China Yangtze Power: The Company That Owns the World's Biggest Dam

I. Introduction & Episode Roadmap

Stand on the observation deck at Sandouping, in Hubei province, and the thing in front of you does not read as a building. It reads as geology. The 三峡大坝 Three Gorges Dam is 2,335 metres of concrete wedged across the 长江 Yangtze River, holding back a reservoir that stretches roughly 600 kilometres upstream toward Chongqing. When the sluice gates open in flood season, the water leaving the spillway carries enough kinetic energy to be felt in your chest from a kilometre away.

Now here is the part that should interest an investor rather than a tourist. Every kilowatt-hour that comes out of that dam is sold by a single company listed on the Shanghai Stock Exchange. Not leased. Not licensed on a renewable concession. Owned outright, along with five more of the largest hydroelectric stations ever constructed, all sitting on the same river, all dispatched as one machine.

That company is 中国长江电力 China Yangtze Power Co., Ltd. — CYPC, ticker 600900 in Shanghai. As of the end of 2025 it controlled 71.795 million kilowatts of hydropower capacity, of which 71.695 GW sits inside China and accounts for roughly 16% of the entire country's installed hydro base.¹ In 2025 it generated 309.735 billion kilowatt-hours of electricity, booked revenue of RMB 86.24 billion and net profit attributable to shareholders of RMB 34.50 billion.¹ On August 19, 2026 the market valued it at about RMB 695 billion — around 19 times trailing earnings, with a dividend yield near 3.6%.²

The tension that makes this a story rather than a fact sheet

There is a version of this narrative that writes itself: unrepeatable asset, zero fuel cost, hundred-year concrete, government protection, dividends forever. That version is mostly true and almost entirely useless, because it describes the past twenty years and says nothing about the next ten.

The more honest framing is this. For two decades, CYPC was not really an operating company at all. It was a distribution mechanism — a listed vehicle into which its state parent, 中国长江三峡集团 China Three Gorges Corporation (CTG), periodically dropped finished, de-risked, world-scale dams in exchange for cash and shares. Minority shareholders got handed proven cash flows; the parent got recycled capital to build the next dam. It worked beautifully, for both sides, four times over.

And then it stopped. Not because anyone made a mistake, but because the river ran out. There are no more finished, mainstem, giant-scale Yangtze dams left for CTG to sell down. The growth engine that reliably produced step-changes in earnings since 2003 has completed its final cycle, and what remains is an operating business whose reported profit is driven more by rainfall than by management.

So the questions this episode actually has to answer are uncomfortable ones. What does a company do when its entire historical growth mechanism is structurally exhausted? Is a payout commitment of 70% of profits credible when leverage is elevated and a new, less-proven capex cycle is beginning? And what happens to a bond-proxy valuation when the regulator starts moving the very tariffs that made it bond-like?

The roadmap

We will move quickly through the origin — a 2002 corporate shell holding one aging dam — then slow down for the four asset injections that built the company, the Peruvian acquisition that was CYPC's first real test in an open market, the parent's failed European takeover that tells us something uncomfortable about state-owned discipline under pressure, the management now steering it, and finally the moat, the risks, and what breaks the case. Let's start where the concrete does.

↑ Back to Top

II. Born From a Dam: Founding and the 2003 IPO

In the autumn of 2002, the Three Gorges project was still a construction site. The first generating units would not spin until 2003, the full 18-turbine right-bank station not until 2008. The state entity building it — then the China Three Gorges Project Corporation — was doing what every mega-project developer eventually does. It was running out of money.

The engineering was going well. The financing was the problem. Three Gorges had consumed the better part of two decades and enormous quantities of capital, resettled well over a million people, and absorbed political capital in proportion. What the project needed was a way to convert future electricity revenue into present-day construction funding, without waiting for the last turbine to be bolted down.

The answer was a corporate structure that, in hindsight, looks less like Chinese state planning and more like a leveraged sponsor's playbook.

The seed asset

On November 4, 2002, China Yangtze Power was registered with paid-in capital of RMB 5.53 billion.³ CTG seeded it not with Three Gorges — which was unfinished and unfinanceable — but with 葛洲坝 Gezhouba, the older, smaller run-of-river station downstream that had been generating since the 1980s. Joining as co-founders were a roster of state energy names: 华能国际 Huaneng Power International, 中国核工业集团 China National Nuclear Corporation, 中石油 PetroChina, and 葛洲坝集团 Gezhouba Group.³

The design logic matters more than the cast list. Gezhouba was boring, small, and — critically — finished. It threw off predictable cash. A public market could price it. Three Gorges, by contrast, was a construction risk wrapped in a resettlement controversy wrapped in a schedule.

So the state did the obvious thing: it split them. The listed company would hold the completed, cash-generating asset. The unlisted parent would carry the construction risk, the cost overruns, the political exposure, and the debt. And a promise was attached — as Three Gorges units were commissioned and proven, they would be injected into the listed company.

October 2003

The IPO priced on October 28, 2003: 2.326 billion ordinary shares at RMB 4.30 apiece, raising net proceeds of RMB 9.826 billion. The public tranche began trading in Shanghai on November 18, 2003.³ For its time this was a very large domestic offering, and the stated use of proceeds was blunt — the money would go toward buying Three Gorges generating units from the parent.³

Read that carefully, because it establishes the entire architecture of the company. The IPO was not raising money to build anything. It was raising money to buy something the parent had already built. From day one, CYPC's job was to be the balance sheet that monetised CTG's engineering.

What it means, structurally

There is a genuinely elegant risk transfer here, and it deserves to be named rather than assumed. Public shareholders in most infrastructure stories are asked to fund construction and absorb schedule risk — the ten-year gap between first concrete and first cash flow, during which anything can go wrong and frequently does. CYPC's shareholders were systematically spared that. They bought assets after the turbines were spinning and the output was contracted.

The flip side is equally important and much less flattering. A listed company whose growth comes entirely from buying assets from its own controlling shareholder has, by construction, no arm's-length price discovery on the single most important decision it makes. The seller sets the queue, the timing, and — subject to appraisal rules and regulatory review — meaningfully influences the price. That is a governance question that never went away, and we will return to it repeatedly.

For the first six years of its listed life, though, CYPC was a modest business: one old dam, some minority stakes, and a standing IOU from its parent. The IOU came due in 2009.

↑ Back to Top

III. Becoming the River: The 2009 Full Listing and the 2015–2016 Cascade Buy-In

By 2008 the promise had become awkward. On October 30, 2008, the last of the right-bank units came online, taking Three Gorges to its designed 18,200 MW.³ The world's largest power station was fully operational — and it was sitting inside the unlisted parent while a listed company with the word "Yangtze" in its name owned a 1980s dam downstream.

