument, and the street has already marked them down from the Q2 spike. Total orders were $24,216 million in Q2, +95.9% versus $12,364 million. One October 20 sell-side note put the orders consensus at $22.2 billion, about +52% versus Q3 2025 orders of $14,608 million, and its own estimate at $21 billion. That is a deceleration, and it is what Strazik already said in September: Q3 orders still strong because reservations are converting, most of the summer’s incremental 12 GW of 2030–31 supply contracts in Q4, and data-center orders fall from almost 40% of first-half Electrification orders to about 20% in each of Q3 and Q4. FMP does not publish an orders number. This is not a case of management repeating a demand story the street refuses. The street believes the July P&L. The proof management keeps pointing at and will not show on Wednesday — equipment margin in backlog, the 2027 outlook, the 2030 case — was deferred to January and to a capital markets day in spring 2027.

What to watch on October 28

• Power organic growth versus 17–19%, and whether gas shipments actually step to about 5 GW. Q2 shipments were 3.3 GW, −37% versus 5.2 GW. The revenue guide was built on that step. Missing 5 GW does not by itself miss a dollar range that was set in July. It does change whether the 24 GW and 30 GW ladder is a factory fact.
• Power margin versus 17–18%. Q2 was 18.8%. The Q3 range is below that on purpose, because Q3 is the seasonal services trough. A print under 17% is the Power miss that matters. 17.0% is in-guide, not a beat. Versus Q3 2025’s 13.3%, 17–18% is still +370 to +470 bps.
• Electrification revenue versus $3.8–$4.0 billion. This is the only consolidated-scale dollar range, and the prior quarter’s pointer was beaten through the top: Q2 printed $3,637 million against a $3.3–$3.5 billion pointer from the Q1 call. Below $3.8 billion is the first crack in the segment that has funded the revenue raises.
• Wind at about breakeven. First-half Wind EBITDA was −$382 million plus −$275 million, about −$657 million. A full-year loss of about $400 million needs the second half to earn roughly +$257 million. If Q3 is about zero, almost all of that sits in Q4. They missed this same loss figure in 2024 and again in 2025. This is the aggressive line.
• Free cash flow positive, and much smaller than Q2. First-half free cash flow is $4,791 million plus $5,107 million, or $9.9 billion, against an FY range of $11.5 billion–$12.5 billion. A collapse versus Q2 is the guide. Treating it as a miss ignores the only cash sentence they gave for the quarter.
• Tariffs and onshore wind orders, which they have not cleared. The April net tariff figure of $250–$350 million was not restated in July or at Laguna. First-half Wind shipments had fewer contractual protections. Section 232 and permitting are still why they will not call a U.S. onshore order inflection. Silence is not a cut and it is not a clearance. The July guide was built with that impact already inside it.
• Anything that is not in this print. The Venture Global frame was signed October 9, after quarter-end, and is shared with Baker Hughes. It is not Q3 backlog. The India HVDC award of almost $1.5 billion was placed in Q4. Solid-state transformer orders were described as a 2027 event.

Hold, because the disagreement is not on this quarter’s P&L. Revenue has beaten for six straight public quarters since Q1 2025, but the surprise shrank from +9.0% in Q3 2025 to +0.9% and then +2.9%. The two large recent EPS beats are a $2,572 million tax benefit and a $4,494 million Prolec gain. The last clean EPS print, $2.47 versus an FMP estimate of $3.17, missed by 22.1% on the same day management raised the year. Internal models, SharePoint, OneNote, Outlook, and Excel were not available. There is no internal EPS. The bias into Wednesday is a small revenue beat or an in-line print at the high end of the segment color, not another 9-point surprise, and no benefit of the doubt on GAAP EPS.

GE Vernova guides the year, and it gives quarter-ahead segment ranges. It does not guide consolidated quarterly revenue, EPS, gross margin, or operating margin. The July 22 ranges are still the outlook. No 8-K or IR release between July 22 and October 26 changed a financial number. The October 13 release only set the October 28 webcast. Laguna added orders, price, and timing. It did not add a dollar to 2026 revenue, the 12–14% margin, or free cash flow.

Classification: conservative on the lines they have historically beaten, in line to slightly tight versus a street that has already walked up, aggressive on Wind. Power organic growth, Electrification dollars, and free cash flow are the lines they guide low and then beat, usually after a raise. Margin ranges are wide and they land inside them, nearer the middle than the high end. The Wind loss was too small in FY2024 and too small again in FY2025, and the FY2026 guide repeats about $400 million after a −$598 million year. First-half 2026 is already through that number. A 100% beat rate would have made the guide uninformative. It is not that. It is informative on Wind, and the information is that the loss runs hot.

