CGN Power

Phase A — Understand the business

Lens 1 · Company Overview

CGN Power is the listed operating vehicle of China General Nuclear Power Corporation (CGNPC), the larger of China's two nuclear champions, and it is the largest nuclear power operator in China and one of the largest in the world by units in operation. The business is simple to state and hard to disrupt: it owns and operates nuclear power stations, generates electricity, and sells that electricity to provincial grids. Electricity sales are ~95% of revenue; the rest is construction/installation/design and technical services rendered to sister projects.

Scale (as of 31 Dec 2025): 28 nuclear power units in operation, 31.838 GW total managed installed capacity, with roughly 20 units under construction — the largest new-build pipeline of any single operator on earth. FY2025 on-grid generation was 232.648 TWh.

How money is actually made — the contract structure that matters:

• Regulated on-grid tariff. Each unit sells to its provincial grid at a benchmark on-grid tariff historically anchored to the local coal-benchmark price, designed for cost recovery plus an allowable return. Reference points: Hongyanhe ~CNY 0.4142/kWh, Ningde ~CNY 0.43/kWh (2015 vintage).
• Market-based trading — the growing, and dangerous, slice. A rising share of output is sold via provincial spot/bilateral markets. By H1-2023 market-based volume was already ~55.5% of on-grid generation at ~CNY 0.4022/kWh. This is the crux of the whole thesis: the more output that clears the market rather than the benchmark, the more CGN's revenue floats down toward the marginal (increasingly renewable-depressed) provincial clearing price.
• No take-or-pay. Unlike a Western PPA-backed IPP, CGN is a merchant-plus-regulated hybrid, not contracted. There is guaranteed dispatch priority (nuclear runs baseload) but not a guaranteed price for the traded portion.

Customers: provincial grid companies (State Grid / China Southern Power Grid subsidiaries) in Guangdong, Fujian, Guangxi, Liaoning, Zhejiang — Guangdong is the single largest exposure. Suppliers: fuel from CGN's own fuel-cycle arm and CGN Mining (uranium); reactor islands increasingly the indigenous Hualong One (HPR1000) and CAP1000. Competitors: its A-share twin China National Nuclear Power (CNNP, 601985.SS) is the only true domestic peer; beyond that, coal, wind and solar compete for the same provincial demand and increasingly set the marginal price CGN's traded MWh clear against.

Take: this is a policy asset dressed as an equity — a state-directed baseload builder whose revenue line is a regulated/administered price, not a market one it controls. The bull frame ("nuclear renaissance, AI power demand") and the bear frame ("administered-price utility with eroding tariffs") are both true; which dominates is a question about China's tariff regime, not about CGN's operations.

Lens 2 · Supply Chain

Upstream → CGN → end customer, named at every link:

Fuel & front end:

• Uranium: sourced via CGN Mining (1164.HK) and CGN's overseas mines/offtake (Husab in Namibia via Swakop Uranium; Kazakhstan JVs) — the group is vertically integrated into uranium, a genuine differentiator vs. a pure operator.
• Conversion/enrichment/fabrication: handled within the CGN group fuel-cycle companies and CNNC-affiliated enrichers (China is largely self-sufficient in the fuel cycle).

Reactor technology & construction:

• Reactor islands: indigenous Hualong One (HPR1000) and, for newer approvals, CAP1000 (Westinghouse-derived AP1000 localisation). Legacy fleet includes French-derived M310/CPR-1000 and EPR (Taishan) designs.
• EPC / heavy equipment: China National Nuclear Corporation (CNNC) engineering arms, Dongfang Electric, Shanghai Electric, Harbin Electric for turbines/generators/pressure vessels; CGN's own China Nuclear Power Engineering and CGN Engineering for construction management.

The company (CGN Power): operates the plants, runs O&M, manages refuelling outages, and dispatches to grid.

End customers: China Southern Power Grid (Guangdong, Guangxi) and State Grid subsidiaries (Fujian, Liaoning, Zhejiang), which on-sell to industrial/commercial/residential load.

