CGN Power: A Real Nuclear Moat, but No Margin of Safety Yet

That is why CGN Power is best described not as high-quality compounding growth and not as a simple mature cash cow. It is a state-backed nuclear build-out compounder with utility-style current earnings and infrastructure-style deferred value. The closest portrait label is company in transition. The installed base already behaves like a mature cash generator. The balance sheet, capex program, and earnings bridge behave like a company still moving from one fleet size to another. In 2025, attributable net profit was RMB 9.77 billion, but operating cash inflow was RMB 29.97 billion and fixed-asset investment was RMB 35.98 billion. That is not the profile of a completed cash-harvester. It is a fleet operator funding a second life-stage.

That leaves the stock, today, priced on the A-share line at about 22.5 times trailing earnings and roughly a 2.1% indicated yield: not distressed pricing, and not an obvious bargain either. The H-share line remains much cheaper at HKD 2.95; converted at roughly HKD/CNY 0.865 as of July 22, that is about CNY 2.55, implying a very large A/H valuation gap for the same company. But this report keeps the valuation conclusion on the A-share basis, because that is the line under review and the one already tracked in the user’s price feed.

The cleanest present judgment is this: CGN Power owns real assets, a real moat, and a real policy runway. It also sits in the awkward middle ground where the market already recognizes all three. That does not make the stock bad. It makes the next 12 months much more about execution and cadence than about discovering the story for the first time. Investors are not deciding whether China will build more nuclear. They are deciding how much of that future they want to prepay while 2026 earnings are being distorted by outages and marketized tariffs.

Company vertical history

CGN Power exists because China wanted a listed vehicle for a business that had become too large, too capital-intensive, and too strategically important to remain financed only at the parent-group level. The parent, China General Nuclear Power Group, grew out of the Daya Bay project and the broader reform-era push to secure electricity for coastal China through imported nuclear technology, local engineering capability, and eventually domestic standardization. EDF’s account of Daya Bay still captures the origin story: French know-how seeded the first large coastal plants, and the Guangdong base became the template for later expansion. The listed company, formed on March 25, 2014, was therefore not a startup. It was a carve-out of already-operating strategic assets into a market-financed platform.

The first stage was asset crystallization and market listing. The H-share debut in December 2014 brought in external capital at HKD 2.78 per share and made CGN Power the group’s public nuclear-generation platform. What investors bought was not an invention, but a portfolio of long-lived reactors wrapped in a dividend promise and a state-backed growth story. The one feature that distinguished CGN from a normal thermal utility was not higher short-term growth. It was scarcity. Nuclear licenses in China are few, new sites are approved centrally, and the owners able to develop them safely are even fewer. That scarcity has shaped the entire capital-markets narrative since listing.

The second stage was the slow, uneven normalization of China’s post-Fukushima nuclear cycle. The industry had gone through years of caution and approval delays. CGN’s listed business had to operate like a mature utility before the country was ready to treat it like a growth utility again. Through this period, the company’s job was to keep the fleet safe, integrate newer units, deepen technical capability through subsidiaries such as operations, engineering and research entities, and preserve enough balance-sheet flexibility for the next approval wave. The business model remained broadly the same. What changed was the market’s willingness to pay for pipeline value. When approvals were scarce, CGN traded more like an income utility. When approvals resumed, it traded more like an infrastructure compounder.

The third stage began when China’s approval cadence picked up again. Reuters reported six new units were approved in 2023, 11 in 2024, and 10 in 2025. At the group level, CGN said it secured approvals for 16 units during the Fourteenth Five-Year Plan. By the 2025 annual report, the listed company had 20 units under construction, including units managed on behalf of the controlling shareholder. By June 2026, after Huizhou Unit 1 and Cangnan Unit 1 entered commercial operation in April, that construction count had fallen to 18, but the picture was more useful than the headline: three units were in commissioning, two in equipment installation, seven in civil construction, and six in FCD preparation. This is the defining fact of CGN Power’s current era. The company is not just feeding a stable fleet. It is stair-stepping toward a larger one.

Several key nodes changed the company’s shape in ways that still matter.

The first was the 2019 A-share listing. This was not needed to prove the business existed. It was needed to widen funding options for a capital-intensive expansion cycle and to insert the company into mainland valuation regimes, where large strategic SOEs often command different multiples from their H-share lines. The A-share IPO priced at RMB 2.49 and raised about RMB 12.57 billion. The subsequent dual-listing structure still matters today because the company’s A-shares and H-shares trade at a wide discount to each other, which says as much about capital-market segmentation as about fundamentals.

The second was the return of annual nuclear approvals from the State Council. That did not change CGN’s current cash flow immediately, but it changed the duration of the growth story. In infrastructure businesses, future permitted projects are part of today’s valuation, especially when the license to build is scarce. The years 2023 to 2025 therefore mattered more than any single quarter: they moved CGN from “high-quality incumbent with limited near-term growth” to “incumbent with another decade-long asset build cycle in sight.”

The third was the wave of intra-group asset injections and consolidation changes in 2025 and early 2026. CGN Power acquired Taishan Second Nuclear from its parent in January 2025, then in October 2025 acquired 82% of Huizhou Nuclear and full stakes in Huizhou Second Nuclear, Huizhou Third Nuclear and Zhanjiang Nuclear, all from the parent. In January 2026, Ningde Second Nuclear became a subsidiary after a concerted-party agreement with China Datang’s nuclear unit gave CGN effective leadership in shareholder and board votes there. These moves tighten the parent-listed company relationship. They also complicate period-to-period comparisons, because common-control transactions and consolidation scope changes can flatter or muddy trend lines.