May 2009: the whole-listing

On May 15, 2009 the CYPC board approved the major asset restructuring plan that market participants called 整体上市 — the whole listing.³ What moved across was the core of the machine: the eighteen 700 MW Three Gorges units, the associated generating facilities, and six auxiliary specialist companies.

The consideration was roughly RMB 107.5 billion, and the funding structure is worth understanding because it set the template for everything that followed. CYPC assumed about RMB 50 billion of debt attached to the assets, issued roughly 1.552 billion new shares to the parent at RMB 12.89 per share for about RMB 20 billion, and paid the balance of roughly RMB 37.5 billion in cash.⁴ At the time it was the largest restructuring A-share investors had seen.⁴

Notice the shape of it. Half the price was paid in debt assumption — meaning CYPC took on leverage against assets that immediately began generating cash to service it. A fifth was paid in stock, which kept the parent's ownership high and its interests aligned. The rest was cash, raised from a balance sheet that could now borrow against the largest hydro station on Earth.

This is the moment CYPC stopped being a utility and became the utility. But it also created a problem the company would spend the next fifteen years managing: a single asset, on a single river, with a single revenue driver that no one controls.

The upriver logic

Which brings us to the most under-appreciated idea in this entire story: a cascade is worth more than the sum of its dams.

Here is the intuition, stripped of engineering. Imagine a staircase of bathtubs down a mountainside, each draining into the next. If you only own the bottom tub, you take whatever water arrives, whenever it arrives. If a storm dumps water faster than you can pass it through your turbines, you spill it — and spilled water is revenue thrown away. If a dry month arrives, you have nothing.

If you own the whole staircase, the calculus inverts. You can hold water in the upper tubs when the lower ones are full, release it when they have room, and shift generation from moments when power is abundant and cheap into moments when it is scarce and valuable. The same rainfall produces more electricity and better-timed electricity.

CYPC could not build that staircase. But its parent already had.

November 2015 to March 2016

In November 2015 the plan surfaced publicly: 溪洛渡 Xiluodu (13,860 MW) and 向家坝 Xiangjiaba (7,750 MW), the second- and third-largest hydro stations in China at the time, sitting on the lower 金沙江 Jinsha River directly upstream of Three Gorges, would be injected into CYPC.⁵ The vehicle holding them was 川云公司 Chuanyun Company, and the price for 100% of the equity was RMB 79.735 billion, paid through a combination of new shares and cash.⁵

The transaction roughly doubled CYPC's installed base in one step. More importantly, it gave the company four linked stations across the lower Jinsha and the Yangtze mainstem, dispatched jointly.⁵

Was the price fair? The question nobody quite answers

Here is where an independent reading has to depart from the celebratory one. These were related-party transactions between a controlling shareholder and its listed subsidiary. Chinese regulation handles this through mandatory independent appraisal, minority-shareholder voting with the interested party recused, and exchange-level scrutiny. Those are real protections. They are not the same as competitive bidding.

The structural bias runs in a specific direction. When a state parent sells a de-risked asset to its own listco, the parent is both seller and, through its majority stake, the largest owner of the buyer. It is internalising a large share of any overpayment. That is a genuine alignment mechanism — but it is alignment, not arm's length. And the appraisals used are asset-based and income-based valuations produced by firms engaged in the transaction, not market clearing prices.

What can be said with evidence: shareholders who held through both injections received assets that were finished, commissioned, and already selling power under established grid arrangements. They did not fund a single cubic metre of construction risk. Whether they paid two or fifteen percent more than an open auction would have cleared at is unknowable, because no auction ever happened — and it never could have, since no other buyer on earth could be permitted to own a Chinese mainstem mega-dam.

The playbook by 2016 was fully visible: CTG builds for a decade, proves the cash flow, then sells into the listco. And by then everyone with a spreadsheet knew exactly which two dams were next.

↑ Back to Top

IV. The Final Piece: Wudongde, Baihetan, and the End of the Easy Growth Runway

On June 28, 2021, the first units at 白鹤滩 Baihetan came online — a 289-metre double-curvature arch dam in a gorge on the Jinsha, with sixteen turbines of 1,000 MW each. Each single unit at Baihetan is roughly the size of a large nuclear reactor. Together with 乌东德 Wudongde downstream, commissioned a year earlier, the last two giants of the lower Jinsha were finished.

Everyone knew what came next. The only questions were when, and for how much.

December 2021 to January 2023

CYPC disclosed the restructuring plan on December 10, 2021: acquire 100% of 云川公司 Yunchuan Company, the entity developing Wudongde (10,200 MW) and Baihetan (16,000 MW), via share issuance plus cash.⁶ The appraised value of Yunchuan's equity, at a January 31, 2022 valuation date, was RMB 80.484 billion.⁶

The deal did not sail through unexamined. The Shanghai Stock Exchange issued a formal inquiry letter to CYPC over the proposed acquisition, pressing on valuation and transaction terms.⁷ That is worth noting precisely because the transaction is so often described as a formality: the exchange treated it as something requiring justification, which is the minimum one should want from a related-party deal of this size.

The equity transfer completed on January 10, 2023, and Yunchuan became a wholly-owned subsidiary.⁸ Installed capacity jumped roughly 26 GW — a 36.5% increase in a single step.

The cascade is complete — and so is the runway

With that transfer, CYPC owned the full sequence: Wudongde, Baihetan, Xiluodu, Xiangjiaba, Three Gorges, Gezhouba. Six stations, one river, one dispatch system.¹

This is the single most important inflection point in the investment case, and it gets remarkably little airtime relative to its significance. For twenty years, the answer to "where does growth come from?" was always the same: the next dam upstream. That answer no longer exists. The Yangtze mainstem and lower Jinsha are built out. There is no seventh giant waiting in the parent's portfolio to be dropped down.

Everything CYPC does from here — pumped storage, overseas acquisitions, capacity uprates, market trading — has to be judged as a new competence, not an extension of the old one. The old competence was, essentially, having a rich parent with excellent assets and a need for cash.

2023: the preview

The market got an early demonstration of life after injections almost immediately.

In 2023, with roughly 36.5% more installed capacity than the year before, CYPC's six stations generated 276.263 billion kWh, up 5.34% year on year, and net profit attributable to shareholders came in at RMB 27.24 billion, up 15.44%.⁹

Sit with that arithmetic for a moment. Capacity rose by more than a third. Generation rose by five percent. Profit rose by fifteen. The gap is water — 2023 was a 来水偏枯 year, meaning inflow to the reservoirs came in below the long-run average, so brand-new world-class turbines spent a great deal of 2023 not spinning.

