GD Power Development: The Thermal Engine and Clean Energy Pivot of China's Energy Colossus

I. Introduction & Episode Roadmap

On the morning of June 26, 2026, in a steep gorge in the Aba Tibetan and Qiang Autonomous Prefecture of western Sichuan, water that had been rising behind a wall of rock and clay for the better part of two years finally did what it was engineered to do. It turned a turbine.

The wall is 315 metres high—the tallest dam of any kind on Earth, a rockfill embankment with a gravel-clay core, built over eleven years in a seismically active valley at the headwaters of the 大渡河 Dadu River.¹ The first of its generating units synchronised to the grid that day, marking the initial phase of a 2,000-megawatt station designed to produce roughly 7.7 billion kilowatt-hours a year on its own.¹ Yet the physical machine is only part of the story. 双江口水电站 Shuangjiangkou Hydropower Station functions as a battery made of water: a multi-year regulating reservoir positioned at the head of a cascading series of hydropower plants, engineered to store summer floodwaters and release them during dry winter months when power is scarce and expensive. Its operator estimates that this storage capability alone will add about 6.6 billion kilowatt-hours annually to the output of plants downstream—generating electricity through existing turbines at essentially zero incremental capital cost.¹

The corporate entity capturing the bulk of that cash flow is 国电电力发展股份有限公司 GD Power Development Co., Ltd., listed in Shanghai under ticker 600795. Yet massive as it is, the Shuangjiangkou dam represents only a fraction of GD Power's overall operations.

At the end of June 2026, GD Power controlled 130,291 megawatts of generating capacity across 30 Chinese provinces, municipalities, and autonomous regions.² That total of roughly 130 gigawatts exceeds the entire installed capacity of national grids such as France, Brazil, or the United Kingdom. Of that portfolio, 83,023 megawatts was thermal—overwhelmingly coal—while hydro, wind, and solar accounted for 47,268 megawatts, representing a non-fossil share of 36.28%.² In the 2025 financial year, the company generated 467.4 billion kilowatt-hours of electricity and recorded RMB 170.24 billion of revenue (approximately $24 billion).³

Yet the underlying financial picture is more complex than the scale of its generating assets suggests. In 2025, GD Power's net profit attributable to shareholders fell 27.15% to RMB 7.16 billion.³ Earnings contracted further in the first half of 2026, falling 18.25% year-over-year to RMB 3.01 billion, even though the company generated 6.25% more electricity than a year earlier.² Furthermore, of the RMB 13.73 billion in total group net profit generated in 2025, only RMB 7.16 billion belonged to shareholders of the listed entity.³ Nearly half was distributed to minority partners.

These dynamics outline the central questions surrounding the company's trajectory:

How did the 2002 breakup of 国家电力公司 State Power Corporation of China establish five state generation champions, and why did GD Power—originally a Dalian district-heating enterprise—become the flagship listed vehicle for one of them? What did the 2017 consolidation of 神华集团 Shenhua Group and 中国国电集团 China Guodian Corporation into 国家能源投资集团 CHN Energy alter about the underlying economics of coal-fired power? While GD Power procures 97% to 99% of its coal through long-term contracts to reduce fuel expenses, does this represent a permanent competitive advantage or a regulatory policy structure subject to revision? How is the coal fleet adapting as state power market reforms transition compensation from pure volume generation to capacity availability? Is the 大渡河 Dadu River concession a high-margin strategic asset or a capital-intensive project that has absorbed RMB 43.8 billion on a single dam? And why, during a period of declining profits, is the company increasing its shareholder payout ratio?

The analysis begins where China's modern power sector originated: with a national grid unable to supply enough electricity to keep its factories running.

A note on the analytical framework: GD Power's financial and operational disclosures are unusually detailed for a listed Chinese enterprise. Every six months, the company discloses segment-level capacity, generation, realized tariffs, utilization hours, net profit, and capital expenditure, alongside coal purchase volumes and unit costs. That granularity challenges simplified narratives, as the detailed numbers frequently diverge from broad market assumptions. Where data conflicts with conventional framing, this analysis follows the empirical evidence.

---

↑ Back to Top

II. Genesis & Historical Context: Unbundling China's Electric Grid

For most of the twentieth century, Chinese electricity was not a business; it was an administrative allocation.

Under the Ministry of Electric Power and its successor, 国家电力公司 State Power Corporation of China—formed in 1997 as a single vertically integrated monopoly owning generation, transmission, and distribution—power plants were built where the central plan dictated, operated when scheduled, and sold output at prices set by decree.⁴ Economic dispatch merit orders, marginal cost pricing, and market signals did not exist; there were only production quotas. Throughout the 1980s and 1990s, as Chinese industrial growth consistently outpaced official forecasts, this rigid planning produced chronic electricity shortages, forcing factories onto three-day workweeks and making rolling brownouts a routine operational hazard.

By the conventions of Western utility investing, the system operated without standard financial metrics. Power plants did not compete for dispatch, fuel hedging was irrelevant under state-allocated coal prices, and capital moved directly from the state budget without explicit return expectations. Traditional utility concepts like regulated rate bases, rate cases, and allowed returns on equity were absent. The system's sole mandate was relieving physical power deficits through bureaucratic expansion.

GD Power did not originate as a national utility in Beijing. It began in the northeastern port city of Dalian as a municipal district heating and cogeneration enterprise. Approved by Liaoning province's economic restructuring commission, the company was incorporated in December 1992 as 大连东北热电发展股份有限公司 Dalian Northeast Thermal Power Development Co., Ltd. through a targeted share placement. It listed on the Shanghai Stock Exchange on March 18, 1997 under ticker 600795, trading initially as 东北热电 Northeast Thermal Power.⁵ Its corporate registration remains anchored in the Dalian Economic and Technological Development Zone, a structural artifact preserved in annual reports that otherwise detail a national enterprise directed from Beijing's Chaoyang District.³

In January 2000, State Power Corporation restructured the Dalian company, acquired a controlling interest, and renamed it GD Power Development.⁵ The enterprise was transformed from a regional operator into a national vehicle by administrative directive—establishing a pattern of state-driven reorganization that recurred throughout its history.

The turning point for China's electric power industry arrived two years later.

In 2002, the State Council issued the Electric Power System Reform Plan—国发〔2002〕5号, known industry-wide as Document No. 5—and unbundled the national power monopoly.⁴ The core mandate rested on four key characters: 厂网分开, signaling the structural separation of power generation from grid networks.⁴ Transmission and distribution assets were split between two monopoly grid operators, 国家电网 State Grid Corporation of China and 中国南方电网 China Southern Power Grid. Power generation assets were divided among five newly created central state groups known as the Big Five: 华能 Huaneng, 大唐 Datang, 华电 Huadian, 中国电力投资 China Power Investment (later 国家电投 SPIC), and 中国国电 China Guodian.⁴ Document No. 5 outlined broader reform objectives, including the separation of auxiliary businesses, structural unbundling of transmission from distribution, and a planned transition to competitive on-grid power bidding—a market design that ultimately took two decades to fully implement.⁴ Assigned to China Guodian, GD Power became the group's flagship listed subsidiary and primary vehicle for accessing public equity markets.⁵

The policy rationale behind establishing the Big Five was competition by benchmark. Five national state groups, evaluated against one another on operational efficiency, heat rates, and capital deployment, were intended to drive performance improvements that a single utility could not achieve. In practice, the structure sparked an aggressive capacity expansion race. Possessing similar access to state banking capital, standardized thermal technology, and identical policy incentives, the five generators focused heavily on expanding total wattage—the key metric governing state performance evaluations. Over the subsequent fifteen years, China expanded its national power grid at an unprecedented rate, largely driven by rival state generators competing for market share.

During the post-reform expansion, GD Power operated under a conventional state-owned asset injection model. China Guodian developed or acquired power generation assets, permitted and constructed the facilities, and transferred operational plants into GD Power once cash flows stabilized. The listed subsidiary then leveraged its public balance sheet to raise equity and debt, funding subsequent asset acquisitions. Through the 2000s, this mechanism expanded GD Power's thermal fleet across northern and eastern China, while China Guodian simultaneously established early hydro development positions in Sichuan Province.

The 2002 reforms created structural competition in generation, but stopped short of introducing market pricing for power output. On-grid electricity tariffs remained strictly controlled by 国家发展和改革委员会 the National Development and Reform Commission (NDRC). Concurrently, domestic coal prices were progressively deregulated and exposed to market dynamics. As a result, the Big Five generators operated facilities with state-regulated revenue caps while purchasing their primary fuel input at market-floating spot prices.