Chokepoints & single-source dependencies:

1. Regulatory approval is the true bottleneck, not equipment. New units require State Council sign-off; the pace of approvals (10-12/yr nationally in 2022-2025) gates growth, and the government controls it.
2. Uranium price/security — mitigated by CGN Mining vertical integration but not eliminated; a spot-uranium spike raises fuel cost with a lag.
3. Grid absorption — in high-renewable-penetration provinces, nuclear's traded share must clear against near-zero-marginal-cost solar/wind, compressing captured price. This is a chokepoint on price, not volume.
4. Localisation is a moat, not a risk — because the Hualong One supply chain is ~85%+ domestic, CGN is largely insulated from Western sanctions/export controls (the ASML/foundry vulnerability that plagues Chinese semis does not apply here).

Names or it didn't happen — delivered. The supply chain is unusually self-contained within the Chinese state-industrial complex, which is exactly why this name screens as sanction-proof relative to almost anything else on the energy/hardware frontier.

Lens 3 · Competitive Advantages (moats)

The moat is regulatory + scale + fuel-cycle integration — and it is deep, but it caps the upside as much as it protects the downside.

1. Licence & approval moat (durable, state-conferred). You cannot build a nuclear plant in China without being CGN or CNNC. There are structurally two players; new entrants are effectively barred. This is the strongest moat CGN has — but it is granted by the state, and the same state sets the tariff.
2. Scale & operating experience. 28 units and >7,550 average utilisation hours/unit in 2025 — well above the national all-source average — reflect a fleet that runs hard and reliably. Operating know-how across HPR1000/CAP1000/EPR is genuinely hard to replicate.
3. Fuel-cycle vertical integration via CGN Mining — bargaining power over the uranium input that few global peers (bar Rosatom, KEPCO) possess.
4. Cost position. Once built, nuclear's marginal cost is very low; CGN's baseload dispatch priority means near-100% capacity-factor economics. The problem is not cost — it's the administered price on the revenue side.
5. Non-competition agreement with CGNPC = a contractual asset-injection pipeline (see Lens 5/11). CGN Power has the right to buy the parent's maturing plants — an embedded, low-execution-risk growth optionality worth several GW over the decade.

Bargaining power — who needs whom? Over suppliers: high (state-directed EPC, captive fuel). Over customers (the grid/regulator): low — the NDRC and provincial pricing bureaus set the tariff, and CGN is a price-taker on the regulated portion and a price-follower on the traded portion. This asymmetry is the whole reason a monopoly-ish operator can still see earnings fall 10% in a good demand year.

Take: it is a wide-moat business with a capped ROE — the moat protects the franchise but the regulator harvests the surplus. That is the correct mental model for pricing it: closer to a regulated utility than to a merchant IPP, and nowhere near a growth compounder despite the reactor-count headlines.

Lens 4 · Segments

segments.csv is empty (research layer unavailable) — all figures below are ``, and a full product/geography EBIT split by unit is n/a at the segment-P&L level. What is sourceable:

• By activity (FY/H1-2025): electricity sales dominate. In H1-2025, total electricity-sales revenue was ~RMB 30.658B = 78.27% of operating revenue; the balance is management services to associate/JV projects, construction/technical services, and power-sales-company throughput.
• A telling sub-split: in H1-2025 the group's own subsidiaries achieved grid-connected sales of 89.265 TWh, while the power-sales companies handled ~11.712 TWh of third-party consumption — i.e. CGN is building a retail/trading layer on top of generation, which grows revenue but at thin margin and adds price-risk.
• By geography: not broken out numerically in public snapshots, but exposure is concentrated in Guangdong (largest), then Fujian, Guangxi, Liaoning, Zhejiang. Guangdong's market-based tariff trajectory is therefore the single most important external variable.
• Consolidation scope note: headline "managed capacity" (31.838 GW / 28 units) includes JV/associate units not fully consolidated — generation figures often say "(including joint ventures)". Consolidated revenue reflects a smaller effective share than the 28-unit headline implies. This gap between managed and consolidated is a recurring source of overstatement in bull write-ups.