That was the preview: two of the largest hydro stations ever built joined the fleet, and the earnings lift was materially smaller than the capacity lift because the sky did not cooperate. It is not a criticism of management — no one dispatches rain. It is a statement about what the earnings stream actually is. For twenty years, asset injections provided a growth signal loud enough to drown out hydrological noise. Without them, the noise is the signal.

Which raises the obvious question: if the growth is over, what exactly is the thing that remains?

↑ Back to Top

V. How a Dam Prints Cash: The Business Today

Start with a single number that explains more about this business than any margin ratio: in 2025, CYPC's entire domestic hydropower cost line — the whole thing — was RMB 25.88 billion, and the annual report labels it "depreciation cost and various financial levies and charges."¹

There is no fuel line. There is no coal price, no gas hedge, no uranium contract, no scrubber retrofit. The input is rain, and rain does not send invoices.

The economics, in plain terms

The domestic hydropower segment produced revenue of RMB 75.66 billion in 2025 at a gross margin of 65.79%, up 3.28 percentage points year on year.¹ Everything else — the Peruvian utilities, electricity retailing, technical services — contributed RMB 10.32 billion at a 31.54% margin.¹ So roughly 88% of revenue and a much larger share of profit comes from six dams.

Think of the cost structure as a mortgage rather than a factory. A thermal plant's economics change every time coal moves. CYPC's do not change at all in that sense: the dam was paid for once, the accountants write it down on a schedule, and the cash cost of producing an additional kilowatt-hour is close to nothing. Depreciation dominates the P&L while producing no cash outflow, which is why operating cash flow of RMB 60.56 billion in 2025 dwarfed reported net profit.¹

That gap — cash generation running roughly 75% above accounting earnings — is the mechanical reason a 70% payout ratio is possible at all. And it gets better with time in one specific respect: concrete gravity dams have physical lives far longer than their depreciation schedules. At some point the accounting charge runs off while the turbines keep turning. That is a genuine, if slow-moving, tailwind that most utility comparisons miss.

The lever that actually moves earnings

If you want to understand why CYPC's profits move, ignore almost everything management does and look at utilisation hours — how many hours per year each station effectively ran at full output.

In 2025 the pattern was strikingly uneven. Three Gorges ran 4,303 hours, up 13.44% year on year. Gezhouba ran 7,026 hours, up 11.87%. Meanwhile, upstream, Wudongde ran 3,625 hours, down 6.63%, and Xiluodu fell 1.51%.¹ Same company, same year, same management, and the stations diverged by twenty percentage points.

That is not strategy. That is where the rain fell. The middle Yangtze was wet and the upper Jinsha was comparatively dry.

Total 2025 generation from the six domestic stations reached 307.194 billion kWh, up 3.82%, crossing 300 billion kWh for the first time and pushing cumulative lifetime generation past 4 trillion kWh.¹ Management's own 2026 plan makes the dependency explicit in a way few companies would dare put in print: the target of 306 billion kWh is conditioned on Wudongde reservoir inflow of no less than 130 billion cubic metres and Three Gorges reservoir inflow of no less than 480 billion cubic metres.¹ The guidance is, quite literally, a weather forecast with a caveat attached.

The KPIs that matter

Strip everything else away and three numbers carry this company.

First, annual generation from the six cascade stations against the ~300 billion kWh benchmark, read alongside reservoir inflow. This is the volume driver and it explains most of the year-to-year swing in profit. Everything else is second-order.

Second, the share of output sold through market mechanisms rather than administered tariffs, and the realised price that results. In 2025, market-traded volume was 110.45 billion kWh against total on-grid volume of 305.56 billion kWh — 36.1%, down 2.5 percentage points from 38.6% in 2024.¹ That ratio is the price-risk dial.

Third, the payout ratio against interest-bearing debt. This is the capital allocation tell — whether the dividend is being funded by cash generation or by the balance sheet.

The dispatch edge — real, but hard to measure

Management describes a genuine operating capability here, and it deserves both credit and scepticism. In 2025 the company reported 14.01 billion kWh of "water-saving power generation" — output achieved through dispatch optimisation that would otherwise have been spilled — and said comprehensive water consumption rate, water utilisation rate, and short-term forecasting accuracy all hit record levels.¹ It also managed nine flood events exceeding 25,000 cubic metres per second at the Three Gorges reservoir within fifty days, and handled what it called the strongest western China autumn rains since 1961.¹

The 14 billion kWh figure is roughly 4.5% of annual output. If broadly accurate, that is meaningful value created by coordination rather than capital. But note the epistemics: this is a company-defined metric, computed internally, with no external audit and no peer benchmark. It is plausible and it is directionally consistent with owning the whole staircase. It is not independently verified, and an investor should treat it as management's characterisation rather than an established fact.

The customer problem

Now the uncomfortable structural feature. In 2025, CYPC's top five customers accounted for 100% of domestic sales — and functionally there were only two. 国家电网 State Grid took RMB 58.92 billion of power, or 68.8% of sales. 南方电网 China Southern Power Grid took RMB 26.72 billion, or 31.2%.¹

Two buyers. Both state-owned. Both monopolies in their own territories. There is no third option, no export route, no direct-to-consumer channel of consequence. A dam cannot ship its product anywhere except into the wires that happen to be attached to it.

For most of CYPC's history this did not matter, because prices were set administratively and were stable. It matters enormously now, because the pricing regime is changing — which is Section IX's problem.

The dividend as the product

Which brings us to what CYPC actually sells to shareholders, as distinct from what it sells to the grid.

Since 2021 the company has committed to distributing no less than 70% of annual net profit as cash dividends, and in August 2025 it extended the same 70% floor across 2026–2030.¹ For FY2025 it proposed RMB 1.00 per share, RMB 24.47 billion in total — a payout ratio of 70.92%.¹ It has also begun paying interim dividends, distributing RMB 0.21 per share, or RMB 5.14 billion, at the 2025 half-year mark, the second consecutive year of doing so.¹⁰

This is not incidental. A multi-decade asset with near-zero marginal cost and a contractually committed payout ratio is, functionally, a very long-duration bond with an inflation-linked and weather-linked coupon. That is precisely how a large slice of the shareholder base treats it — and precisely why the valuation behaves the way it does.

But a bond with a weather-linked coupon is only as good as the issuer's balance sheet. And CYPC did not get to six dams for free.

↑ Back to Top

VI. Crossing Borders: Peru, Portugal, and the M&A Report Card

On April 24, 2020, with global markets still reeling from the first pandemic wave and cross-border deal-making largely frozen, a transaction closed in Lima that had almost nothing to do with the Yangtze.