The raises are part of the record, and they have gotten smaller. FY2025 revenue was held at $36 billion–$37 billion all year and beaten. FY2026 revenue has already been raised twice, from the initial $44 billion–$45 billion to the Q1 range of $44.5 billion–$45.5 billion and then to $45.5–$46.5 billion. The street is now inside the new range. The remaining revenue conservatism is smaller than it was in January. Free cash flow was raised by $5 billion at both ends in July, from $6.5 billion–$7.5 billion. The residual second-half bar is no longer a sandbag. The margin band moved from 11%–13% to 12–14% at Q1 and then froze while revenue and cash were raised again. That is the tell. They will talk growth and cash in dollars. They will not sponsor 14%.

FMP’s Q3 EBITDA average of $1.647 billion on $11.993 billion of revenue is a 13.7% margin. That sits in the top half of the full-year band, in a quarter management called a Power services trough, with Q4 supposed to be the EBITDA high. It is also not adjusted EBITDA. Use it only as a sign that the street’s profit stack is the segment math, not as a company target. FY2026 EPS consensus of $30.35 stacks Q1 GAAP EPS of $17.44 under the later quarters. FY2027 EPS consensus of $24.43 is lower for that reason. The company does not guide EPS. Do not score a beat or a miss against $30.35 as if it were run-rate earnings.

Current quarter — Q3 2026

The organic Power base is $4,789 million, not the reported $4,838 million. Do not put 17–19% on the reported dollar and call it the guide. The FX gap has been small — Q2 organic Power was +14% against reported growth of about +15% — but the guide is the organic rate. “Low double digits” on Wind is read here as about −10% to −15%. That reading is an interpretation of the phrase, not a company range. Eliminations are not guided. Q2 2026 eliminations were −$37 million.

The stack that puts the street at the high end, calculated and labeled as such. Power at +18% on the $4,789 million organic base is about $5.65 billion. Electrification at the midpoint is $3.9 billion. Wind at −11% on the $2,619 million organic base is about $2.33 billion. Subtract $119 million of year-ago eliminations and the total is about $11.76 billion. Power at +19%, Electrification at $4.0 billion, and Wind at −10% is about $11.94 billion. FMP’s low estimate is $11.765 billion and the high is $12.270 billion, 12 analysts. The low of the street is the midpoint of management’s color. The average is the high end.

Q3 is also not supposed to be the order peak inside the second half. Gas under contract is guided to at least 125 GW by year-end, from 116 GW at June 30. At Laguna, directionally 20 GW of second-half commitments was called “likely to prove conservative,” and Q4 was called the larger contractual quarter. A strong Q3 orders print can still exclude most of the summer’s incremental 12 GW.

Trajectory — eight reported quarters

Columns are periods. Growth is versus the year-ago quarter. Q3 2026 is not on this table because it has not printed. Adjusted EBITDA is the series the releases print. Q4 2025 and Q1 2026 diluted EPS are not operating earnings.

Revenue growth troughed at +3.8% and has reaccelerated for two quarters. Orders have accelerated for four quarters, from +55% to +96%, which is why a $22.2 billion Q3 orders figure would be a deceleration even if it is still a very large quarter. Margin expansion never stopped, but the year-over-year delta narrowed from +540 bps in Q3 2025 to +40 bps in Q4 2025 and has since widened again. The $13.39 and $17.44 EPS prints are a tax benefit and the Prolec gain. The operating EPS compare in the recent window is Q2 2026 versus Q2 2025, +32.8%. Orders YoY before Q3 2025 is left blank because this table does not carry the 2023 order base. Dollars are millions.

Segment scorecard the print will actually be graded on. Latest YoY is Q2 2026 versus Q2 2025.

FQ+1 — Q4 2026

No numeric Q4 revenue, EPS, margin, or free-cash-flow range was given. The qualitative cadence matches the shape of consensus: Q4 is the high quarter. FMP revenue of $13.901 billion is +26.9% versus Q4 2025 revenue of $10,956 million, and it is $1.91 billion above the Q3 consensus. That is the right direction. It is not a company guide, and it is not a sequential growth rate to underwrite as if management had given one.

H1 2026 revenue of $20.443 billion, the sum of $9,339 million and $11,104 million, leaves $25.06 / $25.56 / $26.06 billion of second-half revenue to hit the FY low, midpoint, and high. FMP’s H2 (Q3 $11.993 billion plus Q4 $13.901 billion) is $25.89 billion, $0.34 billion above the midpoint path and $0.16 billion below the high. The street is inside the year, closer to the top. A Q3 revenue miss versus $11.99 billion puts any further raise on Q4.