Trend & cause: volume is accelerating (generation +6.93% YoY in H1-2025, full-year on-grid 232.65 TWh) while revenue is decelerating/declining (FY2025 rev −4.1%). Volume up, revenue down = price down. That single sentence is the segment story.

---

Phase B — Measure performance

Lens 5 · Earnings Result (latest print — FY2025, reported Mar 2026)

The headline: growth in every physical metric, decline in every profit metric.

H1-2025 detail (the cleaner read on the trend): revenue RMB 39.167B (−0.53% YoY); net profit RMB 5.952B (−16.3% YoY). Management's own attribution of the H1 decline: (1) lower average tariff, (2) reduced VAT refunds, (3) higher operating costs.

What drove it:

• Tariff is the primary driver. As more output clears via market trading in high-renewable provinces (Guangdong especially), the blended realised price falls even as regulated benchmarks hold. 2024's market-based tariff was already −3.9% YoY; 2025 extended the erosion.
• VAT refund step-down (see Lens 10) reduced a below-the-line subsidy that had been flattering net profit — a policy hit, not an operating one, but permanent for affected units.
• Operating costs rose with new-unit commissioning (D&A on newly commercial reactors, staffing, first-cycle fuel).

Balance-sheet flags: net gearing ~133% at end-2024, net debt/equity ~108.9%, total debt ~CN¥207.8B vs equity ~CN¥174.3B. This is a capital-intensive, heavily-levered balance sheet — normal for a nuclear builder mid-cycle, but it means rising rates or a construction-cost overrun bite hard, and it constrains dividend growth (payout already stretched — see below).

Market reaction: the H-share sits mid-range (52-wk HK$2.47–3.77, last ~HK$3.48–3.52 ); the stock did not collapse on the −9.9% print, which itself is the tell — the market is still paying a growth/defensive premium for a shrinking-earnings utility (Morningstar's repeated "shares overvalued" verdict across 2024-2026 reports ).

Flag vs. own history: a ~116% headline payout ratio appears in one data source vs. a historical 26-53% range — almost certainly a distortion from the dividend-vs-adjusted-earnings basis or a data error, but worth verifying against the audited accounts; if real, the dividend is being part-funded from the balance sheet, which would be a red flag for a levered builder. n/a — not independently sourced; verify against the FY2025 annual report.

Lens 6 · Earnings Calls (sentiment trend)

transcripts/ is empty — CGN does not host US-style quarterly calls with public transcripts; disclosure is via HKEX results announcements and Chinese-language investor briefings. Sentiment is therefore reconstructed from results-announcement language and analyst readouts across 2024→2025:

• 2024 interims: "Disappointing results due to lower tariff and higher costs" [Morningstar report title, 2024].
• Late-2024 / early-2025: "Downward pressure on CGN Power's average tariff" [Morningstar, 2024-25].
• Mid-2025 (H1): "Weak average tariff continues to pressure profit" [Morningstar, Aug 2025].
• 2025 later: "Positive outlook for Guangdong market-based tariff; shares remain overvalued" — a small tonal thaw on tariffs, but the "overvalued" tag persists [Morningstar, 2025].
• 2026: "Tariff Risks Persist for CGN Power" [Morningstar, 2026].

Recurring management themes: generation-volume growth, unit-commissioning milestones (Hualong One batch construction, fuel-loading at Taipingling/San'ao), 15th-FYP nuclear expansion, "green/low-carbon baseload" positioning. What they've stopped emphasising: tariff upside — the framing shifted from "stable regulated returns" toward "volume growth will offset price." That pivot is itself bearish: it concedes price is no longer the friend it was.

Sentiment arc: defensive-and-deteriorating on price, confident-and-improving on volume. Management is leaning on the reactor pipeline narrative precisely because the tariff narrative broke.