Sempra Energy completed the sale of its Peruvian businesses for approximately $3.59 billion in total cash proceeds — an 83.6% stake in Luz del Sur S.A.A., Peru's largest electricity distributor, plus the construction services arm Tecsur and the generation business Inland Energy.¹¹ The buyer was China Yangtze Power International (Hongkong), with the purchase assigned to Yangtze Andes Holding.¹¹ The agreement had been signed in October 2019; Peruvian regulatory clearance arrived on April 10, 2020.¹¹ Luz del Sur serves southern Lima and holds roughly 29% of Peru's distribution market.¹²

Why this deal is categorically different

Everything CYPC had bought before came from its own parent, at an appraised price, with no competing bidder and no possibility of losing. Peru was the opposite in every respect. It was an auctioned asset in an open process against international bidders. It was priced in dollars, in a foreign regulatory regime, funded with debt rather than parent-subsidised equity.

It was also a completely different business. Hydro generation is a physical asset with an output and a price. Regulated distribution is a rate-base business — you own wires, a regulator sets an allowed return, and your job is to invest capital efficiently into a formula while collecting from millions of end customers. The skills barely overlap.

So: how has it gone?

The scorecard, six years on

The evidence is better than the sceptics expected and more modest than the promoters claim.

By 2025, Luz del Sur delivered total profit of $298 million, which CYPC states was 42.91% higher than at the time of acquisition.¹ It also earned a Fitch rating of BBB+ — above Peru's sovereign rating, making it one of the highest-rated corporates in the country.¹ Earning a rating above your own sovereign is a genuine credit achievement; it reflects cash flow stability and structural protections that rating agencies apply sparingly.

CYPC has since built on the position rather than sitting on it. During 2025 it completed unit takeover, generation start-up and initial operation and maintenance at the SanGaban III hydropower station in Peru, and acquired the Peru Red Coral project — adding 135,700 kW of wind capacity and making CYPC the largest wind power operator in the country.¹

The analytical read: this looks like a competently digested acquisition rather than a trophy. Profit growth of roughly 43% over roughly five years is respectable for a regulated distributor — not spectacular, but regulated distribution is not supposed to be spectacular. The credit rating outcome and the subsequent bolt-ons suggest CYPC learned the market rather than merely occupying it.

The caveats are equally real. Peru contributes a small fraction of group profit. The returns are exposed to Peruvian regulatory review cycles, sol/dollar/renminbi currency movement, and a political environment that has cycled through multiple presidents since the deal closed. And a single well-executed acquisition is not yet a capital allocation track record.

The cautionary tale one level up: Portugal

To understand what "state-owned capital discipline" looks like when a deal gets genuinely contested, you have to look above CYPC, to the parent.

In December 2011, CTG bought 21.35% of 葡萄牙电力 Energias de Portugal (EDP) for roughly €2.7 billion, winning a privatisation auction against E.ON and Brazilian bidders.¹³ For seven years it behaved as a strategic minority holder.

Then in May 2018, CTG went for the whole thing. Already holding about 23% of EDP, it offered €3.26 per share for the rest — valuing the balance at roughly €9.07 billion, and representing a premium of only about 4.8% over the prior close.¹⁴ EDP's board rejected it as undervaluing the company.¹⁴ On April 24, 2019, EDP shareholders blocked the bid outright, refusing to remove the voting-rights cap that CTG had made a condition of the offer.¹⁵

Two readings compete here, and both contain truth.

The charitable one: CTG bid a disciplined price, refused to chase, and walked away rather than overpay. That is exactly what you want from a capital allocator, and the partnership with EDP survived — the two later restructured their arrangement to focus on joint growth in Latin America.¹⁶

The uncharitable one, which is harder to dismiss: a 4.8% premium for control of a company the acquirer already partly owned was never going to succeed. Launching a hostile offer for a national champion in an EU capital, in 2018, as a Chinese state-owned enterprise, on those terms, reflects a misjudgement of both price and politics. It cost a year, considerable reputational capital, and it failed publicly.

What the two deals together tell an investor

The honest synthesis: CYPC's domestic acquisition record is shareholder-friendly by construction, not by demonstrated skill. Buying finished assets from a captive seller at appraised prices is not the same competence as competing in an open market. Peru is the only meaningful data point where CYPC competed and then had to actually operate, and that data point is encouraging.

The parent's EDP episode is a warning that state ownership does not confer immunity from ambition outrunning judgement. When CYPC eventually deploys serious capital into its next act — and it will have to — the relevant question is not whether the state will approve. It is whether the people approving it have ever had to be right in a market that could tell them no.

Which makes it worth asking who, exactly, those people are.

↑ Back to Top

VII. Who's Running the River: Current Management, Ownership, and Capital Allocation

The two most senior people at China Yangtze Power arrived from opposite directions, and the contrast is the most revealing thing about how the state thinks about this asset.

The chairman: a water engineer, not a utility executive

刘伟平 Liu Weiping was born in November 1964 and spent more than forty years inside China's water bureaucracy. He rose to Chief Engineer of the 水利部 Ministry of Water Resources and then, in 2021, to Vice Minister.¹⁷ On April 15, 2024 he was appointed chairman and Party secretary of CTG, succeeding 雷鸣山 Lei Mingshan.¹⁷ In August 2024 he assumed the CYPC chairmanship and legal representative role.

Read the résumé for what it signals. This is not a career power-sector operator, nor a financier, nor a market-facing chief executive. It is a hydrology and water-resources technocrat, moved directly from a ministry that regulates rivers to a company that operates them.

That tells you how Beijing categorises CYPC: as strategic water infrastructure that happens to sell electricity, not as a listed utility that happens to sit on rivers. The Three Gorges complex does flood control for the middle Yangtze basin, maintains navigation depth, and releases water for downstream agriculture and drinking supply during dry seasons. Those functions have no revenue line and enormous political weight. Appointing a former Vice Minister of Water Resources is the personnel equivalent of stating that the public-good mandate is not subordinate to the P&L.

For minority shareholders this cuts both ways. It underwrites the regulatory protection that makes the asset un-attackable. It also means that when flood control, navigation, and generation conflict — and they do, every summer — generation is not automatically what wins.

The president: forty years of the same six dams

刘海波 Liu Haibo was appointed general manager on April 9, 2025.¹⁸ Born in October 1971, a senior engineer by training, his career reads like a tour of the company's own assets: deputy director of the Baihetan plant preparation office, deputy plant manager at Baihetan, then plant manager and Party secretary at Baihetan — through its construction and commissioning — followed by deputy general manager of CYPC while concurrently running Baihetan, then the same arrangement at the Three Gorges plant.¹⁸

He was, in other words, personally running the world's second-largest hydropower station while it was being built and brought online, and then ran the largest one.

The pairing is deliberate. The chairman's seat rotates with state personnel cycles and carries the political mandate. The operating seat is filled from inside, by someone who has physically commissioned the assets. For a business where the difference between good and great is measured in dispatch efficiency and unit availability — CYPC reported a 100% high-quality overhaul rate for equipment for three consecutive years¹ — insider continuity is defensible.