FY guide — standing FY2026, and FY2027 which is not guided

“FY+1” in the working notes is the standing FY2026 guide into this print, not a year the company has set for 2027. FY2027 consensus is shown so it is not mistaken for a guide. The 2028 framework from the Q4 2025 communication — revenue $56 billion, adjusted EBITDA margin 20% — was not reaffirmed at Laguna. Strazik said the 2027 outlook comes on the January call.

FY2025 reported revenue used in the YoY rates is $38,068 million, the sum of the four 2025 quarters in the trajectory table, which matches segment revenue of $38,519 million plus eliminations of −$451 million. FY2025 adjusted EBITDA was $3,196 million at an 8.4% margin. First-half 2026 adjusted EBITDA of $2,146 million, $896 million plus $1,250 million, is a 10.5% margin on $20,443 million of revenue. On a $46.0 billion midpoint, the 12% low end needs about a 13.2% second half. The 13% midpoint needs about 15.0%. The 14% high end needs about 16.8%. The low end is the step-up they have pointed at. The high end is a margin level no recent half has printed. Corporate costs stay at $450 million–$500 million and are uneven by quarter, so a single-quarter miss there is noise.

Confidence sits on orders, price, and the gas-capacity ladder. It does not sit on a tighter 2026 margin. At Laguna on September 16, Strazik got more specific on the $200 billion backlog, on price, and on when the incremental gigawatts contract. He did not put a new number on 2026 revenue, adjusted EBITDA margin, or free cash flow. He did say the third-quarter call “will continue to affirm both the growth and margin expansion.” That is a confidence statement. It is not a guidance raise.

The mechanisms under the Q3 ranges are shipment and mix, not a demand surprise.

Power. Output is supposed to step from about three gigawatts a quarter to five in both Q3 and Q4, in the same factories, not a greenfield. About 325 new machines were in the gas factories at the Q2 call, “over 350” by Laguna, and about 400 by year-end. Labor for the 15-to-20 GW step was hired about a year early and was a margin drag while those workers were underutilized. The assumption is that this drag is already inside the 17–18% Q3 margin. Price is the other leg: first-half 2026 gas equipment orders were priced more than 20% above Q4 2025 equipment orders, and second-half orders were expected at the high end of +10 to +20 points per kilowatt versus that Q4 base. The margin page that proves the backlog is January, not October.

Electrification. Q1 revenue of $2,959 million plus Q2 of $3,637 million is $6,596 million. Adding the Q3 band puts nine months at about $10.4–$10.6 billion, which leaves roughly $3.9–$4.6 billion for Q4 to land the $14.5 billion–$15.0 billion year. Q4 does not have to be heroic. Parks said Prolec contributed nearly $900 million of Q2 revenue and that organic growth was 29%. A Q3 print inside the dollar range can be a sequential dollar increase and a slower reported growth rate at the same time. That is the acquisition anniversary math, not evidence the segment rolled over. Data-center orders are the lump. More than $5 billion in the first half will “not likely” repeat in Q3 and Q4.

Wind. About 70% of 2025 onshore equipment orders were booked in the second half of 2025 and are what gets delivered in the second half of 2026. Q3 breakeven assumes offshore project costs come down, with Dogger Bank B the named project, while onshore services margins keep expanding. That is an execution close-out, not a growth engine. U.S. onshore orders are the sentence they have refused to retire since Q1: too early to call an inflection, permitting delays, Section 232. Repower is the part they like, more than 30 GW of U.S. install base, production tax credits expiring at the end of 2030, and they still called it early.

Cash. Q2 free cash flow of $5,107 million was driven primarily by down payments on gas slot reservations and by Electrification orders. Parks was direct that first-half cash will be substantially higher than the second half as reservations convert to orders. At Laguna, Strazik added the quality point: the increase in the progress-payment balance is still smaller than the increase in margin in backlog. That is the defense against “the FCF raise is just customer deposits.” It is also why a modestly positive Q3 cash number can be in-guide and still look like a collapse versus Q2. Firm gas equipment backlog was 53 GW against slot reservations of 63 GW. The second-half claim is that conversions flip that relationship. A quarter that adds reservations faster than it converts them, again, grows the book without improving the quality of it.

Tone, across the calls. Demand-and-price language has stepped up at every communication since January. The financial-guide tone peaked at the July raise and then went quiet. September was more bullish on the order book than July, and no more bullish on 2026 earnings.