That Ningde Second Nuclear agreement is worth being precise about. It was not a generic cooperation memorandum. The filing says China Datang Group Nuclear Power agreed to act in concert with Guangdong Nuclear Power Investment at the shareholders’ meetings and board meetings of Ningde Second Nuclear, enabling CGN’s group to lead the relevant activities there. The agreement took effect on signing and remains in force for the duration of Ningde Second Nuclear. That means the economic significance is governance control over an existing project company, not just joint bidding rhetoric. As of the latest public disclosure reviewed here, there is no evidence of a broader merger, cross-shareholding expansion, or shift beyond that project-level control arrangement.

The fourth node was tariff reform rather than corporate action. In 2013, the NDRC set a national benchmark nuclear on-grid tariff of RMB 0.43/kWh for newly commissioned units after January 1, 2013, with local coal benchmark tariffs or adjusted treatment applying in certain circumstances for demonstration and first-batch third-generation units. Over time, CGN’s realized pricing mix has shifted further toward market-based transactions. By 2025, approved tariffs for operating units were unchanged, but the average market-based tariffs fell by about 8.8%, and market-based volume accounted for 56.2% of on-grid generation. This matters because the historical idea of nuclear as quasi-fixed-price baseload is no longer fully true in the Chinese provincial power market. CGN is still a regulated asset owner. It is increasingly also a dispatch and pricing manager.

The current stage, then, is not simply “maturity.” It is mature operation plus pipeline monetization under more marketized pricing. One side of the company reads like a bond with outages. The other side reads like a rolling construction program with sovereign policy sponsorship. That duality explains why the stock never sits comfortably in a single bucket for long.

Financial vertical review

CGN Power’s long financial arc is easier to understand if revenue is treated as a function of three variables: electricity volume from the operating fleet, the realized settlement price mix between approved and market-based tariffs, and changes in project ownership or consolidation scope. Operating revenue in 2023 was RMB 82.55 billion, essentially flat year on year, while attributable profit rose 7.6% to RMB 10.72 billion because finance costs fell and newly commercial units helped. Profit rose slightly further in 2024, to RMB 10.84 billion. Then in 2025 revenue fell 4.1% to RMB 75.70 billion and attributable profit dropped 9.9% to RMB 9.77 billion, even though on-grid generation still rose 2.36%, because the price side deteriorated: market-based tariffs fell by about 8.8%.

That revenue pattern says something important about the business. Nuclear utilities look volume-driven from the outside, but once plants are already operating at high utilization, price mix often matters more than incremental volume. In 2025, average utilization hours for CGN’s 28 operating units were 7,767, up from 7,710 in 2024, and capacity factor improved to 92.65%. Yet revenue and profit still fell, because more of the fleet’s output faced weaker market-based pricing: the company did its operating job better and still earned less, a structural feature of the current Chinese power-market reform path rather than a CGN-specific failure.

Earnings quality is better than headline net income suggests, though not in the simple way that always means “cheap.” The operating-cash-flow profile has been consistently stronger than attributable net profit: RMB 31.37 billion in 2022, RMB 33.12 billion in 2023, RMB 37.51 billion in 2024, and RMB 29.97 billion in 2025, against attributable profit of roughly RMB 9.96 billion, RMB 10.72 billion, RMB 10.84 billion and RMB 9.77 billion in those years. That gap exists because nuclear plants are heavily depreciating physical assets with long lives. It tells investors that accounting profit understates cash generation from the operating fleet. It does not mean all that cash is distributable, because much of it is being recycled into the build-out pipeline.

The balance sheet is sound for a nuclear utility and demanding for anything else. Total assets rose to RMB 505.66 billion at the end of 2025, total liabilities were RMB 329.44 billion, the asset-liability ratio was 65.2%, and debt to equity was 142.6%. Those numbers would look aggressive in a normal industrial company. In a utility with long-duration project finance, state-backed ownership and multi-decade assets, they are elevated but not abnormal. The more relevant point is that leverage trended up in 2025 as construction intensity and parent-asset acquisitions increased. The company itself said financing inflows rose mainly because of higher external borrowings. This is why the stock cannot be analyzed as a pure yield vehicle. The balance sheet is being used to bridge a capex cycle.

Free cash flow, on an all-in basis, is thin to negative during the build phase. In 2025, operating cash inflow was RMB 29.97 billion and investment cash outflow was RMB 32.34 billion; fixed-asset investment alone was RMB 35.98 billion. Management said those investments were mainly for construction of nuclear generating units, technical improvement at operating plants, and related R&D. The business therefore behaves like two entities consolidated into one statement: a mature fleet that throws off ample operating cash, and a development arm that absorbs it. That accounting shape is why headline FCF screens can misclassify CGN as a weak cash generator when the better description is “internally funding growth of a licensed asset base.”

Returns on capital have softened from a high single-digit base rather than collapsed. Return on equity excluding non-controlling interests was 9.7% in 2023, 8.7% in 2024, and 7.8% in 2025. Return on total assets slipped from 6.4% in 2023 to 5.9% in 2024 and 4.8% in 2025. This is exactly what one would expect when new capital is being deployed ahead of commercial operation and when realized tariffs weaken. The returns are still credible for a regulated infrastructure owner. They are not currently trending like a scarcity business with strong pricing power.

