Siemens Energy

Phase A — Understand the business

Company Overview

Siemens Energy is the energy-technology company spun out of Siemens AG in September 2020 — the "hard-iron" half (power generation + grid) that Siemens the industrial-automation group wanted off its balance sheet. It is not a utility and not a pure-play renewables name; it is a capital-goods OEM + long-cycle service annuity across the entire electricity value chain, from the gas turbine that makes the electron to the HVDC link and transformer that moves it. Today it is a sum of four parts, and understanding the company means understanding that those four parts are at wildly different points in their life:

1. Grid Technologies (GT) — the crown jewel. Transformers, high-voltage direct-current (HVDC) links, switchgear, grid stabilization/FACTS, and grid software. This is the pick-and-shovel play on the AI-datacenter electrification supercycle: FY2025 comparable revenue +25.4%, profit margin before special items 15.8% — the highest-growth and highest-margin segment — with FY2025 order intake above €21bn (a record). It sits in a three-way global oligopoly with Hitachi Energy and GE Vernova's Grid Solutions.
2. Gas Services (GS) — the cash cow, unexpectedly re-rated. Heavy-duty and industrial gas turbines, plus the very high-margin installed-base service annuity (LTSAs). Left for dead in the "energy transition" narrative, it has become the near-term way to power a data center now: FY2025 revenue +14.2% comparable, margin 13.0%, 194 gas turbines sold, ~€3.2bn of segment free cash flow (>2x cash conversion).
3. Transformation of Industry (TI) — the swing option. Compression, industrial steam turbines & generators, and electrolyzers (hydrogen). FY2025 revenue +13.5%, margin 11.3% — profitable, but the hydrogen leg is still an early-stage capital question rather than an earnings driver.
4. Siemens Gamesa (SG) — the wound that is healing. Onshore + offshore wind turbines. Nearly killed the company in 2023 (below); FY2025 margin still −13.1% but H1 FY2026 losses collapsed and management guides to break-even in FY2026.

Contract structure & payment terms — the load-bearing fact of the whole thesis. These are large, multi-year, milestone-billed projects (a HVDC link or a turbine island is delivered over 2–4 years). Siemens Energy collects substantial customer advance payments on order — so a record order book is also a record cash inflow ahead of revenue recognition. Contract liabilities rose €2,816m in H1 FY2025 alone. That is why FY2025 free cash flow pre-tax (€4,663m) ran ~2x the previous year and FY2026 FCF is guided to ~€8bn while net income is only ~€4bn — the business is negative-working-capital-financed on the way up. Keep this in view; it is the single most important thing to model (Lens 10 & 11).

Group FY2025 scorecard:

• Revenue €39.1bn (+15.2% comparable)
• Orders €58.9bn (+19.4% comparable); order backlog €138bn at year-end, since risen to a record €154bn at Mar-31-2026 (Q2 FY26, book-to-billBook-to-billNew orders divided by orders filled. Above 1 means the backlog is growing; below 1 means the company is working through it faster than it is replacing it. 1.72)
• Profit before special items €2,355m (6.0% margin) · Net income €1,685m · EPS €1.63 · FCF pre-tax €4,663m

The one-sentence business model: Siemens Energy sells the physical bottleneck of the electricity system — turbines and grids — into a demand shock (AI power) it cannot supply fast enough, and gets paid in advance to do it.

Supply Chain

Map the chain, name the names — a generic supply-chain answer fails this lens.

Upstream inputs → Siemens Energy → end customer:

• Raw materials / critical inputs:
• Grain-oriented electrical steel (GOES) — the core of every transformer. This is the chokepoint. In the US, Cleveland-Cliffs is the single domestic GOES producer (Pennsylvania/Ohio); steelmakers are actively re-allocating melt capacity away from GOES toward non-oriented steel for the EV/auto market's higher margins. Global GOES is concentrated (Nippon Steel, POSCO, Baowu, thyssenkrupp, JFE, AK/Cliffs). Single-source risk is real.
• Copper (windings, cabling), large castings/forgings (turbine rotors, casings), rare-earth magnets (generators), power semiconductors / IGBTs for HVDC converters — semiconductor lead times "up to 40 weeks".
• The company (conversion): GT plants (Germany — Nuremberg/Berlin transformer & HVDC; new US transformer factory, ~$421m, North Carolina, first units ~2026–2027, ramping to 57 large power transformers/yr ); GS turbine plants (Berlin, Charlotte NC, Mülheim); SG blade/nacelle plants (Spain, Denmark, Germany, India).
• Distribution / EPC layer: heavy-lift & project logistics (transformers and turbine components are super-heavy cargo — Siemens Energy runs a "heavy-lift surge" logistics operation ).
• End customers:
• Grid operators / TSOs & utilities (HVDC, transformers) — e.g. a >€1bn Baltic Sea HVDC award.
• Hyperscalers & data-center developers (gas turbines + on-site grid) — US data-center-related grid orders in the "high triple-digit-million-euro" range; >25% of the Gas Services backlog now data-center-driven.
• IPPs / oil & gas / industrials (compression, turbines).

Chokepoints & single-source dependencies: (1) GOES — the binding constraint on transformer output industry-wide; (2) the company's own factory capacity — turbine slots are sold to 2029–2030, some 2028, and management openly says "it's not so easy to ramp up"; (3) casting/forging and IGBT supply. The strategically important nuance: the chokepoint is upstream of Siemens Energy and inside it, not downstream — demand is not the problem, throughput is. That is a pricing-power position (Lens 3) but also caps how fast revenue can convert (Lens 11).

Competitive Advantages (moats)

The moat is oligopoly + backlog + installed base, sitting on top of a physical bottleneck.

• Grid Technologies (strongest moat). HVDC and large power transformers are a ~3-player global game (Siemens Energy, Hitachi Energy [ex-ABB Power Grids], GE Vernova Grid Solutions). Barriers: multi-year qualification, grid-code certification, brutal capital intensity, GOES access, and a 4-year order-to-delivery lead time that is itself a moat — a customer who needs power by 2028 cannot switch supplier without going to the back of an equally long queue. Grid automation is projected to grow ~8%/yr to 2035, with software ~40% of grid-systems value — a mix-shift toward stickier, higher-margin software/service.
• Gas Services. Heavy-duty gas turbines are a ~3-player oligopoly — Mitsubishi Power (~35% of MW orders, 2023), Siemens Energy (~24%), GE Vernova (~16%); top-3 >60% of the market. The real moat is the installed-base service annuity: once a customer's fleet is Siemens turbines, decades of high-margin LTSAs follow — "lifetime maintenance contracts generate more revenue than the turbine itself". Switching cost is a fleet-replacement decision.
• Bargaining power (who needs whom more). Right now, decisively tilted toward Siemens Energy: customers pre-pay, accept multi-year waits, and the supplier is deliberately rationing capacity to protect price. Peers are "limiting investments to protect margins… to reduce risks of overexposure if data-center demand fails" — a disciplined oligopoly, not a capacity war. That discipline is the bull case for margins and the tell that everyone remembers the last gas bust.
• Where the moat is thin: Siemens Gamesa. Wind turbines are a commoditizing, subsidy-dependent, quality-cursed business with Chinese entrants (Goldwind, Envision, Mingyang) undercutting Western OEMs. SG's "moat" is really offshore scale + a fixed installed base; onshore is a margin desert. The 2023 crisis proved the technical moat can invert into a technical liability.

Segments

Hard sourcing note: the sources pulled disclose segment growth rates and margins (, from the Q4 FY25 earnings release) but **not a clean euro revenue split** — so the euro revenue lines below are from FY2025 group revenue (€39.1bn) apportioned by the disclosed growth mix, and should be treated as approximate. The margins, growth rates, and orders are ``-sourced and load-bearing; the euro splits are directional.