Lens 7 · Comps

The one comp that matters — CGN vs CNNP:

• CNNP grew FY2025 net profit +6%; CGN fell −9.9%. Same country, same regulator, same tariff reform.
• The difference: CNNP is ~33.6 GW of wind+solar on top of its nuclear — its renewables/diversification absorbed the nuclear-tariff hit, while CGN is a near-pure nuclear play and took the full blow.
• Yet CGN's H-shares trade at a P/E premium to the "renewable peer average of ~8.1x" and roughly in line with-to-above CNNP's ~13.8x TTM — i.e. the slower-growing, less-diversified, more tariff-exposed name is not cheaper.

Comp verdict: on the only apples-to-apples pair, CGN screens expensive for its growth and risk profile. The H-share's ~4% yield is the one genuinely attractive number; the P/E is not.

Lens 8 · Stock-Price Catalysts (>5% moves, last ~5 yrs)

Mostly ``; a full tick-by-tick 5% event log is n/a without price-history data, but the pattern is clear from what moved the name:

• A-share IPO (Aug 2019) and subsequent A/H convergence — the Shenzhen listing created a persistent A-share premium to the HK line; A-share ATH CNY 5.25 on 25 Jul 2024, A-share low CNY 2.49 (Jul 2021). The 2021→2024 A-share run (roughly a double off the low) tracked the nuclear-approval acceleration and the "China nuclear renaissance / dual-carbon" narrative.
• Unit-commissioning milestones (e.g. Fangchenggang-3 commercial op Mar 2023; Taipingling-1 first Greater-Bay Hualong One op Apr 2026) — supportive but incremental.
• New-build approvals (State Council batch sign-offs, 2022-2025, 10-12 units/yr) — the biggest sentiment catalyst; approvals = future earnings visibility.
• Tariff/VAT policy news (MOF VAT Announcement No. 10, 2025; market-pricing reform from Jun 2025) — the biggest fundamental de-rating catalyst; each incremental step of market-pricing exposure is a negative.
• Dividend/convertible-bond actions — the RMB 4.9B A-share convertible (listed Shenzhen Jul 2025, conversion from Jan 2026) and dividend-driven conversion-price cuts moved the A-line.

What the market actually reacts to for this name: (1) approvals/pipeline (bullish narrative fuel), and (2) tariff & tax policy (bearish fundamental reality). It is a policy-catalyst stock far more than an earnings-surprise stock — the swing factor is Beijing's pen (approvals and pricing), not the quarterly beat/miss.

---

Phase C — Judge people & books

Lens 9 · Management

• Chairman: Yang Changli (Chairman since ~May 2020; re-affirmed 2025) — a career state-nuclear/SOE executive.
• President / Executive Director: Gao Ligang.
• Archetype: these are professional state managers / Party-system executives, not founder-owners. That is exactly right for a strategic SOE — the job is to execute national energy policy safely and expand the fleet, not to maximise per-share value. Insider ownership is immaterial (insider-transactions.csv unavailable; personal stakes n/a but structurally negligible).

Track record: operationally excellent — a large fleet run at >7,550 utilisation hours with a strong safety record, and the world's biggest new-build program executed on the indigenous Hualong One. On delivery of physical capacity, management is best-in-class globally.

Capital allocation: this is where the shareholder and the state diverge. Capital is directed toward massive multi-year new-build capex (guidance ~RMB 70-100B cumulative 2024-26 ), funded with debt (gearing ~133%), green bonds, state-bank and export-credit facilities, and a convertible. Dividend payout has run ~35-45% historically (one source shows a distorted ~116% for the latest period — verify). ROE is structurally mid-single-to-high-single digits and falling with the tariff — capital is being reinvested at regulated, declining returns. From a pure equity-value lens that is value-neutral-to-dilutive reinvestment; from a national-strategy lens it is the point of the company.

Red flags (governance): the 58.91% parent (CGNPC) control means related-party transactions are the business model, not an exception — asset injections from parent, fuel purchases from group, EPC from affiliates. All are disclosed and regulated, but minority holders are structurally junior to state policy objectives. No promotional behaviour, no strategy whiplash, no comp scandal surfaced.