One more caution belongs here. Comparability across the last two years is imperfect because common-control acquisitions have changed consolidation scope, and the Q1 2026 report explicitly states retrospective restatement due to merger under common control. That means any neat trend line from 2024 to 2026 should be read with more care than usual. On the underlying economics, however, the message is clear enough: operating cash remains well ahead of reported profit, accounting profit is currently depressed more by tariff mix and outages than by demand, and all surplus capital continues to be pulled into construction.

Price and valuation history

CGN Power’s capital-market history has moved through three broad valuation identities.

The first was IPO scarcity. The 2014 H-share listing and the 2019 A-share listing both landed into markets where pure-play nuclear exposure was rare. The 2019 A-share deal, priced at RMB 2.49, quickly became one of the year’s largest mainland IPOs. Press coverage at the time framed CGN as the first nuclear operator to achieve both A- and H-share listing, and early trading pushed the company into Shenzhen’s “RMB 200 billion club” within days. That was a scarcity premium more than a revision to near-term earnings power.

The second identity was utility torpor. Once the post-IPO excitement faded, the market often treated CGN less as a strategic growth story and more as an income utility with a low-beta policy overlay. That happened whenever reactor approvals paused, tariff reform increased uncertainty, or outage-heavy quarters reminded investors that “stable baseload” still has very lumpy reported earnings if refuelling schedules bunch together. The company’s 2023 and 2025 disclosures show how sensitive the equity story can be to operational calendars and pricing mix.

The third identity is the current one: strategic low-carbon SOE with a visible new-build runway. The reason the valuation center shifted was not just lower market rates or thematic fashion. China actually restored a high nuclear approval cadence, which made the future fleet larger, not merely more fashionable. Reuters summarized the recent pace as 10 to 11 approvals annually in 2022 through 2025. In that environment, investors stopped valuing CGN only on current reactors and started paying more for reactors not yet in the earnings base.

Today, the A-share line trades at about 22.5 times trailing earnings with an indicated dividend yield near 2.1%, based on quoted market data around July 22–23, 2026. The H-share line trades at HKD 2.95. Even allowing for different liquidity pools, index membership, and investor bases, that remains a striking spread for the same company. The market is effectively saying the mainland line deserves a meaningfully higher multiple than the Hong Kong line. For A-share investors, that is not free alpha. It is a reminder that the domestic valuation already includes strategic-premium thinking.

Business model and moat

CGN Power’s economic engine begins with selling electricity from operating nuclear units and ends with financing, building and eventually moving additional units into that same electricity-sales pool. In 2025, the company’s operating units delivered 232,648 GWh of on-grid power including associates, and 56.2% of that passed through market-based transactions. Revenue is not formally split in the summary results announcement by reactor site, but the business is still dominated by nuclear generation. Construction, design, operations, and technical-service businesses exist largely to support the fleet and the pipeline around it. This makes CGN different from a diversified utility with many unrelated earnings buckets. It is more integrated than diversified.

The cost structure is classic nuclear infrastructure. Fixed costs are enormous: depreciation, plant staffing, safety systems, maintenance planning, refuelling infrastructure, and financing. Variable fuel cost per kWh is low compared with thermal generation, but the business is capital heavy and availability sensitive. When output rises from the same reactor base, margins can improve nicely because the cost base is already sunk. When one or two large units sit in outage at the same time, profit can drop quickly because the cost base remains in place while kWh disappear. Q1 2026 was a near-perfect demonstration. A 10% drop in on-grid generation translated into a 13.25% revenue drop and a 14.86% drop in total profit. That is operating leverage in reverse.

The first real moat is the regulatory and licensing barrier. Nuclear power in China is not a sector where capital alone buys entry. Site approval, safety credentials, project management capability, reactor ecosystem depth, and political trust all matter. The NDRC tariff framework, State Council approval process, and the concentration of ownership among a handful of state groups make this one of the least contestable power-generation segments in the country. A potential challenger does not merely need more money or better software. It needs a place inside the state’s nuclear system. That barrier is still real.

The second moat is scale with repetition. CGN’s operating fleet gave it 28 units at the end of 2025, and the group said its total capacity including construction exceeded 56 million kW during the Fourteenth Five-Year Plan. By June 2026 the listed company was still managing 18 units under construction. Scale matters here not because customers love the brand, but because repeated construction, commissioning, operations and outage management reduce execution risk over time. That advantage becomes more visible when the industry standardizes around families of reactors and repeat site development. It is a scale moat rooted in safety learning curves rather than consumer awareness.

The third moat is capital access. CGN Power is controlled by CGNPC, which is itself controlled by SASAC. That ownership does not eliminate project risk, but it does shape financing capacity, lender confidence, and access to asset injections from the parent. The 2025 acquisitions of Taishan and Huizhou-related entities from the parent are examples. In a business where construction lead times are measured in years and commercial value often arrives long after cash outlay begins, access to stable funding is part of the moat. The trade-off is a governance discount: minority investors must accept that the listed vehicle is part of a wider state group and will continue to transact with it.

The fourth moat is customer stickiness at the system level, not the retail level. Nuclear is baseload power. Grid operators and provincial systems value it for reliability, carbon intensity, and fuel-security characteristics. Individual end users can switch suppliers in a marketized setting, but the system still needs the asset class. That is different from saying CGN has consumer pricing power. It does not. What it has is system relevance. That relevance protects dispatch quality and policy support more than it protects spot pricing.