Trend & cause:

• Grid Technologies is accelerating and is the profit-mix story — fastest growth and fattest margin, expanding as a share of group profit every quarter. Cause: AI-datacenter electrification + grid-replacement supercycle meeting an oligopoly with pricing power.
• Gas Services is the surprise re-rater — margin at 13% and rising on favorable pricing that management expects to persist; the "bridge-fuel + power-a-data-center-today" trade.
• Siemens Gamesa is the delta that matters for group margin — going from −13.1% toward break-even removes a ~€1.3bn+ annual drag. Management's FY2028 group-margin target of 14–16% (Lens 11) is largely a Gamesa-stops-bleeding + GT-mix-rises bridge.
• Geography: the US is the demand epicenter (data-center grid + gas orders; the reason for the $421m NC transformer plant and $1bn US manufacturing program). Europe is grid-replacement; Middle East/Asia is gas.

---

Phase B — Measure performance

Earnings Result (latest print: Q2 FY2026, reported 2026-05-12)

The print was a beat that triggered a guidance raise:

• Orders €17.7bn — an all-time high (+29.5% comparable), book-to-bill 1.72, backlog to a record €154bn. Driven by record Gas Services orders and a sharp Grid Technologies increase.
• Revenue €10.3bn (+8.9% comparable) — note revenue growth (~9%) lags order growth (~30%) badly: the constraint is throughput, not demand (Lens 2 confirmed in the tape).
• Profit before special items €1.16bn (from €906m yoy), margin 11.3% · Net income €835m.
• Siemens Gamesa operating loss shrank to €44m (from €249m); H1 loss €90m vs €623m — an ~85% improvement, with first orders for the SG 7.0 onshore platform (successor to the cursed 5.X).
• Guidance raised (the important part): comparable revenue growth 14–16% (was 11–13%); margin before special items 10–12% (was 9–11%); net income ~€4bn (was €3–4bn); FCF pre-tax ~€8bn. Driver: "stronger-than-expected Grid Technologies" + advance-payment cash inflows.

Balance-sheet / cash flags: FCF strength is real cash but flattered by advance payments on the record order book (contract liabilities rising) — see Lens 10. Net cash position solid; the state guarantee is gone (Lens 9). Dividend resumed and a €6bn buyback launched (Lens 9) — the clearest management signal that the balance sheet is healed.

Market reaction / what was priced in: despite the beat-and-raise, the stock is €152.66 as of 2026-07-10, down ~20% from its €191.66 all-time high (Apr-24-2026). That divergence — record results, falling stock — is the current setup: the fundamentals are still accelerating while the multiple is de-rating on peak-cycle fear (Lens 12/13). Prior-quarter Q1 FY26 print (Feb 2026) was similar: profit ~doubled to €1.16bn, orders ~€17.6bn, and it still met profit-takers.

Earnings Calls (sentiment trend)

No transcripts on the shelf (web-only); synthesized from release language and reporting across the last ~4 calls (Q3 FY25 → Q2 FY26).

• Tone has shifted from "prove the turnaround" to "manage the boom." Twelve–eighteen months ago management was defending the Gamesa fix and the balance sheet; now the recurring frame is capacity, throughput, pricing discipline, and cash. CEO Bruch's own line: "The combination of gas-fired capacity and modern grid infrastructure is driving our growth".
• Recurring phrases (rising): "record orders," "favorable pricing," "data-center demand," "book-to-bill," "advance payments," "disciplined capacity ramp."
• Phrases receding: "state guarantee," "stabilization," "Gamesa charges," "restructuring" — the crisis vocabulary is exiting the script, replaced by capital-return language ("dividend," "buyback," "mid-term targets").
• What management is focused on: (1) converting a €154bn backlog without blowing execution (the 2023 scar tissue); (2) holding pricing rather than chasing volume (the disciplined-oligopoly stance); (3) getting Gamesa to break-even and keeping it there; (4) signaling normalization via capital returns. Sentiment is confident, bordering on victory-lap — which is itself a mild contrarian caution flag near a cyclical/valuation peak.

Comps

Peer set: the global power-tech oligopoly.