This structural mismatch created severe earnings volatility across the sector. During economic expansions, generators added thermal capacity rapidly to meet state growth targets and capture volume. When global coal prices surged—most notably in 2008 and 2011—administrative tariff adjustments lagged rising fuel costs, pushing power generators into net operating losses despite rapid national demand growth. Consequently, the power sector faced high operational leverage combined with fixed top-line pricing, exposing balance sheets to unhedged commodity risk.

Unlike traditional utilities in developed markets that operate under fuel-cost pass-through mechanisms and guaranteed returns on rate bases, Chinese thermal power producers absorbed full commodity fuel risk without capital cost protections. Functionally, GD Power and its peers operated as merchant generators subject to capped realization prices and variable fuel costs. This asymmetric model suppressed returns on equity across the sector, causing industry profitability to fluctuate wildly across commodity cycles.

For GD Power, the strategy between 2002 and 2015 focused primarily on absorbing thermal assets injected by China Guodian across northern and eastern China and expanding existing plant capacity. However, one long-term capital decision diverged from this thermal focus. In Sichuan, China Guodian secured exclusive development rights across the main stem of the Dadu River—one of China's thirteen designated national hydropower bases. Originally viewed as capital-intensive, long-payback infrastructure, these Dadu River rights eventually evolved into GD Power's highest-margin generating assets.

By 2016, chronic fuel price volatility and systemic overcapacity rendered the traditional thermal generation model increasingly unsustainable. Rather than implementing immediate wholesale tariff deregulation, state planners responded with structural consolidation.

↑ Back to Top

III. The Inflection Point: The 2017 CHN Energy Mega-Merger

By early 2017, Chinese thermal power generators faced stark operational arithmetic.

Coal prices had rallied sharply through 2016 as Beijing's supply-side capacity cuts collided with steady electricity demand. Because on-grid tariffs were set administratively, they did not adjust to absorb higher fuel expenses. Across the Big Five generators, income statements reflected identical pressures: revenue remained flat, fuel costs rose steeply, and operating margins compressed toward zero. This imbalance was a structural defect in market design rather than an operational failure. Resolving it required a system-level realignment beyond the capability of any single generator: consolidating coal mining and power generation onto a unified balance sheet, turning inter-company fuel pricing into an internal management allocation rather than an external market exposure.

Understanding this mechanic requires examining how thermal coal is priced in China. The primary benchmark is the spot assessment for 5,500 kcal thermal coal delivered at Qinhuangdao, the northern transshipment hub connecting Inner Mongolian and Shanxi mines to coastal load centers. That index fluctuates based on mine safety inspections, rail freight allocations, port inventory levels, southwestern hydrology, and seasonal weather patterns in Guangdong. A generator purchasing coal at spot rates carries a cost structure resembling a volatile commodity derivative. Conversely, a generator buying under annual contracts within state-supervised price bands operates under a predictable cost baseline. Across a fleet consuming over one hundred million tonnes of coal annually, that structural cost differential formed the strategic rationale for mega-merger consolidation.

On August 28, 2017, 国务院国有资产监督管理委员会 the State-owned Assets Supervision and Administration Commission (SASAC) announced that China Guodian Corporation and Shenhua Group would merge to form 国家能源投资集团 CHN Energy, reducing the number of centrally administered state enterprises to 98.[^6] Shenhua brought the world's largest coal mining operation alongside integrated heavy-haul railways and ports, while Guodian contributed a vast national generation fleet. Upon formation, the combined entity stood as the world's largest coal producer and one of its largest thermal power operators.⁶

The scale of the integrated group remains unprecedented. By 2024, CHN Energy was producing approximately 620 million tonnes of coal annually, selling roughly 850 million tonnes, and generating 1.26 trillion kilowatt-hours of electricity—accounting for about one-sixth of national coal output and one-eighth of national power generation.⁷ The group holds over four billion tonnes of approved coal capacity and operates an extensive heavy-haul rail network, including the 神朔铁路 Shen-Shuo Railway and its extensions, moving coal from mine-mouth to port to power plant without relying on spot market logistics.⁷

This captive logistics architecture provides a crucial structural moat. In standard coal supply chains, rail transport introduces third-party margin extraction and capacity bottlenecks that can turn price inflation into physical fuel supply constraints. Shenhua constructed its rail network to eliminate third-party logistics risk, allowing the merged group to move coal from pit to power station entirely within an internal operating network. When GD Power's filings refer to an integrated "coal-power-rail-port-shipping" supply chain, they describe a physical asset corridor spanning from mine seam to boiler under unified ultimate ownership.²

For GD Power's public shareholders, the practical value of this parent-level integration lies in fuel supply security rather than subsidized pricing. GD Power does not receive unpriced coal; it secures 长协煤 long-term contract coal under the National Development and Reform Commission's (NDRC) medium-and-long-term agreement framework, which enforces price caps and volume commitments. In 2025, GD Power purchased 188 million tonnes of raw coal, with 184 million tonnes—or 97.73%—secured through long-term contracts.³ In the first half of 2026, that long-term contract share reached 99%.² Management refers to this contract framework as a 压舱石 (ballast stone), attributing the structural protection to an integrated "coal-power-rail-port-shipping" arrangement rather than informal state favor.²

The necessity of this supply architecture became clear during the commodity shocks of 2021 and 2022. Thermal-heavy peers without captive mining support—such as 华能国际 Huaneng Power International—faced severe exposure to spot coal price inflation. GD Power was not immune: in January 2022, the company guided to a 2021 net loss attributable to shareholders of between RMB 1.6 billion and RMB 2.3 billion—a decline of 138% to 154% year-over-year—citing an into-furnace standard coal unit price of roughly RMB 890 per tonne, up nearly RMB 300 from the prior year.[^9] Every flagship listed generator among the Big Five recorded net losses that year.[^9]

The resulting supply crisis marked the moment state planners recognized that price design itself was destabilizing the grid.

In September 2021, more than twenty Chinese provinces instituted electricity rationing. Manufacturing hubs in Guangdong, Jiangsu, and Zhejiang restricted factory operations to three days per week or forced production into off-peak night shifts, while municipal grids in the northeast experienced sudden blackouts. The disruption stemmed not from physical capacity shortages—China possessed ample thermal generation capacity—but from tariff rigidity. Spot coal prices had escalated to levels where thermal generation produced severe operating losses, prompting generators to minimize output. The nation's power fleet faced rationing because administrative price controls suppressed dispatch incentives.

Beijing responded on October 12, 2021, when the NDRC issued Notice 1439, which took effect three days later. The directive expanded the allowed tariff fluctuation band around benchmark coal power prices to 20% in either direction—replacing a 10% cap on upward adjustments—and mandated that all commercial and industrial users enter the power market directly rather than purchasing at government catalogue rates.⁸ Tariffs for residential and agricultural customers remained strictly protected.⁸

While Notice 1439 immediately allowed generators to pass through fuel inflation, its structural legacy was a permanent shift toward market pricing. Pushing commercial and industrial buyers into wholesale markets converted non-residential demand into market-based trading volume. Consequently, GD Power's market-traded volume share exceeded 93%, meaning that the reform that shielded generators during a fuel spike later exposed their realized tariffs to market pressure during periods of surplus.² Policy mechanisms engineered for commodity scarcity produce fundamentally different margin dynamics during supply abundance.

Thus, integration with CHN Energy alters the structure of operational risk rather than eliminating it. Long-term contract books transform volatile spot input costs into predictable cost baselines, shifting management focus to contract compliance and fuel quality. The arrangement functions as a variance-reduction mechanism rather than a total cost shield.

The 2017 consolidation also addressed long-standing corporate governance constraints. Prior to the merger, Guodian and Shenhua operated overlapping thermal assets in shared regional markets, creating a 同业竞争 (horizontal competition) conflict under Chinese listing rules. To resolve these overlaps pursuant to regulatory requirements, GD Power announced a major asset restructuring on August 28, 2017. Under the agreement, GD Power and 中国神华 China Shenhua Energy injected their overlapping thermal power assets across Shanxi, Inner Mongolia, Liaoning, Jiangsu, Zhejiang, Anhui, Ningxia, and Xinjiang into a joint venture. GD Power contributed assets valued at approximately RMB 37.37 billion against China Shenhua's RMB 29.27 billion, securing operational control of the combined entity.⁹

This structure explains a persistent feature of GD Power's financial statements: controlling capacity significantly exceeds equity-weighted attributable capacity. At the end of June 2026, GD Power controlled 130,291 megawatts of capacity, but its equity-weighted capacity stood at 75,189 megawatts—just 58% of the total—while its equity share in coal-fired assets was only 44%.² This structural gap explains the widening divergence between group net profit and attributable shareholder profit, and highlights why the ongoing resolution of the parent's horizontal-competition commitments remains central to future asset injection prospects.²

This corporate structure imposes clear operational constraints alongside strategic benefits. When a controlling parent holds competing assets, regulatory frameworks require formal commitments to rectify overlaps. GD Power's corporate filings explicitly highlight the continued implementation of this horizontal-competition resolution as a primary mechanism for future asset injections.² This arrangement provides access to operational generating assets without greenfield construction risk, though the valuation, asset composition, and execution timing remain governed by the controlling shareholder.