Take: trust the operators to build and run reactors; do not expect them to run the company for the minority H-share holder. Governance is competent and clean but agency is aligned to the state, not to you.

Lens 10 · Forensic Red Flags

financials.csv unavailable → this lens is qualitative + policy-driven, every figure labelled.

1. The VAT-refund step-down — the single most important accounting/earnings item, and it is structural, not one-off. Under legacy Notice 38/2008, nuclear units enjoyed a generous VAT refund (a real cash subsidy running through profit). MOF Announcement No. 10 (2025) re-cut the regime: units approved but not in commercial operation before 31 Oct 2025 get only a 50% refund for 10 years; grandfathered older units keep the old treatment. Implication: the entire new-build pipeline — the very growth the stock is priced for — will earn less subsidy per MWh than the legacy fleet. The market narrative ("more reactors = more earnings") collides with a declining per-unit after-tax economics reality. This is the most under-appreciated line in the whole story.
2. "Managed" vs. "consolidated" capacity gap. Headline 31.838 GW / 28 units includes JV/associate units. Bull models that apply full-fleet generation to consolidated economics overstate attributable earnings. Watch for "(including joint ventures)" qualifiers.
3. Leverage & interest. Net gearing ~133% (2024) — a construction-heavy balance sheet where capitalised interest during construction flatters the income statement now and converts to D&A + interest expense as units go commercial. The −15.1% adjusted net-profit fall (worse than the −9.9% headline) suggests non-recurring gains propped the headline — the underlying trend is worse than the top line.
4. Dividend sustainability. The disputed ~116% payout figure — if genuine — means the dividend exceeds attributable profit, i.e. part-funded by debt/balance sheet on a business already at 133% gearing. Flag to verify.
5. Cash-flow vs. earnings / receivables: grid receivables from state utilities are low-default but can stretch on timing; n/a at the working-capital-detail level. Nuclear decommissioning & spent-fuel provisions are long-dated liabilities whose adequacy is opaque in summary data — standard nuclear-sector caveat.

Regulatory findings (required sub-section):

• SEC (EDGAR EFTS — LR + AAER): regulatory/regulatory-findings.md records zero findings — CGN Power has no CIK and is not an SEC registrant, so no EDGAR enforcement search is possible.
• Non-SEC enforcement (web search: "CGN Power" + FTC/DOJ/consent decree/fine/penalty): the material item is US-side, and it concerns the parent group, not the listed CGN Power directly — CGNPC and affiliates (notably CGN and a subsidiary China General Nuclear Power / and the CGN-linked entity involved in the Ho/TVA case) were placed on the US Entity List (Dept. of Commerce, 2019) over alleged diversion of US nuclear tech to military use; a related DOJ economic-espionage matter (US v. Allen Ho / Szuhsiung Ho, 2016, conviction) named CGN's US nuclear-consulting activity. This is a parent/geopolitical overhang, not an accounting-fraud finding against 1816.HK — but it (a) constrains CGN's access to US/Western nuclear technology and (b) is a live sanctions-risk factor for the equity. Label: material geopolitical/regulatory overhang.
• Item 3 / Legal Proceedings from a 10-K: n/a — no 10-K exists (HKEX/Shenzhen filer). The equivalent HKEX annual-report litigation note was not machine-ingested; verify against the FY2025 annual report on hkexnews.hk before relying on a "clean" litigation read.
• Net: No accounting-fraud or securities-enforcement finding against the listed entity via SEC EFTS or web search as of 2026-07-06. However, there is a real, sourced US Entity-List / sanctions overhang on the CGN group that any thesis must carry — verified via SEC EDGAR EFTS (nil, no CIK), web search (parent Entity-List hit), and the absence of an ingested HKEX Item-3 equivalent (unverified — chase the annual report).

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (EPS, next 3 fiscal years — FY2026/27/28)

Grounding: financials.csv/guidance.csv empty → this is a web-anchored ``, arithmetic shown. Base actuals: FY2025 net profit RMB 9.765B; ~50.5B shares → FY2025 EPS ≈ RMB 0.193. (Consensus "next-year EPS ~CNY 0.21" implies the Street models a recovery, above trailing.)