Read-across: Siemens Energy at ~39x EV/EBITDA is expensive in absolute terms but trades at a ~20–35% discount to GE Vernova on forward FCF yield and EV/EBITDA per Barclays. Hitachi looks "cheap" (12x) but is a diversified conglomerate — not a like-for-like. The honest conclusion: the whole electrification complex is priced for a supercycle; Siemens Energy is the cheaper of the two pure-plays but neither leaves margin of safety, and GEV's ~60–95x multiples are the definition of a crowded trade.

Stock-Price Catalysts (>5% moves, ~last 5 years)

The chart is a near-perfect crisis-to-mania arc:

• Jun 21–23, 2023: −~35%. Gamesa quality bombshell — faulty components in 15–30% of turbine models; the defining crash. Reveals: wind quality risk is the tail that can take down the whole company.
• FY2023: −€4.4bn loss; Nov 2023: German government €7.5bn counter-guarantee (part of a €15bn package). Reveals: existential balance-sheet/guarantee dependence for a long-cycle EPC.
• 2024 → 2026: +~1,600% off the 2023 low; one of the best DAX performers (a YTD +146% vs DAX +21% stretch reported late 2025). Reveals: the market reacts most violently to the two swing factors — (a) the AI-power/data-center order narrative and (b) Gamesa loss reduction. Every beat-and-raise on those two axes = leg up.
• Dec 2025: near record high on "activist pressure, buybacks, AI power boom".
• Apr-24-2026: all-time high €191.66.
• ~Q2 2026: de-rating to ~€153 — Barclays downgrade on peak-cycle valuation (below) is the proximate catalyst; the stock now reacts to valuation/cycle-timing, not just order flow. The market's fear has rotated from "can they execute?" to "is this the peak?"

Pattern verdict: this name trades on narrative inflection more than on the quarter — data-center demand headlines and Gamesa milestones up, peak-cycle/valuation calls down.

---

Phase C — Judge people & books

Management

• CEO — Christian Bruch (since May 2020, i.e. from the spin-off). Ex-Linde: CEO of Linde Engineering and 15+ years across Linde's board/operations. Track record, quantified: took the company public, then steered it through the 2023 Gamesa near-death (a €4.4bn loss year and a state bailout) to a ~1,600% recovery, record €154bn backlog, resumed dividend, and unwound government guarantees by 2025. That is one of the more complete large-cap turnarounds in European industrials. The fair caution: he was also CEO when the Gamesa quality crisis detonated — the 2022 acquisition/squeeze-out of Gamesa minorities into an unfixed quality problem happened on his watch. Credit for the fix must carry the debit for the cause.
• CFO — Maria Ferraro (since May 2020). Ex-CFO Siemens Digital Industries; ex-CFO/CDO Siemens AG (2019–2020). Architected the balance-sheet rescue → guarantee-unwind → dividend/buyback normalization. Credible, tenured, aligned with the same spin-off cohort as Bruch.
• Tenure & skin in the game: both are founding-listing executives (~6 years) — long enough to own both the crisis and the cure. Insider ownership is modest (professional managers, not founders) — our figures absent; specific holdings n/a.
• Capital-allocation history: early years = survival (raise guarantees, pause dividend, fix Gamesa). FY2026 = normalization: €0.70/share dividend (first since 2022, 99.99% AGM approval Feb-2026) + €6bn buyback through 2028. Judgment call: buying back stock at ~40x earnings near an all-time high, while the core constraint is factory capacity, is a debatable use of cash — it signals confidence but may not be value-maximizing versus capacity or the transformer-plant build-out. Watch this.
• Ownership overhang: parent Siemens AG has steadily sold down — 35.1% at spin-off → 17.1% (Dec-2023, after an 8% transfer to Siemens Pension-Trust) → ~14.96% (Jan-2025) → announced intent to sell a further 6% (€2.5bn) in Jul-2025, leaving ~9%. Plus Siemens Pension-Trust ~8–10%. A recurring, telegraphed supply overhang — not a governance threat, but a capped-upside factor.
• Archetype: professional-manager turnaround operators, not founders — exactly the right archetype for a heavy-industrial de-risking-and-scaling phase. Founder-style risk appetite is not what this asset needs now.