Management and governance are typical of a large Chinese central SOE, which is both comfort and constraint. Yang Changli is the group chairman; Pang Songtao is the company’s president and executive director. Official biographies show both are long-time nuclear-industry executives rather than market-driven outsiders. That is appropriate for plant safety and capital discipline in a nuclear operator. It is less useful if investors expect aggressive portfolio pruning, buybacks, or a Western-style obsession with equity returns. The capital allocation record supports this reading: CGN pays dividends consistently, with a 2025 final dividend of RMB 0.086 per share and a payout ratio of about 44.47%, but it has prioritized construction and parent-asset integration over share repurchases.

On governance quality, there is no obvious red flag in the latest filings: the 2025 annual results carried an unqualified audit opinion from KPMG Huazhen, and the company says it complied with the Hong Kong code provisions reviewed in the announcement. The real governance issue is not fraud. It is the standing reality of state control, related-party transactions, and minority investors sitting lower in the priority stack than national build-out goals. That deserves a discount, though in this industry it is also part of the company’s privilege set.

Industry and cycle

China’s nuclear industry is growing again, but it is growing in a very Chinese way: centrally approved, strategically justified, and financially distributed across a small number of state-backed operators. The 2025 NEA briefing said China’s operating and approved-under-construction nuclear capacity had surpassed 120 GW after the 2025 approvals. The China Nuclear Energy Association’s 2025 and 2026 materials point to nuclear taking a larger role in decarbonization and energy security. This is not an early-stage industry in technical terms. It is a re-accelerating industry in policy and capital-allocation terms.

The profit pool sits mainly with licensed plant owners and operators, not with end-market retailers. Upstream fuel and engineering matter, but the durable cash flow is in operating reactors connected to the grid under long-life asset economics. That is why CGN Power and China National Nuclear Power are the key listed references. Suppliers can benefit from the build cycle, but the economic rent of the finished asset belongs to the operator.

This is a policy cycle first, an operational cycle second, and only then a macro cycle. Demand for electricity in the provinces matters. Interest rates matter because the assets are debt financed. But the most important variable is still approval cadence by the State Council and the degree to which market reform pushes nuclear from approved tariffs into competitive transactions. The 2013 NDRC notice set a stable national benchmark framework for new units, while later reforms kept expanding the role of market-based power trading. In 2025, CGN said approved tariffs for operating units were unchanged but market-based tariffs declined materially. That combination is the new normal: industrial policy remains supportive of building nuclear, while power-market reform puts more pressure on realized pricing.

Geopolitics matters less to CGN’s daily dispatch than to technology sourcing and valuation sentiment. Domestic localization has advanced enough that China can keep building large parts of its nuclear program without depending on imported platforms in the old way, but foreign technology and export restrictions still shape the sector’s outer boundary. More important for equity investors is that CGN, as a state nuclear operator, will probably never receive a fully “market” valuation in international portfolios. The same state identity that secures financing and approvals also limits the investor base. That is part of why the H-share line trades so much cheaper than the A-share line.

Horizontal competitor analysis

There is one true listed domestic peer: China National Nuclear Power. China’s nuclear-generation industry is too concentrated for a broad peer set of pure plays. That makes this a “few competitors” case, with China National Nuclear Power as the closest like-for-like comparison and a small set of indirect references for valuation discipline.

China National Nuclear Power is similar in the ways that matter most: it is a state-controlled nuclear operator too, it benefits from the same approval recovery, and it trades as a strategic low-carbon utility. Its A-share line was around CNY 9.06 on July 23 with a market cap around CNY 184.5 billion to CNY 186.6 billion, roughly the same equity size as CGN Power’s A-share line. Google Finance shows a similar trailing P/E, around 22.6 times, and a similar indicated dividend yield near 2.0%. That near-parity matters: the market is not giving CGN a dramatic premium or discount to its one true domestic comparable.

But the businesses are not identical. China National Nuclear Power has broader clean-energy exposure beyond nuclear, which can make its quarterly revenue path look less hostage to outage concentration at a few specific large bases. Reuters’ company profile describes it as engaged in clean energy projects including nuclear, wind and solar. CGN Power is closer to a purer nuclear bet. That purity is a strength if investors want direct exposure to the nuclear approval cycle. It is a weakness if investors want smoother portfolio effects from renewables and cleaner diversification at the listed-entity level.

There is also a difference in market narrative. CGN’s story is tied more tightly to the CGN ecosystem, its parent injection path, its coastal fleet, and the A/H valuation gap. CNNP’s story is more often told as the CNNC-backed national nuclear platform with a broader portfolio. Customers do not choose between the two in a consumer sense. Provinces and the State Council allocate projects through the state system. So the competition is less about stealing retail electric customers and more about who gets which approved sites, what technology packages are used, and how much investor premium attaches to one portfolio rather than another. In that contest, neither company has decisive public-market superiority today.

Indirect references are useful for discipline. China Yangtze Power is not a nuclear operator, but it is a helpful low-risk benchmark for how the A-share market prices a large-scale, state-backed, low-carbon utility with very high asset quality. If CGN ever traded at a significant premium to names like Yangtze purely on “green” narrative while outages and pricing pressure persisted, that would be a warning sign. CLP Holdings is not a domestic nuclear peer either, but it matters because Hong Kong’s power system and legacy Daya Bay arrangement provide context for the stability and system value of baseload nuclear supply. Sister companies like CGN New Energy and CGN Mining show how the market values the group’s adjacent platforms: renewables and upstream uranium are given very different multiples and investor bases from listed nuclear generation.

