Guangdong Hec Technologyholding: The Secret Material Backbone of Global Electronics, EV Batteries, and AI Datacenters

I. Introduction & Episode Roadmap

Drive four hours north from the Pearl River Delta, past the last of the electronics manufacturing hubs, and the landscape changes. Factories thin out as the road climbs into the Nanling mountains. Eventually, it leads to Ruyuan Yao Autonomous County (乳源瑶族自治县), a region better known for waterfalls and ethnic-minority tourism than industrial capitalism. Yet this is the registered address of a company whose components sit inside home air-conditioner power supplies, automotive engine control units, and, increasingly, server racks in AI training clusters.

That company is Guangdong Hec Technologyholding Co., Ltd. (广东东阳光科技控股股份有限公司), listed in Shanghai under ticker 600673.SS. Its registered address sits in a Ruyuan township, and its legal representative remains obscure to most global investors.¹ On August 18, 2026, its market capitalization stood at roughly RMB 110 billion (around $15 billion), with its stock trading at CNY 36.70.²

What catches a fundamental investor's attention is the gap between current valuation and recent earnings. That RMB 110 billion market value sits atop a business that generated RMB 275 million in attributable net profit for the full year 2025 — a 26.5% year-on-year decline — despite revenue expanding 22.4% to RMB 14.94 billion.³ Yet the stock has rallied more than 300% over the past year.⁴ Equity markets are pricing in something beyond trailing earnings.

What this company actually makes

Behind the market activity, HEC operates fundamentally as an electrochemistry business. Its core franchise is formed electrode foil (化成箔), the essential anode material used in aluminum electrolytic capacitors. The product is unglamorous: rolls of aluminum foil chemically etched and re-oxidized under high voltage. But it represents a genuine industry bottleneck. HEC controls roughly 30% of global formed electrode foil capacity across bases in Ruyuan, Yidu (宜都) in Hubei, and Ulanqab (乌兰察布) in Inner Mongolia, making it the only Chinese producer exporting over half of its formed-foil output.⁵

Three legacy segments surround that core: high-end aluminum foil (including air-conditioner fin stock, brazing foil, and lithium-battery cathode foil), fluorine-chlorine chemicals (chlor-alkali, methane chlorides, refrigerants, and PVDF), and energy materials.¹ For two decades, this defined the business model: a vertically integrated materials producer operating from mountain manufacturing hubs, supplying Japanese and Korean component makers.

Then came strategic detours. In 2018, HEC acquired a controlling stake in its parent company's Hong Kong-listed pharmaceutical arm during a surge in anti-flu drug sales. In 2021, after COVID-19 lockdowns severely reduced seasonal flu incidence and crushed drug revenue, HEC divested the stake back to affiliated entities.⁶ Since September 2025, HEC has pursued its most aggressive shift to date: a RMB 28 billion acquisition of Chindata's (秦淮数据) mainland China operations — marking the largest data center deal ever attempted in China — alongside RMB 39 billion to RMB 46 billion in announced multi-year compute-leasing contracts with undisclosed counterparties.⁷⁸⁹¹⁰

The four questions this article tests

Is the materials franchise as strong as the narrative says? The company's segment disclosures for the first half of 2026 highlight a sharp internal divergence. High-end aluminum foil generated RMB 3.67 billion in revenue — the company's largest single segment — at a gross margin of just 4.31%. Meanwhile, fluorine-chlorine chemicals delivered RMB 2.25 billion in revenue at a 45.65% gross margin.¹¹ These figures reflect two distinct business profiles operating under one corporate umbrella, making segment performance far more revealing than consolidated totals.

Was the pharma round-trip value creation or value transfer? HEC purchased the drug asset at a 1.8x premium to book value, absorbed the downturn during a demand collapse, and sold it back to affiliated buyers, leaving public minority shareholders exposed to the full cycle.

Is the AI pivot a capability extension or a balance-sheet bet? Management frames the strategy as "materials as the foundation, compute as the driver" ("材料筑基、算力驱动").¹¹ A more skeptical reading suggests a leveraged industrial producer acquiring GPUs on credit to lease to unidentified counterparties.