Forensic Red Flags

Regulatory file on shelf (regulatory/regulatory-findings.md): SEC = 0 findings, because there is no CIK — Siemens Energy is not an SEC filer, so EDGAR LR/AAER searches are structurally empty (not a clean bill of health, just an inapplicable venue). Forensic analysis therefore rests on IFRS disclosures via IR + web.

Accounting-risk map (IFRS long-cycle EPC — the classic areas):

1. Revenue recognition (highest-attention area). Percentage-of-completion / over-time recognition on multi-year contracts is inherently estimate-driven — margin can be pulled forward or pushed out via cost-to-complete assumptions. Gamesa 2023 is the case study of this risk going wrong: warranty/failure-rate estimates proved badly optimistic, forcing a €472m component-failure charge (rotor blades, main bearings, 4.X/5.X platforms) and, cumulatively, ~€2.2bn+ of wind hits. The lesson: estimate integrity is the whole game here.
2. Advance payments / contract liabilities — the flag that matters now. FCF (~€8bn FY26 guided) is structurally ahead of net income (~€4bn) because a record order book pulls in customer cash before revenue. This is legitimate — but it is timing, not permanence. When order growth normalizes, the contract-liability tailwind reverses into a net-working-capital headwind; Barclays models that turning material from ~2028. Do not extrapolate FY26 FCF as a run-rate. This is the number most likely to embarrass an over-anchored bull.
3. Provisions & warranties. Gamesa warranty/quality provisions and large-project loss provisions are judgment-heavy; the SG 7.0 ramp is a fresh place for optimism to hide. Watch provision release flattering margins vs genuine operational gain.
4. Special items / "before special items" framing. Management steers on profit before special items — a normalization that has historically absorbed billions in restructuring/quality charges. The gap between "before special items" (6.0% FY25) and statutory net income is where the pain lives; the gap is narrowing (good) but is the right place to keep watching.
5. Goodwill/intangibles on the Gamesa acquisition — impairment risk should wind losses re-open. SBC is not a material distortion for a German industrial (unlike US tech).

Regulatory / legal — material finding (do not miss): on Sept-30-2024, Siemens Energy Inc. (the US subsidiary) pled guilty to US federal criminal charges for misappropriating confidential competitor bidding information from GE and Mitsubishi Heavy to gain an edge on a gas-turbine plant bid in Chesterfield, Virginia, and paid a $104m criminal penalty; several employees also pled guilty. This is a genuine integrity red flag on the standalone entity (not just heritage). Separately, heritage context: parent Siemens AG's 2008 FCPA settlement (~$1.6bn) predates the spin-off and is not attributable to Siemens Energy's own conduct, but it colors the compliance-culture lineage. A "Siemens Government Technologies vs Solaria" $9.5m item surfaced in search is a different entity — excluded as a false positive. No material EU-Commission antitrust or sanctions action against Siemens Energy AG surfaced for 2025–2026. Item 3 (Legal Proceedings): n/a — no 10-KForm 10-KA company’s audited annual report to the US regulator. The most complete thing it publishes. on shelf (not an SEC filer); IFRS annual-report legal-proceedings note not separately pulled.

Forensic verdict: books are not a fraud-risk story; they are an estimate-and-timing story. The two things to underwrite are (a) that the advance-payment FCF tailwind reverses (model it, don't extrapolate it) and (b) that Gamesa's improved margins are operational, not provision-release. The 2024 DOJ guilty plea is a real, if contained, integrity mark.

---

Phase D — Project & stress-test

Forward Projection (FY2026 → FY2028)

Built bottom-up from FY2025 actuals + raised FY2026 guidance + management's Nov-2025 Capital Markets Day mid-term targets (low-teens comparable revenue CAGR to FY2028; group margin 14–16% by FY2028, upgraded from 10–12%). All outputs ``; every input labeled. Shares ~857m, drifting down ~1–2%/yr on the €6bn buyback. Per --watchlist rules and task instruction, no our model Brier forecast is logged.