↑ Back to Top

IV. Inside the Thermal Power Core: Economics, Coal Contracts, & Capacity Tariffs

Here is the simplest possible model of a coal-fired power plant, and it is worth building slowly because almost everything that follows is a variation on it.

You buy a tonne of coal. You burn it to make steam, the steam spins a turbine, the turbine makes electricity, and you sell that electricity into the grid at a set price. Your margin per kilowatt-hour is the tariff you receive minus the fuel you consumed to produce it. The efficiency of that conversion is measured in grams of standard coal per kilowatt-hour delivered — 供电煤耗 — and it is the single cleanest measure of whether your machines are any good. GD Power's coal fleet averaged 293.85 grams per kWh in 2025 and 292.80 grams in the first half of 2026.³² To put that in perspective: the global average for coal generation sits meaningfully above 300 grams, and every gram saved across a 80-gigawatt fleet is real money.

GD Power gets there by running very large, very hot machines — 79 units of 600 MW or more, representing 73.86% of coal capacity at end-2025, and 26 units of 1,000 MW or more.³ Bigger, higher-pressure units are simply more thermodynamically efficient, and China has more of the newest ones than anywhere else on Earth.

Now the two prices.

On the cost side, 2025 was a very good year. GD Power's into-furnace standard coal unit price fell to RMB 825.06 per tonne, down RMB 97.11, a decline of 10.53%.³ The company burned 112 million tonnes of standard coal equivalent, so that per-tonne saving is worth roughly RMB 10 billion at the gross level — an enormous number against a shareholder net profit of RMB 7.16 billion. National coal supply was ample, imports fell from historic highs, and prices declined through mid-year before recovering.³

On the revenue side, 2025 was a bad year, and this is the part the bull case tends to underweight. GD Power's average on-grid tariff across all segments fell to RMB 400.90 per megawatt-hour, down RMB 28.92 from 2024 — a decline of about 6.7%.³ Market-traded volume reached 407.6 billion kWh, or 91.73% of total on-grid electricity.³ The premium the company earned on market-traded power over benchmark collapsed to RMB 32.10/MWh from RMB 56.24/MWh a year earlier, and coal units were priced 11.73% above the local benchmark tariff.³

A word on what "market-traded" means, because the phrase conceals a lot of structure. The overwhelming majority of that volume is not spot trading in the Western sense. It is 中长期 medium-and-long-term contracting: annual and monthly bilateral agreements between a generator and industrial buyers or retailers, negotiated province by province, typically settled against a benchmark tariff with an agreed premium or discount. Layered on top is a spot market that has now gone live across most provinces, clearing at fifteen-minute or hourly intervals, where the marginal price collapses when solar output peaks at midday and spikes on winter evenings. A generator's realised price is therefore a blend of contracted volume at negotiated prices and a residual settled at spot.

GD Power's disclosure that its market premium over benchmark halved in a single year is a measure of how much bargaining power moved from sellers to buyers as the system swung from tight to loose.

Read those two paragraphs together and you have the central tension of Chinese thermal power in the mid-2020s: fuel deflation is being handed straight back to customers through market-clearing prices. GD Power captured a RMB 97/tonne coal saving and gave back a RMB 29/MWh tariff. The coal segment still earned group net profit of RMB 9.80 billion in 2025, of which RMB 4.39 billion was attributable to shareholders.³ But the mechanism is not "cheap coal equals profit." It is "cheap coal equals a temporarily wider spread that the market progressively competes away."

The first half of 2026 tested that reading, and confirmed it in the least comfortable direction. Coal prices turned up — the into-furnace unit price rose to RMB 836.84 per tonne, a modest 0.64% increase — while the average tariff fell another RMB 16.40 to RMB 393.31/MWh and market-traded volume climbed to 93.72% of output.² Coal-segment group net profit fell to RMB 3.81 billion for the half.² Fuel costs up slightly, prices down materially: the spread compressed from both ends simultaneously. Management's own explanation of the revenue line is refreshingly unvarnished — volume rose faster than price fell, which is why revenue grew 1.77% while profit fell.²

There is also a volume problem underneath the price problem. Coal fleet utilisation fell to 4,850 hours in 2025 from 5,207 the prior year, and to 2,119 hours in the first half of 2026 from 2,155.³² The cause is not weak demand — Chinese electricity consumption grew 5.0% in 2025 to 10.37 trillion kWh, the first time any country has crossed ten trillion, and grew another 5.3% in the first half of 2026.³² The cause is that new energy took the growth. In 2025, wind and solar accounted for 80.2% of all new capacity added nationally, and new energy supplied 97.1% of the increment in national electricity consumption.³ By June 2026, coal generation had fallen to 49.7% of national output — below half for the first time.²

This is the mechanism that makes the 容量电价 capacity tariff mechanism not a bonus but a necessity.

The idea is straightforward once you see the problem it solves. If a coal plant only gets paid per kilowatt-hour, and the system increasingly wants it to sit idle most of the time and then ramp hard when the wind drops at 7pm in January, the plant will be uneconomic long before it becomes unnecessary. So on November 10, 2023, the NDRC and National Energy Administration established a capacity payment: from January 1, 2024, compliant public coal units receive a fixed annual payment per kilowatt of declared capacity, set against a nationally standardised fixed-cost benchmark of RMB 330 per kilowatt per year.¹⁰ Initially the mechanism recovered roughly 30% of that in most provinces — about RMB 100/kW/year — with around 50% in regions transitioning fastest away from coal. The payment is not free money: units that fail dispatch instructions face escalating deductions, up to 100% for four or more failures in a month.¹⁰

On January 30, 2026, the NDRC and NEA extended and generalised the framework. The new notice covers coal, gas, pumped storage and — for the first time at national level — grid-side battery storage, and lifts the coal fixed-cost recovery ratio to no less than 50%.¹¹ GD Power's own interim report states the 50% floor plainly.²

What is this actually worth? Applied across roughly 81 gigawatts of coal capacity, the step from 30% to 50% of a RMB 330/kW benchmark is on the order of RMB 5 billion of annual gross capacity revenue at the group level — against 2025 group net profit of RMB 13.7 billion, material rather than transformative, and diluted roughly in half at the shareholder line by minority interests. GD Power does not disclose capacity fee income as a separate revenue line, which is a genuine disclosure gap for anyone trying to model the transition; the annual report references it only qualitatively, alongside ancillary-service income from primary frequency regulation, AGC and reactive power compensation.³

The honest analytical framing is this. Capacity payments do not make thermal power more profitable in aggregate; they change the composition of thermal earnings from volume-linked and volatile toward fixed and annuity-like. That is a valuation-multiple story more than an earnings story, and it only pays off if energy-market revenue does not fall by more than the capacity payment rises. In 2025 and the first half of 2026, at GD Power, it did.

There is a third revenue stream growing quietly alongside energy and capacity, and it is the one most consistent with where the system is heading. Ancillary services — primary frequency regulation, automatic generation control, reactive power compensation — pay a plant for the quality of its response rather than the quantity of its output. To earn them, a coal unit has to be able to turn down deeply and ramp back fast, which is not what a 1,000 MW ultra-supercritical boiler was designed to do. GD Power has been retrofitting for exactly this under a programme it calls 三改联动, the coordinated three-way retrofit covering efficiency, heating extraction and flexibility, and management reports that multiple units can now run at 20% of rated output in deep peaking mode.³ A unit that can idle at a fifth of capacity and then climb to full load quickly is a fundamentally different product from a baseload machine, and it is the product the new system is buying.

Heat is the other underappreciated line. GD Power supplied 204 million gigajoules of heat in 2025 and 116 million in the first half of 2026, from cogeneration plants selling steam and hot water to industrial parks and municipal networks.³² Heat contracts are regulated, sticky, seasonal and essentially non-competitive — a district heating customer cannot switch suppliers — and they confer must-run status on certain plants regardless of what the power market is doing. It is the closest thing in the portfolio to a switching cost.