The counterpoint an activist would raise: an entirely internally-grown executive team, at a company entering a phase requiring genuinely new capital allocation competence, is a thin bench for the task. Nobody in the senior team was hired for their record of allocating capital against alternatives in a competitive market. They were hired for their record of running dams superbly.

The owner keeps buying

Then there is the shareholder register, where something concrete happened.

On August 22, 2025, CTG notified CYPC that it planned to increase its holding through open-market purchases over the following twelve months, committing to no less than RMB 4 billion and no more than RMB 8 billion, funded from its own and self-raised funds, with no fixed price band.¹⁹

It then executed. Between August 23 and December 8, 2025, CTG bought 103,342,440 shares through centralised bidding for RMB 2.884 billion excluding fees, lifting the combined stake of CTG and its concert parties from 52.58% to 53.00%.¹⁹²⁰ By December 31, 2025 the cumulative purchase had reached 161,581,335 shares for RMB 4.499 billion — already at the bottom of the announced range with eight months of the window remaining.¹

This is a different category of evidence from a management statement. A controlling shareholder that already holds a majority gains no control benefit from buying more; it simply converts cash into shares at prevailing prices. Doing so with real money, disclosed and verifiable, at a stake level well past the point of necessity, is about as clean a signal of parent-level conviction as this market produces. It is not proof the shares are cheap — CTG is not a disinterested valuer, and there is an obvious policy dimension to state entities supporting their listed subsidiaries. But it is behaviour, not rhetoric.

The payout record: promises kept

On capital returns, the record is unusually clean by the standards of Chinese state-owned enterprises, and it should be evaluated as a multi-year behavioural pattern rather than a single announcement.

The 2021–2025 commitment was a floor of 70% of net profit. Delivered: RMB 18.54 billion on RMB 26.27 billion of 2021 profit, about 70.6%. Then RMB 20.09 billion on RMB 21.31 billion in 2022 — a payout above 94%, the highest in company history, made in a year when profit fell nearly 19%.⁹ Then RMB 20.06 billion in 2023, 73.66%.⁹ Then roughly RMB 23.1 billion for 2024, about 71%.²¹ Then RMB 24.47 billion for 2025, 70.92%.¹

The 2022 number is the one that carries the information. Profit dropped sharply, and the company paid out more than it earned relative to the prior year's absolute dividend rather than cut. Anyone can honour a payout floor in a good year. Honouring it in a bad one, and then extending the same floor for another five years in August 2025, is the behaviour of a company that understands its shareholder base treats the dividend as the primary product.²² Cumulative distributions since listing have run into the hundreds of billions of renminbi.²³

The activist's question — and it is a good one

Here is where the story gets genuinely contested, and it happened in public.

On July 30, 2025, CYPC announced it would invest approximately RMB 26.6 billion in a Gezhouba navigation capacity expansion project — demolishing the existing third ship lock and building two new single-stage locks on the left bank, creating a four-lock configuration capable of handling 10,000-tonne class vessels, over a 91-month construction period.²⁴ The stated rationale was congestion: 2024 cargo throughput through the facility reached 156 million tonnes, and the bottleneck constrains traffic feeding through from the Three Gorges locks.²⁴

Note what this project is. It is not a power station. Ship locks do not generate electricity. A capital commitment of roughly RMB 26.6 billion — around 82% of 2024 net profit — is going into public navigation infrastructure.²¹

Minority shareholders said so, out loud, at the August 2025 shareholders' meeting. They asked whether the dividend would suffer, and whether the project would generate commercial returns at all — vessel tolls, higher power prices, anything.²¹ Management's response leaned on cash-flow capacity: 2024 operating cash flow of RMB 59.6 billion against free cash flow of roughly RMB 45.2 billion and dividends of RMB 23.1 billion, with the incremental annual depreciation from the lock project estimated at around RMB 600 million.²¹

That answer addresses affordability. It does not address returns. And the distinction matters enormously. A company that can afford to spend RMB 26.6 billion on an asset with no identified revenue stream is describing the size of its cushion, not the quality of its investment. This is the "diworsification" question in its purest form — and its resolution is not really a financial one. As long as the controlling shareholder is the state and the chairman comes from the water ministry, some portion of CYPC's capital will be allocated to national infrastructure objectives that a purely commercial owner would decline.

Investors should price that as a structural feature of owning this asset, not as a management failing. But they should price it.

↑ Back to Top

VIII. The Moat: Competitive Landscape, Five Forces, and the Cornered Resource

Here is a thought experiment that ends faster than most.

Suppose you had unlimited capital, complete political access, and a mandate to compete with China Yangtze Power. What would you build, and where?

The Yangtze mainstem is finished. The lower Jinsha is finished. The remaining large-scale hydro potential in China sits on the upper Yangtze tributaries and, most significantly, in Tibet — remote, seismically active, geopolitically sensitive terrain where projects run a decade or more and are assigned by the state rather than won. You could not build a rival to Three Gorges because there is nowhere left to put one, and even if there were, the approvals, land acquisition, and resettlement processes that took CTG decades to navigate are not available to a private challenger at any price.

That is not a competitive advantage in the ordinary sense. It is the absence of a competitive game.

Helmer's Cornered Resource, in its purest form

Hamilton Helmer's 7 Powers framework describes a "Cornered Resource" as preferential access to a coveted asset that independently enhances value. Most claimed examples are soft — a patent portfolio, a talent cluster, a brand.

CYPC's version is concrete and steel sitting in specific geological formations, holding back specific volumes of water, with an exclusive legal right to generate from them. It cannot be reverse-engineered, poached, copied, or disrupted. It is not going to be obsoleted by a better algorithm. The barrier is not merely high; the category of entry does not exist.

Scale tells the same story. At roughly RMB 695 billion of market value, CYPC is several times larger than any domestic power generation peer.² 中国核电 China National Nuclear Power sits around RMB 180 billion; 中国广核 CGN Power around RMB 198 billion; 大唐发电 Datang International around RMB 112 billion; 华能国际 Huaneng Power International around RMB 105 billion; and 三峡能源 China Three Gorges Renewables — CTG's separately listed wind and solar vehicle — also around RMB 105 billion.

That last one deserves a moment. CTG deliberately split its renewables into a different listed company rather than folding them into CYPC. The logic is instructive: hydro is dispatchable, long-lived, and cash-stable; wind and solar are intermittent, shorter-lived, and capex-hungry. Mixing them would have blurred exactly the characteristics that make CYPC investable as a bond proxy. Whatever else one says about CTG's capital markets judgement, that structural decision was sophisticated.