Base-case narrative: Grid Technologies compounds >20% at a 16%+ margin; Gas Services holds ~13–14% on favorable pricing and service mix; Gamesa flips from −13% to break-even (FY26) to modestly positive (FY27–28) — that swing alone adds ~€1.3bn+; group margin marches to management's 14–16% by FY2028. FY28 EPS ~€7.0 puts the stock at ~22x FY28 earnings today — reasonable if the target is hit.

Bull path (FY28 EPS ~€8.5): Gamesa reaches +3–5% margin; GT margin to ~18%; gas pricing holds through 2028; buyback accelerates; backlog converts faster than feared.

Bear path (FY28 EPS ~€5.0, roughly flat vs FY26): peak-cycle bites — gas/data-center orders normalize post-2027 (or a DeepSeek-style efficiency shock hits demand); NWC reverses (advance-payment drag) from 2028 compressing FCF and forcing conservative revenue recognition; Gamesa stalls at break-even; pricing softens as oligopoly capacity finally catches up ~2029. In this path today's ~40x forward multiple is a value trap.

Falsifiable base call (for tracking, not logged): Siemens Energy FY2028 (ending Sep-2028) group margin before special items ≥ 14% and EPS ≥ €6.5, p≈0.55. The disagreement with the market is not on direction but on durability — bulls capitalize FY28 as a plateau, Barclays calls it a peak.

Bull vs Bear

Bull case. Siemens Energy owns two of the three hardest physical bottlenecks in the AI-electrification supercycle — grids and gas turbines — inside disciplined oligopolies, with a €154bn backlog (book-to-bill 1.72) that de-risks 3–4 years of revenue and is pre-funded by customer cash. Grid Technologies is a structural compounder (grid capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. must rise ~50% from $400bn/yr by 2030 per the IEA; data-center power to ~945 TWh by 2030). Gas Services is a re-rated cash machine the "energy transition" narrative wrote off. Gamesa going from ~€1.3bn/yr drag to break-even is a self-help earnings lever independent of the cycle. Management has proven it can execute under stress (guarantees unwound, dividend restored, €6bn buyback). Potential upside surprises: faster Gamesa profitability, GT margin breakout, further mid-term-target upgrades (they've upgraded twice).

Bear case (2–3 permanent-impairment / de-rating risks).

1. Peak-cycle timing. FY2026 is the guided free-cash-flow peak (~€8bn); Barclays argues gas-turbine peak conditions, supply-demand tightness, and peak FCF all arrive together in 2026, with NWC turning into a material headwind from 2028 — i.e. the stock should de-rate from here, which it has begun to (−20% from ATH). Not an impairment of the business, but of the multiple.
2. Data-center demand durability. >25% of the gas backlog leans on data-center power; a DeepSeek-style efficiency shock or DC overbuild could trigger order push-outs/cancellations right as 2028–29 capacity comes online. The last gas boom ended in a bust; peers are rationing capex precisely because they remember.
3. Gamesa relapse. Break-even is a target, not a fact; wind is a structurally poor business with Chinese pricing pressure and a live 4.X/5.X warranty tail. A re-opened charge would re-break the group-margin bridge.

Pre-mortem (18 months out, thesis broke): it's early 2028; a 2027 AI-capex air-pocket (efficiency gains + a hyperscaler capex pause) froze new gas/grid orders; the €154bn backlog stopped growing, so advance payments stopped inflating FCF and reversed; reported FCF halved versus the €8bn FY26 peak; a fresh Gamesa offshore charge landed; and a stock at ~40x priced-for-perpetual-acceleration re-rated to ~20x. Down 40–50% from the highs — no fraud, just peak multiple × peak fundamentals × cycle.

Contrarian view of what the market refuses to see: the bull consensus treats the €154bn backlog as quality when its growth is what secretly funds the cash flow — the backlog is simultaneously the moat and the working-capital sugar high. The market is also under-weighting that management is deliberately not building enough capacity (to protect price), which caps the very revenue growth the multiple assumes. Are current multiples too high? On trailing/near-term, yes (~39x EV/EBITDA, ~40x fwd EPS); on FY2028 targets, defensible (~22x) — so the entire debate is whether you believe FY2028 is a plateau or a peak.