It is worth naming the peer set explicitly, because thermal power in China is not a differentiated product and the comparison is the analysis. Huaneng Power International is the largest listed thermal operator; Huadian Power International and Datang International Power Generation are the other Big Five flagships with heavy coal exposure; 浙能电力 Zheneng Electric Power is the strongest provincial champion, concentrated in the high-tariff Zhejiang market. All of them buy similar turbines, meet the same emissions standards, sell into the same provincial market designs, and fall under the same capacity mechanism. What differs is fuel sourcing, fleet age, provincial mix and equity structure — which is precisely why the coal contract ratio and the equity-versus-controlled capacity gap carry so much of the information in GD Power's case.

Which brings us to the part of the portfolio where the fuel is free.

---

↑ Back to Top

V. The Crown Jewel: Dadu River Hydropower Cascades

The Dadu River drops more than four kilometres from the eastern edge of the Tibetan Plateau to its confluence with the Min River near Leshan. That vertical drop, compressed into a few hundred kilometres of steep gorge, makes it one of China's thirteen national hydropower bases—a natural monopoly forged by topography and gravity that cannot be duplicated by competitors.

GD Power controls this river system through 国能大渡河流域水电开发有限公司 CHN Energy Dadu River Hydropower Development Co., Ltd., in which it holds an 80% equity stake, with the remaining 20% held by Sichuan provincial state capital.²¹² As CHN Energy's primary hydropower arm, the Dadu River subsidiary oversees operations along its namesake river and manages preliminary development for the Parlung Tsangpo basin in Tibet, giving it administrative remit over roughly 30 million kilowatts (30 gigawatts) of total hydropower resources.¹²

The economics of cascade hydropower stand in sharp contrast to thermal generation. Hydropower carries zero fuel expense; once the initial concrete and steel are financed, the marginal cost per kilowatt-hour drops to routine operation and maintenance. Because a dam's operational life spans decades beyond its accounting depreciation schedule, mature hydro facilities generate substantial cash flow long after their construction debt has been ammortized.

The strategic value of a cascade lies in sequential water reuse. As a river flows downstream, each cubic metre of water passes through multiple turbines, generating electricity six or seven times along its path. Consequently, whichever operator controls the timing of headwater releases dictates the generation schedule and revenue profile for every station downstream. Where ownership is fragmented across multiple operators, coordinated dispatch requires complex inter-company negotiations or state regulation. On the Dadu River, unified ownership allows GD Power to execute 流域集控联调—integrated basin-level control and joint dispatch—which management identifies as a primary driver of operational efficiency.³ By comparison, 长江电力 China Yangtze Power—the industry benchmark operating the Three Gorges and Jinsha River cascades—reported a domestic hydropower gross margin of 65.79% in 2025 and attributable net profit of RMB 34.5 billion across roughly 71.8 gigawatts of capacity.¹³

GD Power's hydro assets have not yet reached those benchmark returns. In 2025, its hydro segment operated 15,130 megawatts of controlled capacity, generated 55.12 billion kilowatt-hours, and produced group net profit of RMB 2.14 billion, of which RMB 1.63 billion was attributable to shareholders.³ That translated to roughly RMB 39 of net profit per megawatt-hour generated—a resilient margin compared to thermal power, but well below the profitability of fully depreciated Yangtze assets. Furthermore, dry hydrology caused hydro generation to fall 7.31% in 2025, pushing fleet utilization down by 298 hours to 3,680 hours.³ The segment's average realized tariff of RMB 236.70 per megawatt-hour was the lowest across GD Power's portfolio, sitting roughly 44% below its average coal tariff and reflecting Sichuan's structural wet-season power surpluses.³

Despite these tariff pressures, the Dadu River plants hold strategic value for the regional grid. Accounting for roughly a quarter of all provincially dispatched hydropower capacity in Sichuan, and situated near the primary Chengdu demand center, the cascade functions as the province's primary peaking and frequency-regulation asset rather than a simple bulk power provider.³ That operational role provides long-term pricing support, as dispatchable flexibility commands higher value than unconstrained wet-season volume.

Two structural constraints have historically capped segment returns: weather volatility and seasonal mismatch. Hydrological exposure was highlighted during the severe summer drought of 2022. Inflows into major Sichuan rivers dropped over 40% below historical averages in July and more than 60% by early August. Water levels at the 3,600-megawatt 瀑布沟 Pubugou reservoir—the flagship station on the lower Dadu—fell roughly 30 metres to historic lows, cutting total provincial daily hydro output in half from 900 million to 450 million kilowatt-hours.[^15] In a province where hydropower comprises nearly 80% of generating capacity, the drought forced widespread industrial power rationing during a record heatwave.[^15] The event demonstrated that despite strong structural margins, run-of-river hydro cash flows remain vulnerable to severe climate events.

The second constraint is seasonal generation timing. In a standard water year, run-of-river stations produce peak output during summer flood months when system power prices are depressed, while generating minimal power during winter when grid demand peaks.

The commissioning of Shuangjiangkou Hydropower Station directly targets this seasonal imbalance. Acting as the multi-year regulating reservoir at the head of the upper Dadu, Shuangjiangkou stores summer floodwaters for release during dry winter months. This regulated flow increases winter generation at downstream stations—including 瀑布沟 Pubugou, 深溪沟 Shenxigou, 龚嘴 Gongzui, and 铜街子 Tongjiezi—without requiring additional downstream turbine investments.¹ Management projects that this multi-year regulation will add approximately 6.6 billion kilowatt-hours in annual downstream generation—equivalent to roughly 12% of GD Power's total 2025 hydro output—with the gain captured almost entirely during high-tariff dry seasons.¹³

However, converting this regulatory capability into earnings has required substantial capital expenditure. Shuangjiangkou carries an approved total investment of RMB 43.80 billion. By the end of June 2026, the project was 81.21% complete, with RMB 26.48 billion remaining in construction-in-progress following a RMB 9.65 billion asset transfer to fixed assets during the first half of the year.² To fund this multi-year construction, GD Power accumulated RMB 5.56 billion in capitalized interest, including RMB 348 million capitalized during the first half of 2026 at an effective borrowing rate of 2.35%.²

This accounting structure creates an imminent transition for reported earnings. As Shuangjiangkou's generating units enter commercial service, capitalized interest will shift from balance-sheet construction accounts onto the income statement as direct interest expense, alongside new depreciation charges on the expanded fixed-asset base. Consequently, reported hydro segment earnings in 2027 may show initial accounting margin compression even as physical generation and operating cash flows expand.

That commissioning phase is already under way. In the first half of 2026, GD Power added 1,720 megawatts of hydro capacity, comprising 860 megawatts at Jinchuan, 500 megawatts at Shuangjiangkou, 300 megawatts at Zhentouba, and 60 megawatts at Shaping.² Management expects to bring an additional 1,620 megawatts online before the end of 2026.² Controlled hydro capacity reached 16,851 megawatts by late June 2026, generating RMB 1.45 billion in first-half segment net profit on a modest 1.72% increase in generation—a lag reflecting the initial period required to fill new project reservoirs.² Beyond the Dadu basin, GD Power holds a 55.61% controlling interest in the 开都河 Kaidu River basin developer in Xinjiang, where it is integrating conventional hydro with pumped storage to form a regional multi-energy network.²

Pumped storage represents a strategic extension of this flexibility model. A pumped storage facility uses upper and lower reservoirs connected by reversible turbines, consuming low-cost electricity—such as midday solar power—to pump water uphill, and releasing it through turbines during high-demand periods when spot prices peak. Although the process incurs a 25% round-trip energy loss, it generates arbitrage profits from widening peak-to-valley tariff spreads. GD Power is advancing pumped storage projects at Huoshan in Anhui and Hejing in Xinjiang to complement conventional hydro and mitigate solar curtailment within integrated water-wind-solar-storage systems.³²

The expansion pipeline extends further along the main river stem. In 2025, the Danba hydropower project on the Dadu River secured approval from the National Development and Reform Commission (NDRC) as part of a planned 10-gigawatt integrated energy complex, a project elevated to national strategic energy planning status.³²

From a portfolio perspective, GD Power's hydro assets provide an internal operational hedge against its thermal fleet. Drought years in Sichuan frequently coincide with regional heatwaves that drive national power demand, raising thermal utilization and wholesale electricity tariffs just as hydro output contracts. Although this hedge is partial—given that thermal capacity is larger and geographically dispersed—it stabilizes consolidated group earnings across weather cycles.