The projection's core tension: every path adds reactors and generation; the paths differ almost entirely on realised tariff. This is a price-of-power bet dressed as a volume-growth story. The base case is not a compounder — it is a ~4% yield + low-single-digit EPS growth = high-single-digit total return utility, and only if tariffs merely drift rather than fall.

Brier forecast (logged as the base call): "CGN Power (1816.HK) FY2026 net-profit-attributable ≥ RMB 9.5B (i.e. roughly flat-to-slightly-down vs FY2025's 9.765B), p≈0.55, resolves 2027-03-31." — reflecting that continued tariff erosion makes even flat profit a coin-flip-plus, not a given. (Not written to forecast.ts per --watchlist rules; recorded here for the position seed.)

Lens 12 · Bull vs Bear

Bull case. The world's largest nuclear operator sits at the centre of the only credible zero-carbon baseload in a country adding power demand relentlessly (data-centres/AI, electrification, industry). It has a two-player regulatory monopoly, an indigenous, sanction-proof Hualong One/CAP1000 supply chain, a fuel-cycle vertical integration few peers match, and the largest, most-visible new-build pipeline on the planet (~20 units under construction + a contractual asset-injection queue from a 58.9% parent). China's 15th FYP targets 110 GWe nuclear by 2030; CGN is the prime beneficiary. Buy it for the ~4% H-share yield, defensive baseload cash flows, and a decade of visible capacity growth at a China-utility multiple. If Guangdong's market-tariff bottoms (Morningstar's own "positive outlook for Guangdong tariff" ), the earnings decline reverses and the "overvalued" tag flips.

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

1. Tariff regime is structurally, not cyclically, eroding. As China marketises power (renewables-must-trade from Jun 2025, rising market-share of nuclear output), nuclear's regulated cost-recovery umbrella is being replaced by exposure to a renewable-depressed marginal price. FY2025's volume-up-revenue-down result is the early signal, not the trough. This can compress ROE for years.
2. VAT-refund step-down permanently lowers new-unit economics — the growth pipeline earns less after-tax per MWh than the legacy fleet (MOF Announcement No. 10). The stock is paying a growth premium for units that are less profitable than the ones already running.
3. Minority-holder agency risk. A 58.9%-state parent runs the company for national policy; capital is reinvested at declining regulated returns and related-party transactions are pervasive. You are a passenger, not a principal. Add the US Entity-List overhang on the group as a tail sanctions risk.

Pre-mortem (18 months out, thesis broke): It's early 2028. CGN's traded-output share in Guangdong/Fujian crossed ~70%, blended tariff fell ~5% two years running, and the three units that went commercial in 2026-27 all landed in the 50%-VAT bucket — so net profit printed below RMB 9B for FY2026 and again FY2027 despite +15% cumulative generation. The Street's "recovery to CNY 0.21 EPS" never came; the H-share de-rated from ~14x toward the ~8x renewable-peer multiple and the yield "support" gave way because the stretched payout was cut to protect the 133%-geared balance sheet. That is the plausible break — and note it requires nothing exotic, just the current trend continuing.

Are multiples too high? On the only clean comp (CNNP), yes — CGN is not cheaper than a faster-growing, better-diversified twin, and trades above the renewable-peer average. The ~4% yield is fair; the ~14x P/E on falling EPS is the vulnerable part.