Who is actually in control, and on what terms? On December 29, 2025, Zhang Yushuai (张寓帅) — born in 1987 and a Zhejiang University graduate who initially ran a biological research institute within the group's pharmaceutical arm — became the sole ultimate controller of both listed entities following a share transfer from his mother.¹² Historically, the controlling shareholder group's equity pledge ratio has remained extremely high.¹

The core narrative connects a business that mastered specialized physical chemistry with a corporate structure that has repeatedly leveraged that cash-generative foundation to back large, related-party capital transactions. Both dynamics are central to evaluating the enterprise. The story begins with a founder who left Zhejiang in search of low-cost power.

---

↑ Back to Top

II. Strategic Origins & Ruyuan's Energy Arbitrage (1990s–2005)

In 1997, Zhang Zhongneng (张中能), an entrepreneur from Dongyang (东阳) in Zhejiang province, made a counterintuitive move. He traveled more than a thousand kilometers south to build a factory in Ruyuan, a mountainous county in northern Shaoguan (韶关), Guangdong.⁵ Chinese entrepreneurs of that era typically built businesses within regional networks—Wenzhou founders in Wenzhou, Dongyang founders in Dongyang. Zhang relocated to a Yao autonomous county with a modest population and virtually no industrial base.

He was not drawn by cheap labor. He was chasing cheap water.

Why electricity is the whole business

Understanding why Ruyuan mattered requires examining how a capacitor anode functions. An aluminum electrolytic capacitor stores an electrical charge on the surface of an aluminum foil. Because surface area determines capacitance within a given volume, manufacturers deliberately corrode the metal. A high-purity aluminum sheet passes through electrochemical etching baths that carve millions of microscopic tunnels into it, expanding effective surface area by a factor of one hundred or more—effectively transforming a flat metallic sheet into a porous structure without altering its outer dimensions.

Next comes forming (化成). The etched foil passes through an electrolyte under high voltage, growing an aluminum-oxide dielectric film just nanometers thick across every tunnel wall. If the film is too thin, the component fails; if too thick, capacitance declines; if the chemical reaction is too aggressive, the foil becomes brittle and fractures on winding machinery.

Both steps essentially convert massive amounts of electrical power into chemical transformations. Consequently, electrode foil is among the most energy-intensive components in electronics manufacturing. In 2023, electricity alone represented 16.47% of HEC's total procurement spend.¹ For an industrial manufacturer, power is not merely an operational line item—it is the structural determinant of unit economics.

Ruyuan solved that cost equation. Situated in a mountainous, high-rainfall catchment with abundant hydroelectricity, Ruyuan allowed HEC to position its foil plants directly near low-cost power while keeping finished goods within a short trucking distance of electronics factories in the Pearl River Delta.⁵ The 1998 decision to construct a formed-foil facility in Ruyuan was a strategic bet that in an industry where power accounts for a sixth of input costs, cheap electricity provides a durable competitive edge.¹³

That operational playbook was later replicated at far larger scale in northern China. HEC's Inner Mongolia subsidiary in Ulanqab now accounts for roughly 70% to 80% of group formed-foil capacity, and in 2024 it secured approval from the regional energy bureau to participate directly in the Mengxi (蒙西) electricity market—a regulatory arrangement that enables direct power purchasing at wholesale rates.¹ The founding principle—locating electrochemistry operations wherever electrons are cheapest—remains HEC's core operating doctrine.

The shell

Building production capacity was only half the equation; securing a public listing presented a separate hurdle. During an era when Chinese initial public offering queues stretched over years, the standard shortcut to the A-share market was acquiring a distressed listed company and staging a reverse merger.

The shell HEC acquired was Chengdu Measuring & Cutting Tool Co., Ltd. (成都量具刃具股份有限公司), a machine-tool maker that had listed on the Shanghai Stock Exchange in September 1993 under ticker 600673.SH.¹ HEC assumed control of the shell in 2003 and began injecting its aluminum-foil assets.¹³ The corporate moniker caught up gradually: the listed entity adopted its current corporate name in May 2014, trading under the short name Dongyangguangke (东阳光科), before simplifying the ticker name to Dongyangguang (东阳光) in April 2019.¹

This backdoor listing via a 1993-vintage Sichuan machine-tool maker reveals an important aspect of HEC's corporate DNA. HEC never underwent a traditional IPO, nor did it face institutional investors to defend an initial prospectus. Its entry into public markets was transactional. That origin set a precedent: nearly every pivotal moment in HEC's subsequent history has been executed as a financial transaction rather than an organic operational milestone.