Peer snapshot

The market’s current message in those numbers is restrained. CGN is not being given an obvious premium to CNNP for its pipeline, nor an obvious discount for its A/H complexity. That makes the stock harder to call cheap than enthusiasts of the long-term nuclear story often assume. The argument for CGN has to come from execution and future earnings realization, not from a glaring peer mispricing.

Ecologically, CGN occupies the position of a licensed incumbent with a still-expanding fleet. It is not a nimble challenger. It is one of the two system owners. That means its position gets stronger if the main industry tension is “who can build safely at scale” and weaker if the main industry tension becomes “how much of nuclear’s output must clear through increasingly competitive provincial markets.” The former is its home court. The latter is survivable, but it narrows the valuation premium that scarcity alone can justify.

Current fundamentals and bull-bear divergence

The last four reported quarters have produced a simple headline and a more nuanced underlying picture. The headline is that 2025 weakened and Q1 2026 weakened further. The nuance is that much of the 2026 weakness came from outage timing rather than demand deterioration, and Q2 appears to have recovered. Attributable profit in 2025 fell 9.9% to RMB 9.77 billion on weaker realized tariffs. Then Q1 2026 revenue dropped another 13.25% year on year and attributable profit fell 9.33%, with total profit down 14.86%. Yet by the half-year operating briefing, the generation decline had narrowed sharply from the Q1 rate.

The market is therefore trading two things at once. The first is real fundamentals: whether Huizhou Unit 1 and Cangnan Unit 1 entering commercial operation in April, plus the continuing progress of Huizhou Unit 2 and other projects, can offset the drag from outages and price weakness. The second is narrative: whether investors continue to pay a strategic nuclear premium because China has clearly committed to keeping approvals high. NEA’s April 2025 briefing and Reuters’ approval reporting support the second narrative. The Q1 2026 report and H1 generation briefing caution that the first one can still go wrong quarter to quarter.

The bull case rests on five pieces of evidence. China’s approval cadence is real, not speculative, and CGN’s share of that pipeline is large. The company also still has one of the deepest new-build queues in the sector, with 18 units under construction as of June 2026. 2026 Q1 weakness already appears partly reversed in Q2 generation, and operating cash generation remains much stronger than accounting income, giving the company more self-funding capacity than headline P/E suggests. Parent support and asset-injection optionality remain intact.

The bear case also rests on hard evidence. Realized pricing has become less regulated and more exposed to market-based transactions, and 2025’s 8.8% drop in average market-based tariffs was severe enough to overpower higher generation. Outage clustering can make near-term earnings highly volatile even in an otherwise stable asset class; Taishan’s Q1 2026 weakness was dramatic. Construction intensity keeps leverage elevated and suppresses all-in free cash flow. The A-share market is already valuing the company close to the domestic peer multiple too, so there is little room for a simple “nuclear rerating” without new evidence. And the H-share discount is a reminder that not all investor pools agree with the mainland strategic premium.

Valuation analysis

Historically, CGN’s valuation center has depended less on growth rate alone than on what investors thought they were buying. When the market saw a safe dividend utility, the multiple compressed. When it saw a scarce strategic beneficiary of a revived national build cycle, the multiple expanded. Today’s A-share valuation of about 22.5 times trailing earnings and about a 2.1% yield suggests the market is already pricing more than a mature utility and less than an aggressive infrastructure compounder. That is why absolute valuation matters more here than peer-relative arguments.

Cash-flow passthrough

Over the last four reported full years available in the primary disclosures used here, operating cash flow has run at roughly three times attributable net income: about 3.15 times in 2022, 3.09 times in 2023, 3.46 times in 2024, and 3.07 times in 2025. This is not evidence of aggressive accounting. It is mostly nuclear utility economics: heavy non-cash depreciation on an established operating fleet.

The harder step is separating maintenance capex from growth capex. The company does not disclose the split directly. My working assumption is that only about RMB 8 billion to RMB 10 billion of annual spending represents maintenance and technical-improvement capex required to sustain the current fleet, while the majority of the 2025 fixed-asset investment and investment cash outflow was growth capex tied to the under-construction pipeline. That assumption is grounded in management’s statement that 2025 investment was mainly for construction of nuclear generating units, technical improvement of operating NPPs, and related R&D, plus the fact that the group still had 20 units under construction at year-end 2025. On that basis, owner earnings are far closer to operating cash flow minus maintenance capex than to reported net profit.

Using 2025 operating cash inflow of RMB 29.97 billion and assumed maintenance capex of RMB 9 billion, owner earnings are roughly RMB 21 billion. Against the quoted A-share market cap of about RMB 187 billion, the equity is trading on an owner-earnings yield of about 11%, or an owner-earnings multiple just under 9 times. That is very different from the headline P/E of 22.5 times. Because the gap is comfortably above 30%, the more relevant valuation anchor for CGN is owner earnings and medium-term asset monetization, not raw accounting profit. This does not automatically make the stock cheap, because growth capex still has execution risk. It does mean the headline P/E somewhat overstates valuation richness for investors willing to underwrite the pipeline.