Contrarian view (what the market refuses to see): the market is trading CGN as a "nuclear renaissance growth story" when the numbers say it is a "regulated utility in gentle managed decline of returns." The reactor-count headlines and AI-power-demand narrative are masking the fact that China's own power-market reform is quietly converting CGN's best asset — a protected tariff — into a liability. The pipeline is real; its per-share profitability is deteriorating. Bulls are counting reactors; the regulator is counting yuan.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• Where revenue concentrates & what shifts it: ~95% electricity sales, Guangdong-heavy, increasingly market-traded. A single provincial pricing decision (Guangdong's annual market-tariff outcome) moves the whole P&L. That is enormous single-variable, single-regulator concentration for a "diversified baseload" story.
• Why the moat is weaker than bulls think: the moat protects the franchise but not the return. A licence monopoly is worthless to a minority holder if the licensor (the state) sets the price at cost-recovery-minus. The moat caps ROE at the regulator's chosen number — and that number is drifting down.
• Most dangerous competitor bulls underestimate: not CNNP — renewables + the merchant market itself. Near-zero-marginal-cost solar/wind setting the provincial clearing price is a more lethal threat to CGN's realised tariff than any nuclear rival, because it attacks the price of every traded MWh.
• Worst capital-allocation reality: reinvesting ~RMB 70-100B into new units that (post-VAT-reform) earn lower after-tax returns than the existing fleet, funded on a 133%-geared balance sheet, while paying a possibly-over-earnings dividend. Growth capex at declining incremental returns is value destruction in slow motion, even if strategically mandated.
• Assumptions that must hold for today's price: (1) tariff erosion is cyclical and bottoms soon; (2) volume growth keeps outrunning price decline; (3) the dividend/yield "floor" holds; (4) no sanctions escalation on the CGN group. Break any one and the ~14x multiple is indefensible.
• If growth disappoints 20-30%: if EPS lands at the bear ~0.16 vs Street ~0.21 (−24%), and the multiple re-rates toward the 8.1x renewable-peer average, the H-share has material downside from ~HK$3.50 (SWS DCF fair value is already HK$2.69, ~23% below spot).
• Single scenario that permanently impairs: full nuclear marketisation without a nuclear-specific price floor — if Beijing subjects baseload nuclear to the same must-trade regime as renewables with no CfD-style contract-for-difference backstop, CGN's cash-recovery model breaks structurally. Plausibility: moderate and rising — this is the direction of travel, the only question is pace and whether nuclear gets a carve-out.

Lens 14 · Management Questions (ordered by information value)

1. What share of on-grid generation was sold via market-based trading in FY2025 by province, and what is the blended realised tariff trajectory you are guiding to for FY2026-2028?
2. Post-MOF-Announcement-No.10, how many GW of your under-construction pipeline fall into the 50%-VAT-refund bucket, and what is the estimated per-MWh after-tax margin gap vs the grandfathered fleet?
3. Will baseload nuclear receive a price floor / CfD-style mechanism as power-market reform proceeds, or will it be exposed to the renewables-must-trade regime — and what is your contingency if the latter?
4. What is the consolidated (attributable) vs. managed capacity and generation split, and what attributable-EPS contribution do you expect from each unit commercialising in 2026-2028?
5. Given ~133% net gearing, what is the capex funding plan and target gearing through the construction peak, and at what point does leverage constrain the dividend?
6. Is the FY2025 dividend covered by attributable (not managed) earnings, and what payout policy should minority holders model for 2026-2028?
7. What ROE/ROIC do you underwrite on a new Hualong One unit at today's tariff and VAT treatment, and how does it compare to the fleet average?
8. What is the concrete asset-injection schedule from CGNPC (which plants, what valuation basis, what timing), and how will minority interests be protected on related-party pricing?
9. How exposed is the group to uranium price movements, and what fraction of fuel is secured via CGN Mining / long-term offtake vs. spot?
10. What operational or financial impact does the US Entity-List designation on the CGN group have on technology access, financing, and any Western partnerships?
11. What is your assumed coal-benchmark linkage for regulated units, and how much regulated vs. market exposure will the fleet carry by 2028?
12. How should we think about decommissioning and spent-fuel provisioning adequacy — what is the funded status and discount-rate assumption?
13. What is the expected utilisation-hour trend as more units come online and grid renewable penetration rises (curtailment risk to baseload dispatch priority)?
14. How do you weigh returning capital vs. funding the pipeline given the incremental-return compression — is there a scenario where you slow new-build to protect returns?
15. What is the strategic role of the power-sales / trading companies — is the third-party retail throughput margin-accretive or a hedge, and how large will it get?

---