The pattern is a management team that raises what the backlog has already funded and will not raise the margin until the equipment delivers. Laguna fits the pattern. It pulled a backlog milestone forward. It did not pull an earnings number forward. Red-flag screen: guidance has not been withdrawn, the date was not delayed, and the CEO and CFO are unchanged. The defensive pocket is Wind orders and the deliberate refusal to tighten the margin band or to claim the first-half cash repeats. That is conservatism on two lines, not a deteriorating tone.

What would change the story on Wednesday is narrower than the speech. Power organic growth inside 17–19% at a 17–18% margin, with shipments at the five-gigawatt rate, means the capacity ladder is real. Electrification inside $3.8–$4.0 billion with margin above the Q2 level is consistent with a segment whose annual guide has been raised twice. Wind about zero only keeps the $400 million year alive. It does not prove it. Free cash flow positive but far below Q2 is in-guide. Any cut to $45.5–$46.5 billion, 12–14%, or $11.5–$12.5 billion would be a tone break they have not prepared. A raise is possible on the historical pattern and was not previewed. They are saving 2027 and the backlog-margin page for January.

Consensus is the July guide written down as two numbers: FMP revenue of $11.993 billion and FMP EBITDA of $1.647 billion. One broker, Bank of America on October 20, is a little above that P&L ($12.4 billion of revenue and $1.8 billion of adjusted EBITDA, as carried in the run file) and a little below the orders consensus it cited ($21 billion versus $22.2 billion). Neither figure is an FMP line. The open points are the ones with no consensus: shipments, the split between firm orders and slot reservations, and whether Wind can actually stop losing money.

Equipment orders are the acceleration. Services orders decelerated in Q2, to +16% ($5,275 million versus $4,555 million) from +26% in Q1 ($5,526 million versus $4,392 million). The backlog story is equipment. The margin story, once the equipment ships, is services. Strazik’s September illustration — about $0.5 billion of services revenue over 20 years on each incremental HA gigawatt, and about $100 billion of contracted future services by the end of 2027 — does not bill in Q3. Q3 only has to show the gigawatts leaving the factory. Electrification equipment-order growth, organic, has already cooled from the Q3 2025 spike of +134% to +72% in Q2. The mix shift he previewed sits on top of an equipment-order rate that was already off its peak.

Peer prints that landed before Wednesday cut the same way and do not change the bar. Grid and transformer demand has been strong: LS Electric’s October 21 quarter showed operating profit more than doubling, and Nexans reaffirmed its 2026 guide on October 22. Onshore wind is split: Nordex raised its EBITDA-margin guide on an October 15 preliminary, and Goldwind’s Q3 revenue and profit fell year over year. Bloom Energy’s October 22 raise is onsite power for data centers, not a turbine read. Vestas and Siemens Energy report in November, after this print. HD Hyundai Electric’s IR call is October 27, and Hubbell reports the same morning. Neither result is in hand. Those peer figures are company releases, not Daloopa, so they are not hyperlinked.

Nothing since July 22 cuts the FY2026 range or replaces the quarterly shape. The news is about the quality and the timing of orders. Laguna says third-quarter orders should still be strong because reservations are converting, while the newest 12 GW of 2030–31 supply and the slower second-half data-center mix land more in Q4. Venture Global is the cleanest illustration of that lag: a large, conditional, dual-sourced frame signed after the quarter closed. Songon is the offsetting proof that at least one international gas turbine did book inside Q3, at a size that cannot carry the gas book. Wind’s legal overhang at Vineyard is settled and the dollars are hidden. Nuclear and the LS Electric venture lengthen the 2030s story without touching this quarter’s revenue.

Left out on purpose: the July 22 print itself, the October 13 webcast notice, the September 30 dividend at the same $0.50, sell-side target changes, and the October price tape. Those are not operating disclosures. The $0.50 dividend is the fourth quarter at that rate. Forward $2.00 a year is immaterial next to the free-cash-flow guide. A July 28 Dinorwig pumped-storage selection was announced in the window and was not established as a Q3 order. A secondary account places it in Q2. It is not used as incremental third-quarter equipment demand.

Tariffs did not get a new number. The $250–$350 million company-wide net figure from the April call was not revised in July or September. The constraint that was repeated is U.S. onshore wind: Section 232 classification, permitting, and tariff uncertainty, the same list in April, July, and September. The impact sits where the contracts lack protection, which management has said is the first-half Wind shipments. That is already inside the July guide.