Peer valuation

Relative to CNNP, CGN looks fairly valued rather than mispriced. Both trade around 22 to 23 times trailing earnings, both yield around 2%, and both sit around the same A-share market-cap scale. The H-share line, by contrast, is much cheaper, implying that the domestic A-share multiple includes a clear local-market premium. I do not think the peer comparison alone justifies calling the A-share cheap. It only says the stock is not obviously expensive relative to its one true domestic comparable.

Absolute valuation

For a company like CGN, the best toolkit is a blend of attributable earnings, owner earnings, and medium-term multiple normalization. I do not use a long-horizon DCF as the primary method because unit timing, tariff evolution, construction pacing, and consolidation changes produce a false sense of precision.

Valuation scenarios

This is valuation-scenario analysis within a research framework, not investment advice.

Expectation gap

The market is currently pricing a middle path: China keeps building nuclear, CGN keeps winning and advancing projects, and 2026’s weak quarter does not become a weak multi-year trend. What the market may still be misjudging is how long it takes large approved pipelines to become distributable cash. The next few result cycles matter most on three metrics: realized tariff mix, outage normalization, and milestone progress at Huizhou, Cangnan, Lufeng and Ningde-related projects. If those improve together, the stock can hold a strategic premium. If only approvals remain strong while current earnings stay soft, the multiple is vulnerable.

Margin-of-safety recheck

At the current A-share price, there is not a large margin of safety against the conservative scenario. The conservative fair-value band centers around RMB 4.3. A genuine value-style entry would require a discount to that number, not a price already hovering near it. The most fragile assumption in the base case is not demand. It is the combination of utilization recovery and tariff stabilization. If that paired assumption is cut to 70%, the base-case value falls back into the high-RMB 3s. On a flat-earnings, three-year view with the stock bought around current levels, total return would likely rely mainly on a roughly 2% dividend yield plus modest earnings drift, which is not a generous cushion for a construction-heavy utility. The verdict is plain: margin-of-safety sufficiency is not obvious.

Risk analysis

The biggest business risk is not a collapse in electricity demand. It is a prolonged mismatch between asset availability and realized pricing. Probability is medium, impact high. Nuclear units can miss a quarter or a half-year through longer refuelling or reduced-load operation without changing long-term asset value much, but when that happens at the same time as weaker market-based tariffs, the earnings hit is magnified and investors start questioning the quality of the entire growth story. The transmission path runs from lower generation and lower settlement price to lower earnings, then to lower confidence in project payback, then to multiple compression. Q1 2026 already showed the first half of that chain.

The second risk is capital-intensity risk. Probability is high, impact medium to high. CGN had 18 units under construction as of June 2026 and asset-liability ratio of 65.2% at end-2025. In a stable funding environment that is manageable. If financing conditions tighten, project timing slips, or costs rise, equity holders feel it first through slower dividend growth, weaker free cash generation, and a lower willingness by the market to capitalize distant earnings. The observation point here is not just net debt. It is the pace of financing inflows relative to investment cash outflows and fixed-asset investment.

The third risk is valuation risk from the A/H gap and domestic premium. Probability is medium, impact medium. The H-share line implies the same company is worth much less to Hong Kong investors than to mainland investors. That gap can persist for a long time, but it is still a warning. If domestic enthusiasm for strategic SOEs cools, the A-share line has more room to derate than a peer comparison alone would suggest. The trigger would not need to be bad operating news. A simple rotation out of policy-favored utilities could do it.

The fourth risk is governance by policy objective. Probability is medium, impact medium. State control is part of the company’s advantage set, but it also means minority investors do not control the timing of parent-asset injections, the balance between leverage and equity financing, or the payout-versus-construction trade-off. The 2025 acquisitions from the parent illustrate both the benefit and the constraint. A shareholder should not assume capital allocation will optimize near-term per-share returns when wider group and national objectives are in play.

The fifth risk is policy reform on pricing rather than policy support for building. Probability is medium, impact high over time. Investors often treat “policy support for nuclear” as if it automatically means supporting returns for listed nuclear operators. The 2025 data disproves that shortcut. Approvals were strong, approved tariffs were unchanged, and yet market-based tariffs fell sharply enough to drag earnings down. The risk is not that Beijing abandons nuclear. It is that the power-market architecture keeps asking nuclear operators to absorb more price competition than equity holders expect.

Catalysts and tracking indicators

Positive catalysts are straightforward. Better-than-feared 2026 interim results would help because the market needs confirmation that Q1 was mostly timing noise. Further commissioning progress at Huizhou, Cangnan, Lufeng and Ningde-related projects would underline that the medium-term earnings bridge is intact. Any evidence that market-based tariff pressure has stabilized would matter more than another generic policy statement about nuclear support, because the market already assumes support for build-out.

Negative catalysts are equally clear. A weak 2026 interim print with no visible pricing improvement would make investors more skeptical that the Q2 rebound was durable. Another cluster of long outages at major stations such as Taishan or Ling’ao would revive the debate about how “stable” the fleet really is from an earnings perspective. Slower construction milestones or higher financing dependence would remind the market that the pipeline is valuable only if it converts on time.

Tracking dashboard

Why these matter is simple. CGN is a capacity-conversion story disguised as a utility. Investors should watch whether current reactors generate predictably, whether the pricing mix erodes, and whether construction keeps marching toward commercial operation. The expected next earnings date most often cited by market calendars is August 26, 2026, which is consistent with the company’s usual late-August interim timetable, though investors should still verify the formal company announcement when released.