Process Power: real, but softer than advertised

The second claimed moat is coordinated cascade dispatch — 梯级调度 — the accumulated organisational know-how of running six linked reservoirs as one optimisation problem across flood control, water supply, navigation, and generation.

There is a reasonable case that this qualifies as Helmer's "Process Power": embedded organisational capability that a competitor could not replicate quickly even with the same physical assets. The forecasting models are built on decades of proprietary basin hydrology. The operational protocols were learned through actual floods.

But an investor should be honest about the evidence base. The supporting facts — record water-utilisation metrics, the water-saving generation figure, letters of appreciation from provincial governments, industry recognition — are company-reported and largely qualitative.¹ There is no independent benchmark showing CYPC extracts more energy per cubic metre than a comparably-equipped operator would. The capability is plausible and probably real. Its magnitude is not established, and it is almost certainly a smaller contributor to value than the cornered resource itself.

Porter's Five Forces, run honestly

Threat of new entrants: effectively zero. Covered above. No sites, no approvals, no path.

Supplier power: negligible. The primary input is water, allocated by the state at no fuel cost. Equipment suppliers matter during construction and overhaul, not in operation — CYPC's top five suppliers accounted for 55.29% of domestic purchases against a cost base that is overwhelmingly depreciation.¹

Buyer power: meaningful and rising. This is the force that is actually changing. With two state grid companies taking 100% of domestic output, CYPC faces a duopsony that increasingly sets price through market mechanisms rather than administered tariffs.¹ Two decades of stable pricing reflected policy, not negotiating leverage. As policy changes, the underlying weakness becomes visible.

Threat of substitutes: asymmetric and genuinely two-sided. China added over 300 million kW of new-energy capacity in 2025 alone, taking total renewable capacity to 2.34 billion kW, about 60% of the national total.¹ Solar and wind at scale push down average clearing prices in the hours they generate — that is real erosion. But intermittency is exactly what hydro solves. A reservoir is a battery measured in cubic kilometres, and a grid absorbing hundreds of gigawatts of variable generation needs dispatchable, fast-ramping capacity more, not less. The National Development and Reform Commission's Action Plan for Optimization of Power System Flexibility (2025–2027), issued in January 2025, focuses explicitly on hydropower basin regulation efficiency to support new-energy absorption.¹ Hydro's energy value faces pressure; its flexibility value should rise. Whether the market pays for the second as fast as it discounts the first is an open question — and it is arguably the most important open question in the whole case.

Competitive rivalry: essentially nonexistent within large hydro. There is no comparable asset to compete with.

Why this wins from here, and what breaks it

The honest summary: CYPC's moat is close to perfect against the risks it faces from competitors, and offers no protection whatsoever against the risks it faces from counterparties, weather, and policy.

No one will take its customers. Its customers can, however, change how they pay. No one will build a rival dam. The sky can, however, decline to fill the one it has. That is an unusual risk profile — extremely low business risk, unusually high exogenous risk — and it is precisely why the stock trades like a bond and why the things that break bonds are the things to watch.

↑ Back to Top

IX. Current Risk Radar

In August 2022, the Yangtze basin stopped behaving like the Yangtze basin.

A record heatwave settled over central and southwestern China and did not leave. Rainfall collapsed. The Three Gorges reservoir fell to levels roughly 40% below the average of the prior four years.²⁵ In Sichuan — a province that draws more than 80% of its electricity from hydro — reservoirs emptied while air-conditioning demand spiked, and authorities ordered industrial power rationing.²⁶ Factories in one of the country's manufacturing hubs went dark because a river ran low.

For CYPC, the financial consequence was straightforward: net profit fell from RMB 26.27 billion in 2021 to RMB 21.31 billion in 2022, a decline of nearly 19%, in a business with no fuel cost and no competitors.⁹

That is the risk radar's first and most important entry, and everything else on it is smaller.

1. Hydrology and climate — the structural exposure

The mechanism is worth stating precisely, because "weather risk" understates it. CYPC's revenue is a linear function of water volume passing through turbines. There is no hedge, no inventory, no substitute input, and no meaningful ability to store output. A dry year is not a delayed sale; it is a permanently forgone one.

The company's own risk disclosure concedes the point without much softening, noting that power generation is closely tied to Yangtze inflow and that the uncertainty of that inflow affects production.¹ Its mitigation is entirely informational — better forecasting, deeper cooperation with hydrological and meteorological agencies, more sophisticated joint cascade dispatch.¹ Those measures genuinely improve extraction from whatever water arrives. They cannot create water.

The climate dimension raises the tail. 2022 was described as the most severe such event in the modern record; 2023 delivered below-average inflow; 2025 brought the strongest western China autumn rains since 1961 in the other direction.¹ The distribution of outcomes appears to be widening at both ends. A cascade with large reservoirs buffers seasonal variance well. It buffers multi-year regime shifts poorly.

2. Power market reform — the slow-moving one that matters most

If hydrology is the volatile risk, pricing reform is the structural one, and it is arriving on a published timetable.

The direction of travel began with the February 2025 policy push to move new-energy generation onto market-based pricing rather than fixed benchmark tariffs — a deliberate step toward letting supply and demand, rather than administrative schedules, set power prices.²⁷ Analysts of China's power sector have described this as a foundational shift in how electricity value is determined.²⁸

Then, in February 2026, the General Office of the State Council issued Implementation Opinions on Improving the National Unified Electricity Market System. The targets are explicit: by 2030, all power sources and all users except guaranteed users participate directly in the electricity market, market-based transaction volume reaches about 70% of total electricity consumption, and spot markets enter full formal operation. By 2035 the system is to be fully established, with energy, regulation, environmental and capacity value all determined through the market.¹

Map that against where CYPC sits today: 36.1% of on-grid volume traded through market mechanisms in 2025.¹ The national target implies roughly a doubling of market exposure within five years.

Management's framing is optimistic. The annual report argues that as the unified market deepens, the energy value, flexibility value and green environmental value of large cascade hydropower will be "further transformed into market competitiveness."¹ The company points to hydro green certificates entering the issuance scope in 2025 as a potential incremental revenue source, and its retail arm sold over 6 billion kWh across nine provinces during the year.¹

The bear reading is simpler and harder to dismiss. Market clearing prices are set by marginal cost. Solar and wind have marginal costs near zero and are being added at over 300 million kW per year.¹ In the hours when hydro generates most — the wet season, which overlaps with peak solar output — the marginal price-setter will increasingly be an asset with no fuel bill at all. Hydro's dispatchability should command a premium through capacity and ancillary service payments, but those mechanisms are still being designed, and the company's own risk disclosure lists the gradual refinement of pumped storage capacity pricing among its uncertainties.¹

The falsifiable test is straightforward: watch realised revenue per kWh over the next several years. If flexibility value genuinely offsets energy value compression, average realisation holds. If it does not, it falls — and a bond proxy with a shrinking coupon is not a bond proxy.