Mixed. Revenue beat in 7 of 10 scored public quarters and in 4 of 4 of the last four, six in a row since Q1 2025. The surprise got larger into Q3 2025 and has since shrunk, from +9.0% to +7.4% to +0.9% to +2.9%, while revenue growth itself reaccelerated. EPS beat in 4 of 9 scored quarters and in 2 of 4 of the last four. Both of those recent EPS beats sit below adjusted EBITDA: the Q4 2025 tax benefit and the Q1 2026 Prolec gain. Strip those two out and the last clean EPS print missed by 22.1%. Q3 2026 has not printed.

Actuals are Daloopa diluted EPS and revenue. Consensus is FMP. A beat is any positive surprise. Q2 2024 GAAP EPS of $4.65 is a gain on the Steam sale to EDF and is not scored, because FMP’s stored actual of $0.71 is not the filed number. Q3 2024 uses the loss of $(0.35). FMP’s stored +$0.35 has the wrong sign. Q1 2024 is scored on the company’s adjusted EPS of $(0.41), which is what FMP stored, not diluted EPS of $(0.45). There is no ongoing adjusted-EPS series. The company steers adjusted EBITDA.

Read left to right. The revenue row is red only in the spin year, and the reds are inside 1.4%. The dark greens are Q1 2025 through Q4 2025. The last two cells fade. The company is still clearing the revenue number. It is no longer clearing it by 7–9%. The EPS row is not a streak. The dark red is Q3 2024, a GAAP loss against a small profit estimate, driven by Offshore Wind contract losses of about $0.7 billion in the October 23, 2024 release, partly offset by a $0.3 billion settlement gain. Adjusted EBITDA that quarter was still positive at $243 million. The two darkest greens are Q4 2025 and Q1 2026, and the cell after them is red again.

Magnitude improved, then deteriorated, on revenue. It is not a trend on EPS. The first four public quarters averaged about −0.1% on revenue. The next four averaged +6.5%. The last four still average +5.1%, but that average is the +9.0% and the +7.4%. The path inside the last four is +9.0%, +7.4%, +0.9%, +2.9%. The most recent print, $11,104 million versus a $10,789.5 million estimate, is still a beat. It is a smaller beat than the 2025 cluster. That fade is the street catching a business that was accelerating, not a business that slowed. A smaller beat on faster growth is a higher bar.

On the line that is comparable, the EPS beat has not gotten larger. The operating beats in the first half of 2025 were about $0.44 and $0.38. The clean misses since then are eight cents in Q3 2025 and 70 cents in Q2 2026. The −22.1% is a wider miss than Q3 2025’s −4.7%. The line management actually steers moved the other way: adjusted EBITDA dollars reaccelerated after the Q4 2025 trough of +7.3%, to +96.1% and then +62.3%, and the margin expanded year over year in every quarter that has a comparable margin.

The calls do not walk a consensus EPS bridge. The variance they explain is orders, revenue, adjusted EBITDA, and cash. Parks raised full-year revenue and free cash flow on July 22 and held the margin range. He did not take the EPS miss. The $3.17 estimate sat 70 cents above a clean $2.47 print. That is the street still one step ahead of GAAP EPS after two quarters in which GAAP EPS was not the business.

Read-through for Wednesday. Six straight revenue beats do not mean a 9-point beat is still the base rate. The magnitude has already come in. Hitting $11.993 billion would be about +20.3% versus $9,969 million, in the neighborhood of Q2’s +21.9%, not a deceleration in the growth rate and not a return to the +9% surprise. The EPS bar is the less reliable of the two, and it got harder after a miss. The next estimate is $3.99, above the bar that was just missed, against a year-ago $1.64. Part of a step-up from Q2 to Q3 to Q4 is what management has said to expect, with Q4 the high quarter, so $3.99 is not automatically a leftover of the $17.44. It is still a number the company missed the last time the print was clean, by 70 cents, while adjusted EBITDA was beating the year-ago quarter by 62%.

On the annual scoreboard the same split shows up. FY2025 revenue beat a held range. FY2025 free cash flow beat a raised range, $3,710 million versus a high end of $3.5 billion. FY2025 adjusted EBITDA margin of 8.4% landed inside 8–9%, nearer the middle. Power organic growth of 10% beat a 6–7% guide. The Wind loss of −$598 million missed an about $400 million loss guide by roughly half again. FY2024 Wind was the same kind of miss, −$588 million against a loss guide of $200 million–$400 million. Guide low on growth and cash. Land inside a wide margin band. Miss the Wind loss. That is the record this print is walking into, and the first half has already put 2026 on the miss side of the Wind line.