Ultimately, GD Power's hydro core provides a non-replicable geographic moat, predictable long-term cash flows, and substantial volumetric gains as multi-year storage comes online over the next three years. However, these assets carry lower realized tariffs than thermal generation and bear a balance sheet burden that will temporarily weigh on reported P&L metrics as construction debt converts into depreciation and interest expense.

Among industry peers, the closest operational comparison is the 雅砻江 Yalong River cascade, developed by a joint venture featuring 国投电力 SDIC Power. Like the Dadu, the Yalong relies on single-operator basin management and headwater regulation to maximize system yields. The strategic model is proven across China's power sector; for GD Power, investment returns will depend on how efficiently its balance sheet absorbs the final capital deployment.

↑ Back to Top

VI. The Green Pivot: Wind, Solar, & "Desert-Gobi-Wasteland" Mega Bases

In the Mongolian steppe of western Inner Mongolia, on flat terrain supporting little else, GD Power operates Blue Sea, a 3,000-megawatt photovoltaic station. Management describes its operating indicators as domestically leading and its profitability as excellent.³ It represents an asset class that functions primarily under three concurrent conditions: vast tracts of low-cost land, dramatically lower solar module prices, and ultra-high-voltage transmission corridors capable of wheeling power a thousand miles to eastern load centers.

This deployment forms part of China's 沙戈荒 desert-Gobi-wasteland mega-base program—a state initiative building utility-scale wind and solar complexes across Inner Mongolia, Ningxia, Xinjiang, and Gansu, sized in gigawatts and evacuated through 特高压 ultra-high-voltage grid infrastructure. GD Power has committed heavily to the strategy.

The economics of these mega-bases rely on an equipment cost decline with few historical parallels. As solar module prices dropped to a fraction of their level a decade earlier, the marginal cost of desert solar shifted toward land preparation, mounting structures, inverters, and grid interconnections. Combined with a central state enterprise's low cost of debt—GD Power capitalizes project borrowings at interest rates around 2.35%—the hurdle rate for new solar projects fell sharply.² However, low hurdle rates created an industry-wide dynamic: as capital costs fell and resource access expanded, developer competition migrated directly to realized power tariffs.

The speed of capacity expansion was substantial.

In 2025, the company added 7,265 megawatts of renewable capacity—comprising 6,609 megawatts of solar and 655 megawatts of wind—bringing controlled solar capacity to 18,890 megawatts and wind capacity to 10,495 megawatts.³ Solar generation nearly doubled, rising 93.78% to 21.87 billion kilowatt-hours.³ By the end of June 2026, controlled solar reached 19,743 megawatts and wind reached 10,675 megawatts, with an additional 5,224 megawatts of solar and 1,866 megawatts of wind under construction, concentrated in Tianjin, Sichuan, and Inner Mongolia.² Non-fossil sources reached 36.28% of controlled capacity and 50.72% of equity-weighted capacity—meaning that on an ownership-weighted basis, GD Power's generating portfolio crossed into majority clean energy during the first half of 2026.²

Beyond fleet expansion, GD Power's strategic approach relies on co-locating renewables with its existing thermal fleet. The 上海庙 Shanghai Miao complex in Inner Mongolia serves as the primary operational model, featuring a joint operational design pairing 4,000 megawatts of coal generation alongside 3,000 megawatts of solar on a shared site and transmission corridor.³ The underlying engineering logic is straightforward: standalone solar produces variable output that grid operators must curtail or back up with dispatchable capacity. A solar installation paired with an operational coal plant shares existing grid interconnect capacity while utilizing thermal ramping to absorb generation swings, accelerating grid connection times and reducing curtailment risk. While segment disclosures do not yet isolate the financial returns of co-located assets, the structural model leverages an asset combination few pure-play developers possess.

However, operational data highlights a clear divergence between asset scale and profit contribution.

Renewables represent over a third of GD Power's installed capacity but generate a disproportionately small share of earnings. In 2025, wind and solar combined produced RMB 2.26 billion in group net profit, compared to RMB 9.82 billion from the coal segment, despite operating nearly 30 gigawatts of non-fossil capacity.³ This gap reflects physical utilization limits alongside pricing pressures. Solar fleet utilization reached 1,259 hours in 2025 and 577 hours in the first half of 2026, while wind recorded 2,108 hours and 965 hours over the same periods.³² By contrast, the coal fleet operated for 4,850 hours in 2025. A gigawatt of intermittent solar capacity produces far less electrical energy over a year than a baseload thermal unit.

Earnings in the renewable segments compressed further in early 2026. First-half wind generation fell 9.77% year-over-year as wind utilization dropped 162 hours—the largest utilization decline across the company's generating segments—reducing wind group net profit to RMB 523 million.² Concurrently, solar realized tariffs dropped to RMB 309.89 per megawatt-hour, yielding RMB 285 million in first-half group net profit across nearly 20 gigawatts of installed capacity.² These results reflect modest returns relative to total deployed capital.

Regulatory shifts account for much of this margin compression.

On February 9, 2025, the NDRC and NEA issued Document 136, deepening market-based reforms for renewable power tariffs by phasing out fixed feed-in benchmarks for new projects. Under the rules, assets commissioned before June 1, 2025 retain power purchase support capped at local coal benchmark tariffs, while projects entering service after June 1, 2025 must determine realization prices through competitive provincial auctions.¹⁴ Consequently, baseline revenue for new renewable generation is governed by competitive market bids rather than administrative pricing.

GD Power's stated capital strategy, referenced in its 2025 annual report and 2026 interim filing, centers on the principle 算赢再投 ("invest only when profitability is assured").³² Management states that project approvals are governed by four investment evaluation gates assessing local market tariffs, curtailment risk, and realization trends.² Capital expenditure data reflects this tightened discipline: first-half 2026 preliminary and construction capital spending fell 32.32% year-over-year to RMB 14.34 billion, while new renewable additions slowed to 1,033 megawatts during the half, down from 7,265 megawatts in full-year 2025.² Upcoming disclosures will determine whether this slowdown represents permanent capital rationing or a temporary pause between major mega-base project phases.

Alongside utility-scale renewables, GD Power is developing auxiliary storage and grid-flexibility assets. The company has commissioned a thermal-plus-molten-salt storage project at Suzhou in Anhui, a four-dimensional composite storage system at Penglai, a ten-thousand-tonne low-pressure adsorption carbon capture facility at its Ninghai coal plant in Zhejiang, and a commercial virtual power plant operating through its Shenwan Energy affiliate.³ It is also constructing pumped storage stations at Huoshan in Anhui and Hejing in Xinjiang.³ In the first half of 2026, it brought online a boiler system integrated with flue-gas molten-salt heat storage at Langfang.²

To mitigate wholesale pricing volatility, the company is evaluating 绿电直连 (green power direct connection), establishing dedicated transmission links to industrial off-takers outside provincial grid pools.² If scaled successfully, direct connections could convert merchant renewable assets subject to auction pricing into long-term contracted generation with corporate counterparties, mirroring corporate power purchase agreement structures in international markets. While volumes remain undisclosed and financial contributions are modest, these initiative projects reflect low-cost positioning within a power system increasingly structured around operational flexibility, backed by an organization employing 35,407 people with a 2025 research and development budget of RMB 1.01 billion.³

However, accounting treatments warrant attention when evaluating these technical initiatives: of the RMB 1.01 billion allocated to research and development in 2025, 45.54% was capitalized rather than expensed, while overall R&D spending represented 0.59% of total revenue.³ Capitalizing research costs defers immediate expense recognition onto future periods, requiring analysts to monitor whether capitalized assets generate corresponding long-term returns.

This balance between capital allocation discipline and ambitious asset deployment highlights central questions of corporate governance and operational oversight.

↑ Back to Top

VII. Governance, Capital Allocation, & Current Management

On December 29, 2025, GD Power's board convened its forty-first meeting of the eighth session to process an executive transition.