3. Leverage and refinancing

CYPC bought six dams. It did not pay cash for them.

At the end of 2025, total liabilities stood at RMB 325.85 billion against total assets of RMB 559.21 billion.¹ Long-term borrowings were RMB 172.31 billion, non-current liabilities due within one year RMB 70.20 billion, bonds payable RMB 29.78 billion, and short-term borrowings RMB 15.18 billion — the last down 78% year on year as the company termed out its debt.¹ Cash and equivalents sat at just RMB 4.59 billion, which management described as deliberate minimisation of idle balances.¹

The trend is favourable. Total liabilities fell by roughly RMB 18.7 billion during 2025, and financial expenses dropped 15.81% to RMB 9.37 billion — a reduction of RMB 1.76 billion that flowed almost entirely to the bottom line.¹ That single line item explains a large share of 2025's profit growth. Credit ratings sit at Fitch A and Moody's A1, among the highest for any Chinese corporate.¹

But note what that implies about earnings quality. Growth driven by refinancing at lower rates is real, and it is also finite and non-repeatable. Interest costs cannot fall forever. Meanwhile, the constraint is arithmetic: with roughly 71% of profit committed to dividends and a debt load of this size, incremental capex must come from the residual cash flow or from more borrowing. Add the Gezhouba lock project, an expanding pumped storage pipeline, and potential overseas deals, and the flexibility that looks comfortable today gets tested.

4. The growth cliff and the valuation it supports

At roughly 19 times trailing earnings and a 3.6% yield, CYPC trades at a premium to most global regulated utilities and to essentially every domestic thermal generator.²

The premium is not irrational — the asset quality genuinely justifies something. But an investor should be clear about what is actually being bought at that multiple. Domestic asset injections are finished. Organic capacity growth is limited to capacity-uprate applications filed for Xiluodu and Xiangjiaba and to pumped storage projects that will take years to commission.¹ Underlying earnings growth ex-financing is modest: basic EPS grew 6.17% in 2025, but excluding non-recurring items it grew 2.89%, and weighted average ROE excluding non-recurring items actually fell 0.31 percentage points to 15.41%.¹ The gap was driven substantially by RMB 1.68 billion of fair-value gains on financial assets.¹

The same pattern showed up in Q1 2026: reported net profit rose 30.50% to RMB 6.76 billion, but profit excluding non-recurring items rose 19.20%.²⁹ Strong quarter, genuinely — generation was up 7.19% — but roughly a third of the headline growth came from items that will not repeat on schedule.²⁹

The multiple, therefore, rests substantially on the dividend, and the dividend's attractiveness is relative to Chinese government bond yields. A large share of the register is insurance and pension capital treating this as a fixed-income substitute. If domestic yields rise materially, the relative case weakens, and the de-rating would come from the discount rate rather than from anything the company did. That is a risk management cannot manage.

5. Overseas and geopolitical exposure

Peru is small in group terms but carries risks the domestic business simply does not have: regulatory tariff reviews, currency translation, and political instability. CYPC's own disclosure acknowledges outbound investment has become more difficult amid changes in the international situation and intensified competition, and specifically flags geopolitical risk identification as an area requiring strengthening.¹ The EDP experience is the reminder of what happens when a Chinese state-owned bidder meets a politically sensitive European asset.

6. Execution risk in the next act

The pumped storage build-out is the clearest test of whether CYPC can allocate capital as well as it operates assets.

The flagship is Zhangye in Gansu — 1.4 GW across four 350 MW reversible pump-turbine units, with total investment budgeted at up to RMB 9.57 billion and first unit generation targeted for April 2028.[^30] Cumulative investment in the project reached RMB 2.12 billion by the end of 2025.¹ Beyond it sits a widening pipeline: Guanghanping in Hunan, Caiziba in Chongqing, Housihe in Gongyi, Xunwu in Jiangxi, Lizhuang in Anhui, Binggou in Hebei, plus equity acquisitions of four project companies during 2025 alone.¹ Total construction in progress rose 66.49% to RMB 15.91 billion.¹

Explained simply, pumped storage is a rechargeable battery made of water and altitude: pump water uphill when power is cheap, release it downhill through turbines when power is expensive, and pocket the spread. It is the only mature large-scale storage technology available, and a grid absorbing hundreds of gigawatts of solar needs a great deal of it.

The problem is that the economics depend almost entirely on a capacity pricing mechanism that regulators are still refining — CYPC's own risk disclosure says so.¹ Investor discussions have put typical project IRRs in a 6–8% range.²² That is not a bad return for a regulated asset. It is a materially worse return than the mainstem hydro business it is being compared against, and it is dependent on administrative price-setting rather than a cornered physical resource.

Pumped storage, in other words, is a reasonable business that CYPC is entering as one competitor among many, funded by cash flows from a business where it has no competitors at all. That is a real change in the character of the enterprise, and it is the thing most likely to be underestimated.

↑ Back to Top

X. Bull Case, Bear Case, and the Playbook

Every investment case eventually reduces to a disagreement about one thing. Here, the disagreement is not about asset quality — nobody disputes the dams. It is about whether asset quality is sufficient once growth is finished.

The bull case, and the evidence behind it

The asset is genuinely un-replicable, and that is not a figure of speech. Six stations, 71.7 GW, roughly 16% of China's hydro capacity, on a river system with no remaining mainstem sites.¹ The barrier is physical geography plus a decades-long approval and resettlement process that no challenger can access. This is the strongest form of Helmer's Cornered Resource available in public markets.

The cost structure improves with time. Depreciation dominates costs, is non-cash, and eventually runs off against concrete rated for far longer than its book life. Operating cash flow of RMB 60.56 billion against net profit of RMB 34.50 billion in 2025 shows the gap in action.¹

The capital return commitment has been tested and honoured. Not just declared — delivered through a drought year at above 94% of profit, then re-committed for another five years.⁹²² Five consecutive years of meeting or exceeding a public floor is a behavioural record, not a promise.