Cross-synthesis summary

Vertically, CGN Power has proven one capability above all others: it can take a politically strategic, technically complex, capital-intensive asset class and run it at scale inside China’s state electricity system. That sounds bland until one remembers what nuclear operations require: safety discipline, financing stamina, standardized engineering, periodic outage execution, and the ability to keep regulators, lenders and grid operators confident for decades. The company’s past success did not come from luck, and it did not come from consumer branding. It came from being one of the very few institutions allowed and able to build this kind of fleet. That remains true today.

The problem for equity investors is that a proven industrial capability is not the same as a cheap stock. What made CGN successful historically was a combination of state sponsorship, scarce licenses, engineering repetition and access to long-term capital. Those success factors are still present. What has changed is the way returns are transmitted to listed shareholders. Market-based transactions now account for more than half of on-grid generation, average market-based tariffs fell sharply in 2025, and the company remains deep inside another heavy capex phase. That shifts the investment question away from “is this a privileged business?” and toward “how much of that privilege is already capitalized in the A-share price?”

Horizontally, CGN’s real advantage over peers is not a radically cheaper cost curve or a superior retail channel. It is the combination of a very large existing fleet and one of the deepest visible pipelines tied to recent national approvals. Against China National Nuclear Power, that advantage is meaningful but not overwhelming. The market recognizes this and values the two names similarly. Against lower-risk low-carbon utilities, CGN offers more growth optionality but also more execution and pricing complexity. Against its own H-share line, the A-share already reflects more optimism. That is the heart of the current pricing problem: the company is strong, but the stock is no longer neglected.

I think the market is most likely misjudging the shape of earnings, not the direction. It is right to believe China will keep building nuclear and wrong to treat that truth as if it guarantees smooth yearly earnings progression for the listed operator. Q1 2026’s drop and the implied Q2 rebound make this plain. CGN’s medium-term story is intact, but its short-term reported numbers will stay lumpy because outages, project commissioning, and provincial pricing rules can move several billion renminbi of earnings across quarters. Investors paying a premium multiple for a “stable utility” may be underestimating that lumpiness. Investors writing off the company because of one soft quarter may be missing the embedded pipeline.

For the next year, the critical variables are outage normalization, tariff stabilization, and whether interim results validate the Q2 recovery implied by the operating briefing. For the next three years, what matters most is commercial operation timing at Huizhou, Cangnan, Lufeng, Ningde-related and other approved projects, along with the balance between debt-funded growth and distributable cash flow. For the next five years, the decisive question is whether CGN exits the current build cycle with a larger, higher-cash-yielding fleet before the market fully discounts that future, or whether the market continues to capitalize the pipeline years ahead of cash arrival.

CGN Power would become a better investment than it is now along two routes. One is execution-led: the company starts to show that new units can enter operation on time while pricing pressure eases and owner earnings per share rise faster than reported net income. The other is price-led: the A-share falls into a true margin-of-safety zone while the project pipeline and approval backdrop remain intact. The judgment should be re-examined if market-based tariffs continue to erode, if the construction queue starts slipping across multiple bases, or if the parent-listed relationship becomes more extractive than accretive for minority holders. Any of those deserves a lower multiple even if the national nuclear story stays strong.

Bull and bear reasons

• China’s nuclear approval cycle is no longer hypothetical: China approved 10 to 11 units annually in 2022–2025, and CGN says its group secured approvals for 16 units during the Fourteenth Five-Year Plan.

• The development pipeline is unusually large for a listed utility: CGN still managed 18 units under construction as of June 30, 2026 even after two units entered operation in April.

• Operating cash generation is much stronger than accounting profit, which supports growth funding and means headline P/E overstates the economic multiple of the operating fleet.

• Q1 2026 weakness looks at least partly timing-related rather than structurally deteriorating, because the implied Q2 generation and on-grid output recovered year on year.

• Parent support and project-level control increased through 2025–2026 acquisitions and the Ningde Second Nuclear concerted-party agreement.

• More than half of on-grid power is now market-based, and average market-based tariffs fell about 8.8% in 2025, showing that policy support for building does not guarantee price protection.

• Near-term earnings remain outage-sensitive, and Q1 2026 showed that a handful of longer refuelling outages can cut generation and profit sharply.

• The balance sheet is appropriate for a nuclear builder but still leveraged, with 65.2% asset-liability ratio and continuing heavy investment cash outflows.

• The A-share already trades near the same multiple as China National Nuclear Power, so the easy relative-value argument is weak.

• The H-share line trades far below the A-share on an implied CNY basis, which suggests the mainland strategic premium may already be generous.

Pre-mortem

The most plausible 50% drawdown script is not a nuclear accident scenario; that is too extreme and not the base way listed equity loses half its value. A more realistic script is this: through 2027 and 2028, market-based tariff pressure continues, several key new units slip by one to two years, and another round of heavy refuelling bunching hits major stations. At the same time, the market stops capitalizing distant approved capacity at a strategic premium. Attributable EPS stalls around current levels instead of compounding, and the A-share multiple compresses from about 22 times to about 12 to 14 times. A stock around CNY 4.1 could then re-rate toward the low CNY 2s.