唐坚 Tang Jian, born in March 1971, a Communist Party member and senior engineer, had been elected chairman eighteen months earlier, on June 13, 2024.¹⁵ He resigned as chairman, director, and chair of the board's Strategy and ESG Management Committee, citing a work transfer—the standard formulation for reassignment within the central state enterprise system.¹⁶ Under the company's articles of association, resigning as chairman also vacated the legal representative role. The board designated 赵世斌 Zhao Shibin, director and general manager, to act as chairman and legal representative until a successor is elected.³¹⁷

As of the 2026 interim report, filed in August 2026, Zhao was still signing as the company's responsible person, alongside chief accountant 刘春峰 Liu Chunfeng, who also serves as board secretary.² Zhao had been appointed general manager on October 11, 2024, at the board's thirty-first meeting, and joined the board later that month at age 55.¹⁸

Public disclosures offer limited personal background on either executive, providing titles, birth years, and Party affiliation rather than personal commentary. Tang was born in March 1971 and holds a senior engineer rank.¹⁵ Zhao holds the general manager and acting chairman roles simultaneously, a concentration of executive authority that the board framed as a continuity measure rather than a permanent arrangement.¹⁷ What can be observed is the operational structure beneath them: 35,407 employees at the end of 2025, with just 215 stationed at the parent company and 35,192 deployed across operating subsidiaries—reflecting a thin corporate center managing a vast distributed asset base across thirty provinces.³ That decentralized structure requires operational autonomy at the regional level, giving practical substance to the "one plant, one strategy" management approach highlighted throughout corporate filings.³

This executive sequence warrants objective scrutiny. GD Power experienced a chairman appointment, a departure, and eight months under an acting leadership arrangement within roughly two years, all while deploying tens of billions of renminbi annually into thirty-year infrastructure assets. While executive rotations reflect standard administrative transfers within central state enterprises rather than governance failure, frequent leadership changes create a potential disconnect in multi-decade capital allocation strategies, as incoming executives inherit commitments made by predecessors.

Executive incentive structures also differ fundamentally from investor-owned utilities. GD Power maintains no equity incentive plan and no employee shareholding scheme.³ Instead, senior executives at central state enterprises are evaluated by SASAC against structural targets—including economic value added, research intensity, clean energy capacity share, financial leverage stability, and state energy security mandates. The dual obligations of state ownership are explicit in operational disclosures: management highlighted fulfilling national energy supply guarantee tasks during major political and civic events, including the Fourth Plenum of the 20th Party Congress, the September 3 military parade, Spring Festival, and the annual Two Sessions.³²

From an investment governance perspective, several structural factors shape GD Power's balance sheet and earnings profile.

First, minority interests dilute reported shareholder returns. Group net profit of RMB 13.73 billion in 2025 yielded RMB 7.16 billion for shareholders of the listed entity.³ This earnings gap is a structural feature of a company expanded primarily through joint-venture vehicles—including the 2017 asset restructuring joint venture, the 80%-owned Dadu River company, the 55.61%-owned Kaidu River developer, and dozens of provincial operating platforms.² As a result, controlled generating capacity metrics exceed attributable equity capacity by approximately 42%.

Second, portfolio complexity limits asset-level transparency. The 2026 interim report lists more than forty separate investee companies receiving capital allocations within a single half-year, spanning wind, solar, hydro, and thermal assets across a dozen provinces, many structured as single-purpose development vehicles.² While standard for Chinese renewable development, this multi-entity structure aggregates hundreds of individual projects within consolidated accounts, offering high-level segment data without disclosing project-level return metrics by vintage or province.

Third, leverage constrains balance sheet flexibility. GD Power's gearing ratio reached 73.45% at the end of 2025, up slightly from 73.40% a year earlier.³ Total liabilities stood at RMB 382.04 billion against RMB 520.15 billion in total assets.³ By mid-2026, long-term borrowings reached RMB 187.48 billion, while bonds payable rose 28.13% to RMB 26.40 billion as the company expanded debt issuance in public bond markets.² While a capitalized borrowing rate of 2.35% on project debt reflects low state enterprise funding costs, elevated leverage leaves limited balance sheet capacity to absorb potential tariff declines.²

Fourth, related-party transactions require structural oversight. Controlling shareholder CHN Energy holds 9,038,709,571 shares, representing a 50.68% equity stake in the listed company.² Because CHN Energy acts as both controlling parent and primary coal supplier, fuel transaction pricing is a central governance consideration. Current disclosures indicate that long-term contract coal is priced within the NDRC's regulated price bands, allowing fuel savings realized in 2025 to flow directly through the income statement to shareholders. Independent auditor BDO China (立信) issued an unqualified audit opinion on the 2025 financial statements.³

Where capital allocation discipline is most evident is in shareholder distribution policy.

In August 2025, the board established a three-year cash dividend plan for 2025 through 2027, committing to a payout ratio of at least 60% of attributable net profit and an annual floor of RMB 0.22 per share.³¹⁹ For 2025, the company paid an interim dividend of RMB 0.141 per share (RMB 2.515 billion) and a final distribution bringing total full-year payouts to RMB 0.241 per share (RMB 4.298 billion), representing 60.02% of attributable net profit.³ GD Power has paid interim dividends for three consecutive years, establishing semi-annual payouts as a normalized mechanism.³ At the annual general meeting on May 14, 2026, shareholders authorized the board to implement a 2026 interim distribution.²

The key analytical point is that management expanded distribution commitments during an earnings contraction. Establishing a floor of RMB 0.22 per share against 2025 earnings of RMB 0.402 per share pre-commits the board to distributing over half of attributable net profit even if earnings decline further.³

However, overall capital intensity remains high relative to cash flow generation. Preliminary planning and construction capital expenditure reached RMB 44.78 billion in 2025 alone, absorbing most of the RMB 53.67 billion in operating cash flow before debt service and dividend payouts.³ Management highlighted capital constraint requirements and cash flow management controls in recent reports, with the first-half 2026 reduction in capital expenditure providing initial evidence of tighter spending discipline.³²

Financial reporting consistency offers additional context for evaluating corporate disclosures. GD Power's strategic positioning and risk disclosures remained consistent between the 2025 annual report and the 2026 interim filing.³² Furthermore, management attributed the 27.15% earnings decline in 2025 to non-recurring baseline effects from 2024—specifically a single-off gain from selling a 50% equity stake in an Inner Mongolian subsidiary and a construction-in-progress impairment at the Daxingchuan station—noting that net profit adjusted for non-recurring items grew 45.23% year-over-year.³

Across the broader central state generator peer group, Huadian Power International earned RMB 6.07 billion in attributable net profit in 2025 on 77.92 gigawatts of controlled capacity, while Datang International Power Generation reported a 63.92% increase in attributable profit as non-fossil capacity reached 42.99%.²⁰ While GD Power's dividend framework provides an income profile comparable to its state enterprise peers, its primary strategic distinction remains its exclusive operating control over the Dadu River hydropower basin.

What generalises from all of this?

↑ Back to Top

VIII. Playbook: Business & Investing Lessons

Lesson 1: In a commodity business with an administered output price, controlling the input is worth more than running the machines better.

GD Power's coal fleet is efficient, but so are the fleets operated by major peers like Huaneng and Huadian. Ultra-supercritical generation technology is widely available to any enterprise with capital. What is not widely available is a parent company that mines 620 million tonnes of coal annually and owns an integrated heavy-haul rail network. The 97.73% long-term contract compliance rate is the operational mechanism through which parent-level asset ownership becomes an income statement advantage.³⁷ Hamilton Helmer would classify this as a cornered resource—though the precise framing matters. The cornered resource is not coal itself, which is physically abundant. It is privileged access to contracted coal at scale, created by a specific corporate integration that competitors cannot replicate without a similar structural merger. The limitation of this advantage is equally precise: it provides variance reduction rather than complete cost elimination, and it was established by policy rather than won in a open commodity market. Advantages granted by policy can always be revised by policy.

Lesson 2: When a regulator changes how an asset gets paid, the asset changes category — but check whether the total pie grew.

The capacity tariff mechanism converts a portion of a coal plant's revenue from volumetric, commodity-dependent sales into a fixed, contracted payment stream. In theory, that structural shift justifies a higher valuation multiple on earnings, because predictable annuities command higher market valuations than volatile commodity spreads. The analytical trap lies in assuming the policy reform is purely additive. In GD Power's case, fixed capacity revenues arrived precisely as market-based energy tariffs declined, resulting in a net negative financial impact through 2025 and the first half of 2026. The reform is best evaluated as financial compensation for a business model undergoing deliberate restructuring—repositioning coal generation from a primary energy source to a strategic grid-balancing resource, where capacity payments make the operational transition survivable rather than structurally high-margin.

Lesson 3: Basin exclusivity is real process power, and it is expensive.

A competitor cannot construct a rival cascade along the same river. Holding exclusive development rights across the Dadu River allows GD Power to capture every downstream generation benefit produced by upstream reservoir regulation, enabling a single headwater dam to add 6.6 billion kilowatt-hours annually across existing downstream turbines.¹ That represents a genuine, non-replicable, compounding operational advantage. However, securing that position requires immense capital commitment, as evidenced by RMB 26.5 billion in construction-in-progress on a single dam project and RMB 5.56 billion in capitalized interest accrued during development.² Cornered resources in heavy civil infrastructure are acquired through multi-year negative free cash flows, with financial returns materializing only after generating units enter commercial service. Investors evaluating this strategic moat must be prepared to look through extended construction accounting cycles.