Devil's Advocate (short-seller)

Dismantling the bull case:

• What structurally breaks the money machine: the advance-payment flywheel. Bulls quote ~€8bn FCF as if it's a run-rate; it is negative-working-capital financing that mathematically reverses when order growth flattens. A short's cleanest trade is "FCF was borrowed from the future."
• Revenue concentration & shift: a quarter+ of gas backlog is data-center-linked, and data-center capex is the single most reflexive, sentiment-driven spend in the economy. If hyperscalers blink (efficiency, a funding tightening, a demand rethink), the marginal order — where the margin is richest — vanishes first.
• Weaker moat than bulls think: in wind it's negative; in gas the moat is the installed base, but new-unit pricing is cyclical spot-tightness that competes away when Mitsubishi/GE capacity lands ~2029; in grid the moat is real but capped by the company's own throughput — you cannot compound 25% forever off factories you are deliberately under-building.
• Most dangerous competitor bulls underestimate: not GE Vernova — Chinese OEMs (grid gear + wind) on export, and Mitsubishi Power's gas-turbine share leadership. Plus GEV's cleaner balance sheet and richer currency for M&A.
• Worst capital-allocation move: a €6bn buyback at ~40x earnings near an all-time high while the binding constraint is capacity — buying stock instead of building the factories that would break the throughput ceiling.
• Assumptions that must hold for today's price: perpetual >12% growth and 14–16% margins and no cycle and no Gamesa relapse and advance payments never reverse. That's a lot of ANDs at ~40x.
• Growth disappoints 20–30%: at ~22x a base FY28 that then gets cut, the multiple and the number fall together — the classic peak-cycle double-derating; 40–50% downside is on the table.
• Single scenario that permanently impairs: a genuine data-center-demand reset (AI capex proves front-loaded) that converts a "sold-out to 2030" book into a cancellation cycle — plausibility moderate, not remote, and rising the more the whole complex prices in permanence.

Management Questions (ordered by information value)

1. Advance payments: of the ~€8bn FY26 FCF, how much is structural cash generation vs a contract-liability timing benefit — and at what order-growth rate does net working capital flip from tailwind to headwind?
2. Backlog quality: what share of the €154bn (and specifically the data-center gas book) is cancellation-protected / take-or-pay, and what are the escalation, FX, and input-cost pass-through terms?
3. Grid Technologies margin: 15.8% today — how much is cyclical spot pricing vs structural mix/productivity, and where is the through-cycle ceiling?
4. FY2028 14–16% target: bridge it by segment — how much depends on Gamesa turning positive vs GT mix vs gas pricing?
5. Gas-turbine peak: with slots sold to 2029–30, what is your normalized new-unit margin once competitor capacity lands ~2029, and how exposed is the 2028–29 delivery window to data-center cancellations?
6. Gamesa durability: beyond FY26 break-even, what is normalized through-cycle wind margin, and what 4.X/5.X warranty tail remains?
7. Capacity vs buyback: why €6bn of buyback at ~40x rather than more capacity to break the throughput ceiling — walk through the reinvestment-return math.
8. Capacity discipline: are you deliberately under-investing to protect pricing, and what demand signal would make you add capacity aggressively?
9. Normalized FCF: what is through-cycle FCF conversion once advance payments stop growing?
10. Siemens AG overhang: expected path/timeline for the residual ~9% stake and the Pension-Trust holding?
11. Compliance: structurally, what changed after the 2024 US DOJ trade-secret guilty plea?
12. Transformation of Industry / hydrogen: is the electrolyzer business a path to profit or a capital sink — decision criteria and timeline?
13. Supply chain: GOES and IGBT single-source exposure — how far is transformer output de-risked by the NC plant and dual-sourcing?
14. Offshore wind political risk: exposure of the Gamesa offshore book to US/EU permitting and policy reversal?
15. M&A: with a healed balance sheet and a rich currency, what is the appetite, and where are the gaps you'd buy?

---