A second script is a balance-sheet-and-sentiment version. Construction remains on track technically, but financing needs stay high because more projects move from approval into concrete and equipment phases before enough of the current queue begins commercial operation. Dividend growth stalls, owner earnings per share fail to lift visibly, and domestic investors rotate out of policy-favored utilities. The market begins to price CGN more like a heavily capexed utility and less like a strategic compounder. In that case, even without major earnings disappointment, the A/H gap could narrow through A-share de-rating rather than H-share catch-up.

Final research conclusion

CGN Power is a real strategic asset owner with a real moat, a real cash engine, and a real multi-year expansion runway. The question is not whether the company matters. It does. The question is whether the A-share line offers enough valuation slack for investors to absorb the three things that make the next few years messy: more market-based pricing, persistent outage volatility in reported quarters, and a construction program that keeps swallowing cash before it turns into earnings.

At the current A-share price, I think the answer is only partly yes. The stock is not expensive in the sense of fantasy pricing, and it is certainly not a broken business. But it is also not offering a classic value entry. The operating fleet and pipeline justify ownership interest; the current quotation does not yet offer the buffer I would want for a capital-intensive utility whose quarterly cadence can still surprise on the downside. What would change my mind? Either clearer evidence that the 2026 soft patch is passing and new units are converting on schedule into per-share earnings, or a lower entry price that moves the stock decisively into a margin-of-safety zone.

【Company-profile scores】

• Fundamental quality: high
• Growth: medium
• Moat: strong
• Financial soundness: medium
• Management credibility: medium
• Valuation attractiveness: medium
• Risk level: medium
• Suitable investor type: long-term growth / value / dividend

【Investment rating】

• Rating: Hold
• One-line thesis: Scarce nuclear assets and a large build pipeline are real strengths, but the A-share already prices much of that while tariff and outage risk remain real.
• 【Ideal Buy Price】3.3–3.6 CNY Basis: at least a 20% discount to the conservative fair-value range around RMB 4.2–4.4.
• Acceptable hold price: 3.8–4.7 CNY
• Clearly overvalued price: 5.6 CNY and above
• Current-price classification: acceptable hold
• Whether to wait for a better price: yes. A buy is more attractive below about CNY 3.6, provided the approval backdrop and project milestones remain intact. The opportunity cost of waiting is giving up a modest dividend and some upside if execution improves quickly.
• Target holding horizon: 3–5 years
• Expected annualized return: conservative 3%–5%; base 7%–10%; optimistic 12%–15%
• Max-loss risk: about 40%–50% in the pre-mortem case, triggered by prolonged tariff pressure, project slippage, and multiple compression toward a plain-utility valuation
• Reassessment-trigger signals: if market-based tariff realization deteriorates sharply again; if multiple major projects slip beyond current milestone windows; if operating cash flow weakens materially relative to profit; if leverage rises further without corresponding commissioning progress; if interim or annual filings show parent-related transactions becoming clearly less accretive to minority holders

【Valuation Range】

• current: 4.11 (close as of 2026-07-22)
• bear (conservative · ideal buy zone): [3.3, 3.6]
• base (fair · acceptable hold zone): [3.8, 4.7]
• bull (optimistic · above the clearly-overvalued line): [5.0, 5.6]

Key data tables

Operating and financial snapshot

† 2024 on-grid generation is derived from 2025’s disclosed 2.36% year-on-year increase.

Pipeline and operating scale

The pipeline table is the single most important bridge between today’s earnings and the medium-term story. CGN’s valuation cannot be judged only on the current fleet, because too much capital and too much regulatory permission already sit in assets that have not yet reached commercial contribution.

Research uncertainties

The largest uncertainty is the maintenance-versus-growth capex split. The company discloses aggregate investment and fixed-asset spending, but not the exact amount required merely to sustain the current fleet. That means owner-earnings estimates are necessarily judgment-based.

The second uncertainty is pricing granularity. The company discloses the direction of average market-based tariffs and publishes approved tariff tables, but it does not give a simple site-by-site realized blended tariff bridge in the summary materials used here. That limits precision when modeling earnings sensitivity by province.

The third uncertainty is how far the parent-asset injection path will continue. Recent acquisitions and the Ningde Second Nuclear control arrangement show the listed company can be expanded through related-party deals, but public disclosures do not provide a full medium-term map of what will definitely be injected and when.

The fourth uncertainty is near-term earnings-date certainty. Market calendars widely point to August 26, 2026 for the next results, but the definitive company announcement should still be checked when published.

Sources

Primary sources used most heavily in this report were CGN Power’s 2025 annual results announcement, 2026 first-quarter report, and 2026 second-quarter operating briefing filed through HKEX; CGN/CGNP investor-relations pages; the NDRC notice on the nuclear on-grid tariff mechanism; the National Energy Administration briefing on 2025 nuclear project approvals; and Shanghai Stock Exchange or company-reference disclosures for China National Nuclear Power. Market-reference quotes came from Google Finance, Yahoo Finance, Reuters, and other market-data pages cited inline.

Other tickers mentioned

• 601985.SHG: China National Nuclear Power, the only close listed domestic peer for valuation and positioning
• 600900.SHG: China Yangtze Power, an indirect low-risk utility benchmark for comparing public-market expectations
• 00002.HK: CLP Holdings, relevant as a Hong Kong utility reference and historical Daya Bay offtake context
• 01811.HK: CGN New Energy, a sister listed platform showing how the market prices CGN’s renewable assets differently from nuclear generation
• 01164.HK: CGN Mining, a sister upstream nuclear-fuel platform relevant to the broader CGN ecosystem

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.