Lesson 4: A powerful state parent is simultaneously a moat and a claim on your cash flow.

CHN Energy provides GD Power with lower borrowing costs, contracted fuel supplies, transmission corridor access, and a pipeline of potential asset injections under its horizontal-competition commitment.² Concurrently, CHN Energy functions as a controlling shareholder with 50.68% voting control, operating under broad state mandates that encompass national energy security, regional employment, and emissions reduction targets—objectives that frequently align with public shareholder interests but can occasionally diverge.² Evaluating this relationship requires monitoring observable operational proxies: dividend distribution trends, related-party coal contract pricing, capital expenditure discipline, and project-level economic hurdle rates. Measured against these criteria over recent reporting periods, GD Power's corporate governance execution has demonstrated stronger capital discipline than conventional market stereotypes suggest.

Lesson 5: Watch equity capacity, not controlled capacity.

This principle represents one of the most portable insights across utility analysis. While GD Power's headline operating capacity is reported on a consolidated, controlled basis under standard accounting rules, public shareholders own only the equity-weighted share. The structural gap is substantial: 130 gigawatts of controlled capacity versus 75 gigawatts of attributable equity capacity, and 83 gigawatts of controlled coal capacity against 37 gigawatts of equity coal capacity.² Growth strategies built around joint ventures—including the asset restructuring that created GD Power's main thermal operating vehicle—mechanically widen this divergence. Evaluating reported metrics on an equity-ownership basis reveals a nuanced picture: GD Power's attributable capacity mix is actually cleaner than its controlled portfolio, with non-fossil sources exceeding half of total equity capacity.² Nevertheless, the underlying earnings arithmetic remains unyielding, as roughly half of group net profit is distributed to minority partners before reaching attributable equity holders.

The final test is whether these strategic foundations hold against emerging market realities.

---

↑ Back to Top

IX. Analysis, Risk Radar & Bull vs. Bear Stress Test

The Powers, Honestly Assessed

Before applying analytical frameworks, three consensus market claims warrant a direct empirical check, as each pairs a valid premise with a conclusion unsupported by disclosed data.

Myth: "Cheap contracted coal makes GD Power a structurally profitable thermal generator." The underlying premise is well evidenced: unit fuel costs fell 10.53% in 2025 while the long-term contract compliance ratio held near 98%.³ However, the conclusion fails because fuel savings were not retained on the balance sheet. Realized electricity tariffs declined 6.7% in 2025 and fell by another RMB 16.40 per megawatt-hour in the first half of 2026, causing attributable net profit to contract in both periods.³² Long-term coal contracts determine how much fuel price volatility GD Power absorbs; they do not dictate the price at which wholesale electricity markets clear.

Myth: "The clean energy pivot is well advanced." Non-fossil sources reached 36.28% of controlled generating capacity by mid-2026 and exceeded 50% of equity-weighted capacity.² While these operational milestones are genuine, wind and solar combined generated RMB 2.26 billion in group net profit in 2025 compared to RMB 9.82 billion from coal, and produced RMB 808 million combined in the first half of 2026.³² The energy transition remains a capacity and balance-sheet milestone rather than an earnings engine, and current segment economics suggest that financial realignment will occur gradually.

Myth: "Shuangjiangkou transforms the company." The project transforms basin regulation rather than corporate financial scale. Adding roughly 6.6 billion kilowatt-hours of annual downstream generation against 2025 baseline hydro output of 55.1 billion kilowatt-hours provides a meaningful operational boost to a segment that produced RMB 1.63 billion in attributable net profit—less than a quarter of total shareholder earnings.¹³ Relative to consolidated net profit of RMB 7.16 billion attributable to shareholders, an upgraded hydro fleet serves as a solid earnings contributor rather than a fundamental corporate restructuring.

Evaluating GD Power against Hamilton Helmer's competitive framework indicates that two strategic powers remain strong, two are partial, and three do not apply:

Cornered resource — strong. The Dadu River development concession is geographically unique and protected by topography. Concurrently, the parent company's captive coal-and-logistics supply chain provides structural fuel access that unintegrated listed generators cannot replicate.

Scale economies — moderate. Operating a 130-gigawatt fleet allows GD Power to procure equipment efficiently, while geographic coverage across 30 provinces reduces localized market risk—a operational factor highlighted by management as provincial market rules diverge.² However, scale does not confer pricing power, as tariffs are governed by state market design. Peer comparison underscores this limitation: Huaneng Power International operates approximately 156 gigawatts without achieving structurally superior margins, while Huadian Power International generated RMB 6.07 billion in attributable net profit in 2025 on 77.9 gigawatts of capacity—approaching GD Power's RMB 7.16 billion on a far smaller capacity base.²⁰³

Process power — emerging. Basin-wide cascade dispatch, co-located coal-solar operations at Shanghai Miao, and specialized coal-blending techniques represent real technical capabilities, though disclosed segment figures do not yet demonstrate structural margin expansion over peers.

Counter-positioning — largely absent. While hydro and thermal assets provide an internal operational hedge during dry years, competing central state generators maintain comparable multi-energy generation portfolios.

Switching costs, branding, network economies — not applicable. Commodity electricity carries no brand differentiation or direct consumer switching friction.

Porter, Applied

Rivalry is intense and expanding structurally. Five central state generation groups and regional state champions compete within provincial power markets where exit is restricted. With 93.72% of output cleared through market mechanisms, competitive pressure translates directly into realized tariff compression, as evidenced by the RMB 16.40 per megawatt-hour price decline in the first half of 2026.²

Buyer power is substantial. Grid operators manage physical dispatch, provincial power centers control market design, and industrial off-takers have traded directly in wholesale markets since Notice 1439 took effect on October 15, 2021—a policy that widened coal tariff floating bands to ±20% and moved all commercial users off catalog rates.⁸ While engineered to protect generators during commodity spikes, that flexible pricing band transmits downward pressure during market surpluses.

Supplier power is low, representing GD Power's primary structural protection. Securing 97% to 99% of thermal coal through medium-and-long-term contracts under parent-level agreements limits supplier price extraction.³²

Threat of entry remains negligible. Capital requirements, environmental permitting, grid interconnection quotas, and land constraints prevent new commercial entrants at scale.

Substitution represents a primary operational headwind. Intermittent solar and wind output is directly displacing GD Power's most profitable asset class. As coal generation fell below half of national electricity output in the first half of 2026, midday solar generation increasingly displaces thermal dispatch during peak daylight hours.²

The Three KPIs That Matter

Evaluating GD Power's underlying trajectory relies primarily on three disclosed operational metrics:

1. Into-furnace standard coal unit price and long-term contract ratio. This metric serves as the core thermal margin indicator and tests parent-level supply chain integration. The 10.53% drop in unit coal costs to RMB 825.06 per tonne in 2025 drove a 45.23% increase in net profit adjusted for non-recurring items, while a 0.64% price uptick in the first half of 2026 signaled cost stabilization.³² Investors should monitor whether long-term contract compliance holds near 97% to 99% and whether realized coal costs remain competitive against regional benchmarks.

2. Average realized on-grid tariff and market-traded volume share. Realized tariffs dropped from RMB 400.90 per megawatt-hour in 2025 to RMB 393.31 per megawatt-hour in the first half of 2026, as market-traded volume expanded from 91.73% to 93.72% of total output.³² Crucially, the realized premium over local benchmark tariffs halved to RMB 32.10 per megawatt-hour in 2025, reflecting diminishing pricing power across provincial wholesale markets.³

3. Dadu River hydro generation and fleet utilization hours. This metric tracks the financial return on long-term hydro infrastructure investments. Hydro utilization averaged 3,680 hours in 2025, down 298 hours year-over-year.³ As Shuangjiangkou completes reservoir impoundment and enhances downstream flow regulation, utilization hours and total generation must expand to validate the capital deployment thesis.

Risk Radar

Tariff erosion represents an active financial risk. Realized price declines have offset unit fuel savings over recent reporting periods. As solar capacity expands, midday spot power prices in eastern provinces face downward pressure, reducing thermal margins during historical peak generation hours. Disclosures explicitly acknowledge that widening peak-valley price spreads squeeze off-peak margins, while regulatory tariff mechanisms remain subject to policy adjustment.²

Hydrological volatility remains an exogenous operational risk. The 2022 Sichuan drought demonstrated how severe weather events contract hydro generation.[^15] Although Shuangjiangkou's multi-year storage capability moderates seasonal variations, multi-year drought conditions remain beyond operational control.