The controlling shareholder is buying with its own money. RMB 4.50 billion of open-market purchases by December 31, 2025 against a committed range of RMB 4–8 billion, taking the stake past 53%.¹¹⁹

The energy transition may need this asset more, not less. Every gigawatt of solar added to the Chinese grid increases the value of dispatchable, fast-ramping, storable capacity. CYPC owns the largest such fleet in the world, and national policy is explicitly focused on hydropower basin regulation to support new-energy absorption.¹

The bear case, given equal weight

The growth engine is structurally finished, and nothing has replaced it. Four asset injections over twenty years produced step-function earnings growth. There is no fifth. What is left — capacity uprates, pumped storage at 6–8% IRRs, incremental overseas deals — is smaller, slower, and lower-returning.²²

Recent earnings growth is flattered by non-recurring items and financing. EPS ex-items grew 2.89% in 2025 while reported EPS grew 6.17%; ROE ex-items declined.¹ Much of the improvement came from a RMB 1.76 billion reduction in financial expenses that cannot repeat indefinitely.¹

Leverage is elevated exactly as the growth lever disappears. Total liabilities of RMB 325.85 billion, minimal cash, a 70% payout floor, and a new capex cycle including RMB 26.6 billion of non-generating navigation infrastructure.¹²⁴ Something has to give if any of those three commitments comes under pressure.

The pricing regime that made this a bond proxy is being dismantled on a schedule. Market-based transactions are targeted to reach about 70% of national consumption by 2030 against CYPC's current 36.1% market exposure.¹ The company has not quantified the expected impact on realised tariffs, and its disclosure on the topic remains directional rather than numerical.¹

Hydrology is unhedgeable and possibly worsening. 2022 demonstrated a 19% profit decline with no operational failure whatsoever.⁹

Some capital will be allocated to public goods regardless of returns. The Gezhouba lock expansion is the visible case, and minority shareholders challenged it directly without receiving a return-based answer.²¹²⁴ A chairman drawn from the water ministry makes this a persistent feature rather than a one-off.

The stress test a sceptical investor would actually run

Forget the dams for a moment and interrogate the capital allocation.

Ask why the company holds equity investments in 63 entities with cumulative initial investment of roughly RMB 58.2 billion, generating investment income of RMB 4.96 billion and cash dividends of around RMB 2.6 billion in 2025.¹ That is a substantial financial portfolio for an operating utility. It contributed a meaningful share of the year's fair-value gains — and therefore of the reported earnings growth. Is that core, or is it a source of earnings volatility dressed up as diversification?

Ask why a company with a cornered physical resource needs to be in the electricity retailing business at all, given the retail arm's sub-6 billion kWh scale against 300-plus billion kWh of generation.¹

Ask whether the same-control business combinations that required retrospective restatement of comparative financials in 2025 — the pumped storage entity acquisitions from the parent¹ — represent continued related-party dropdown activity that deserves the same scrutiny the mainstem injections received, at a point when the assets being dropped down are considerably less attractive than dams.

Ask, finally, what the disclosure does not contain. There is no quantified guidance on realised tariff under different market-penetration scenarios. There is no stated return threshold for pumped storage investments. There is no explicit framework for how the board weighs dividends against capex when both cannot be fully funded. Those are the three questions that determine the next decade, and the answers are not in the filings.

The durable lesson

The generalisable insight here has nothing to do with dams.

A monopoly-grade physical asset is a wonderful thing to own, and for as long as there is a runway of similar assets to consume, it produces an exceptional compounding machine that requires very little judgement. CYPC had that runway for twenty years, and it worked exactly as advertised. Shareholders who bought quality and held were richly rewarded, and they were rewarded largely without management having to make a difficult capital allocation decision.

But asset quality and growth are separate variables, and investors routinely conflate them because for long stretches they arrive together. When the runway ends, the asset is exactly as good as it ever was — and the stock is a fundamentally different security, because everything now depends on judgement rather than inheritance.

CYPC is at that transition point today. The dams have not changed. The question has. It is no longer "how good are the assets?" — that is settled and unarguable. It is "how good are these people at deciding what to do with the cash the assets produce, in arenas where they have never been tested and where being wrong is possible?"

Two decades of evidence answer the first question. Almost none answers the second.

↑ Back to Top

XI. Epilogue: What to Watch

The next chapter of this story will not be written by engineers. It will be written by hydrologists, regulators, and whoever sits in the room when the board decides where the cash goes.

Does the 70% floor survive contact with the new capex cycle?

The 2026–2030 commitment is in writing.²² The test is not a good year — it is the first year in which weak inflow collides with peak spending on locks and pumped storage simultaneously. Watch whether the floor holds, whether it holds through incremental borrowing, and whether management explains the trade-off explicitly rather than pointing at cash-flow headroom.

Does market pricing reform actually compress realised tariffs?

Watch the market-traded share against the 36.1% starting point, and more importantly watch revenue per kilowatt-hour.¹ Watch, too, whether disclosure improves — whether CYPC begins quantifying tariff sensitivity rather than describing market reform as an opportunity to convert flexibility value into competitiveness. A company confident in its pricing position should be willing to show the arithmetic.

Is the next major capital deployment priced like Peru or like EDP?

Peru was a competitive process, executed at a defensible price, and has been operated to a rating above its sovereign.¹¹¹ The parent's Portuguese bid was mispriced, misjudged politically, and failed in public.¹⁵ Both templates exist inside the same organisational family. The next material deal — overseas acquisition, large pumped storage commitment, or something not yet visible — will reveal which one is dominant.

Does another 2022 arrive, and how is it explained?

Multi-year drought is the tail risk that no dispatch optimisation can offset. When it recurs — and over a hundred-year asset life it will — the informative question is not the size of the miss. It is whether management quantifies the hydrological shortfall precisely, distinguishes weather from execution, and states what it is doing differently, or whether the explanation stops at "below-normal inflow."

And the number that sits above all of them

Six stations. Roughly 300 billion kilowatt-hours a year. Every other metric in this story — the dividend, the leverage capacity, the multiple, the ability to fund a next act — is downstream of that single figure, and that figure is downstream of the rain.

For twenty years, China Yangtze Power could grow through a dry spell by buying another dam. That option is gone. What remains is the largest, cleanest, most protected generating asset on the planet, an unusually credible commitment to hand most of its cash back to shareholders, and a management team about to find out whether it is as good at allocating capital as it has been at operating concrete.

The river will keep flowing. The interesting question is what gets done with it.

China Yangtze Power Co., Ltd.

Table of Contents

China Yangtze Power: The Company That Owns the World's Biggest Dam

I. Introduction & Episode Roadmap

II. Born From a Dam: Founding and the 2003 IPO

III. Becoming the River: The 2009 Full Listing and the 2015–2016 Cascade Buy-In

IV. The Final Piece: Wudongde, Baihetan, and the End of the Easy Growth Runway

V. How a Dam Prints Cash: The Business Today

VI. Crossing Borders: Peru, Portugal, and the M&A Report Card

VII. Who's Running the River: Current Management, Ownership, and Capital Allocation

VIII. The Moat: Competitive Landscape, Five Forces, and the Cornered Resource

IX. Current Risk Radar

X. Bull Case, Bear Case, and the Playbook

XI. Epilogue: What to Watch

References