Coal pricing and regulatory policy carry two-way exposure. First-half 2026 domestic supply constraints—driven by mine safety inspections and import restrictions—pushed coal prices up during a traditionally soft demand period.² While long-term contract frameworks mitigate spot spikes, any structural revision to state price-cap guidelines would directly impact thermal operating expenses.

Renewable return compression is accelerating. Document 136 phased out baseline feed-in support for new renewable projects, requiring post-June 2025 installations to compete through provincial auctions.¹⁴² First-half 2026 segment performance—characterized by lower utilization, reduced realized tariffs, and modest earnings relative to deployed capital—highlights the impact of marketized pricing, alongside land access constraints and grid curtailment pressure.²

Asset conversion and execution risk. GD Power carried RMB 62.4 billion in construction-in-progress on its balance sheet at the end of June 2026.² Following a construction-in-progress impairment at the Daxingchuan hydro station in 2024, asset conversion requires monitoring as major capital projects transfer into fixed assets during a period of softening power tariffs.³

The Question a Short Seller Would Ask

The central investment debate focuses on a single structural question: what is the long-term value of a thermal coal fleet transitioning from primary baseload generation to a backup balancing resource?

State policy directives explicitly position coal as a grid-balancing asset, and corporate disclosures confirm that thermal utilization hours face ongoing compression as plants absorb deep-peaking and emergency reserve mandates.² Skeptics argue that capacity tariffs merely cushion a secular operational decline, that 83 gigawatts of thermal capacity faces accelerated economic obsolescence, and that RMB 62 billion in construction assets is being transferred to balance sheets into a softening tariff environment.² Conversely, supporters emphasize that national power demand exceeded four billion kilowatts of installed capacity by mid-2026, that grid stability requires thermal backing during peak winter periods, and that state policy guarantees capacity cost recovery of at least 50% for compliant units.²¹¹ Determining corporate valuation depends on how capacity payments and utilization hours balance over the coming decade.

Bull Case

The constructive thesis relies on structural re-rating drivers: 1. State capacity tariffs covering at least 50% of the RMB 330 per kilowatt fixed-cost benchmark convert a growing portion of thermal revenue into predictable cash flow.¹¹ 2. The hydro commissioning pipeline—targeting 3,796 megawatts in 2026, with 3,502 megawatts added in the first half—comes online as headwater storage expands.² 3. Multi-year storage at Shuangjiangkou is projected to generate roughly 6.6 billion kilowatt-hours in incremental downstream output at negligible operational cost.¹ 4. A committed 60% dividend payout floor with a RMB 0.22 per-share minimum through 2027 provides defined shareholder distribution protection.¹⁹

Bear Case

The cautious perspective projects the continuation of current operational trends: 1. Realized tariffs continue to soften as power market trading approaches national targets of 70% total volume.² 2. Thermal utilization hours remain under pressure as renewable additions capture baseline demand growth. 3. Group net profit fell 27% in 2025, and shareholder net profit contracted 18% in the first half of 2026 despite higher total generation, reflecting top-line margin compression.³² 4. Financial leverage of 73.45% and multi-billion renminbi annual capital expenditure commitments constrain balance sheet flexibility.³ 5. Minority partners receive nearly half of total group net profit before distributions reach listed equity holders.³ 6. Executive transitions—including an acting chairman overseeing long-term capital allocation—introduce strategic continuity questions.¹⁶

GD Power maintains structural competitive advantages through its Dadu River hydro concession and captive parent supply chain. However, these cost-side protections must navigate a national power market where output prices are increasingly determined by competitive auction dynamics.

↑ Back to Top

X. Epilogue & Final Verdict

Thirty-four years after a Dalian district-heating company sold shares to local investors, its corporate successor operates more generating capacity than most industrialized nations and consumes nearly two hundred million tonnes of coal annually—almost all of it procured through its controlling parent.

The operational transformation is genuine and remains under way. A generator that suffered severe net losses in 2021 when standard coal hit RMB 890 per tonne now secures 99% of its fuel under long-term contract. An enterprise whose hydropower development was once viewed as a capital-intensive obligation has synchronized the tallest dam in the world at the head of a river cascade no competitor can replicate. A thermal fleet that previously earned revenue solely by generating volume now receives capacity payments for maintaining availability. And a central state enterprise operating without equity incentives and under an acting chairman has expanded its dividend payout ratio into contracting earnings while establishing a per-share distribution floor.

What remains unchanged is the fundamental structural asymmetry created by the 2002 unbundling of China's power grid. GD Power can stabilize its input costs, but it cannot dictate its realized output prices. Chinese regulators have progressively shifted price formation to provincial wholesale markets that are structurally oversupplied, and the company's own corporate disclosures state plainly that core competitiveness is shifting from physical generation scale to market-operating capability.² That acknowledgment highlights that traditional state-owned enterprise moats—sheer scale, baseline reliability, and central state backing—are no longer sufficient to protect profit margins.

The State Council's 2026 implementation guidance reinforced this trajectory, calling for full marketization across all generation types and targeting roughly 70% of national electricity consumption to trade through wholesale power markets by 2030, with all generators and non-protected users participating directly.² Over the next five years, GD Power's realized tariffs will face continuous competitive market pressure.

The most accurate framework for evaluating GD Power is neither as a traditional coal utility nor as a pure-play renewable transition story. Rather, it functions as a portfolio of long-duration physical assets financed through low-cost state enterprise debt, generating a policy-mediated cash flow stream whose revenue composition is being rewritten by the state entity that owns the majority of its shares. The Dadu River will continue to drop four kilometers to the Sichuan basin. Whether those physical assets deliver expanding returns by 2035 depends entirely on the design of Chinese power markets, and GD Power remains a massive, highly leveraged expression of that policy outcome.

In the world's largest electricity market, controlling both contracted fuel supply and a premier hydropower river provides structural operational durability. Yet durability is not identical to investment returns, and the trajectory of realized tariffs over the coming three years will determine what public shareholders ultimately receive.

References

1. 大渡河双江口水电站首台机组并网发电 — 新华网, 2026-06-26 ↩↩↩↩↩↩↩↩

2. 国电电力发展股份有限公司2026年半年度报告 — 上海证券交易所, 2026-08-18 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

3. 国电电力发展股份有限公司2025年年度报告 — 上海证券交易所, 2026-04-15 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

4. 国务院关于印发电力体制改革方案的通知（国发〔2002〕5号） — 中国政府网 ↩↩↩↩↩

5. 国电电力：电力改革见证祖国变迁 — 国家能源集团, 2019-09 ↩↩↩

6. 神华国电宣布重组 国家能投：世界最大的煤化工集团诞生 — 人民网, 2017-09-07 ↩

7. 国家能源集团：安全领航 筑牢高质量发展"压舱石" — 国务院国有资产监督管理委员会 ↩↩↩

8. 关于进一步深化燃煤发电上网电价市场化改革的通知（发改价格〔2021〕1439号） — 国家发展和改革委员会, 2021-10-12 ↩↩↩

9. 国电电力发展股份有限公司出资组建合资公司重大资产重组报告书 — 上海证券交易所, 2017-08-29 ↩

10. 关于建立煤电容量电价机制的通知（发改价格〔2023〕1501号） — 国家发展和改革委员会, 2023-11-10 ↩↩

11. 国家发展改革委、国家能源局完善发电侧容量电价机制 — 国家发展和改革委员会, 2026-01-30 ↩↩↩

12. 国能大渡河流域水电开发有限公司2024年面向专业投资者公开发行公司债券募集说明书 — 上海证券交易所, 2024-11-29 ↩↩

13. 长江电力2025年归母净利润增长6.2%至345亿元 — 腾讯新闻, 2026-04-30 ↩

14. 国家发展改革委、国家能源局有关负责同志就深化新能源上网电价市场化改革答记者问 — 福建省发展和改革委员会, 2025-02-24 ↩↩

15. 国电电力发展股份有限公司关于董事长辞职、选举董事长并聘任董事会秘书的公告（临2024-21） — 上海证券交易所, 2024-06-13 ↩↩

16. 国电电力：唐坚辞去公司董事长等职务 — 每日经济新闻, 2025-12-29 ↩↩

17. 国电电力：董事长唐坚因工作调动辞职 — 中国能源新闻网, 2025-12-30 ↩↩

18. 国电电力总经理变动 — 北极星太阳能光伏网, 2024-10-12 ↩

19. 国电电力：水火共振，红利为锚 — 新浪财经, 2026-07-24 ↩↩

20. 五大发电集团旗舰上市公司2025年业绩大起底 — 新浪财经, 2026-04-16 ↩↩