The starting product set

The company's initial business focused on basic industrial processing: converting high-purity aluminum ingots into plain electronic foil (电子光箔), manufacturing hydrophilic foil for air-conditioner heat exchangers, and producing basic chlor-alkali chemicals. The air-conditioner foil line anchored HEC inside the supply chains of major appliance manufacturers, including Gree, Midea, Haier, Hisense, Panasonic, and LG.¹ It was a high-volume, low-margin business tied directly to Chinese real estate and home appliance cycles.

Yet it established two foundational capabilities that HEC would monetize over the subsequent two decades: industrial-scale aluminum rolling at precise tolerances, and deep commercial ties across the East Asian component supply chain. The next step was taking that foil upstream into higher-value formed foil by mastering chemical etching processes that Japanese manufacturers had refined for decades and guarded closely.

↑ Back to Top

III. Industrial Mastery: Building the Global Monopoly in Formed Electrode Foil (2005–2017)

Start with the qualification test, because that is where the competitive battle is actually won.

When a Japanese capacitor manufacturer evaluates a new foil supplier, it does not issue a standard purchasing tender. It conducts a multi-month physics experiment. Sample rolls are fed into pilot production lines, and finished capacitors are placed on life-test racks under elevated temperatures and voltages for thousands of hours. The critical failure mode—a gradual drift in leakage current as the oxide layer degrades—cannot be detected in a single week. Only after accelerated aging data confirms long-term stability is the foil qualified into a specific part number. Once designed into an automotive engine control unit or industrial power supply, swapping suppliers requires re-qualifying the entire assembly.

That dynamic creates the true barrier to entry: not just proprietary patents, but a multi-quarter qualification timeline.

Why aluminium electrolytics refuse to die

Industry observers periodically predict the demise of the aluminum electrolytic capacitor, noting that ceramic capacitors are smaller, more reliable, and free of liquid electrolytes. In low-voltage, low-capacitance applications, ceramics have indeed taken market share.

However, fundamental physical constraints remain: for storing substantial electrical charge at high voltage within a compact, cost-effective package, etched and formed aluminum foil remains unmatched. As a result, aluminum electrolytic capacitors remain standard components where bulk electrical energy must be smoothed—including industrial motor drives, solar and wind power inverters, electric vehicle onboard chargers, and front-end power supply units for server racks. The expanding electrification of transport and AI infrastructure has increased, rather than diminished, demand for these specific components.

HEC occupies a defined position in this supply chain as an upstream supplier of etched and formed anode foil to global passive-component manufacturers. Its downstream customer base includes Japanese producers Nippon Chemi-Con, Rubycon, Nichicon, and TDK; Korean manufacturers Samwha and Sam Young; Taiwanese suppliers Lelon and Capxon; and domestic peers Aihua and Nantong Jianghai.¹⁴ Through adjacent foil product lines, HEC also supplies international automotive thermal-systems vendors, including Mahle Behr, Valeo, and Hanon Systems.¹

Myth versus reality: three claims worth testing

Three prevailing narratives frequently appear in broker research and marketing coverage of HEC. Each contains an element of truth, but each is overstated in ways that affect how investors evaluate the business.

Myth one: HEC holds a global monopoly in formed electrode foil. A more precise examination reveals scale leadership rather than a monopoly. Across its Ruyuan, Yidu, and Ulanqab facilities, HEC operates approximately 70 million square meters of annual formed-foil capacity and 80 million square meters of etched-foil capacity, controlling roughly 30% of the global market—and remaining the only Chinese producer exporting over half of its formed-foil output.⁵ Within that portfolio, technically demanding medium- and high-voltage formed foil accounted for 56.5 million square meters of annual capacity at the end of 2023, compared to 12 million square meters for low-voltage foil.¹ A 30% market share represents strong scale leadership in a consolidated sector, but it does not constitute a monopoly, nor does it guarantee the unconstrained pricing power often assumed by market commentators.

Myth two: HEC captured roughly half of Japanese tier-one demand. Public disclosures confirm direct supplier relationships with major Japanese, Korean, and Taiwanese capacitor makers,¹⁴ alongside an export ratio higher than domestic peers.⁵ Overseas sales accounted for 16.91% of revenue in 2021, 21.01% in 2022, and 15.41% in 2023, driven primarily by formed foil shipments to Japan and South Korea.¹ While this represents a meaningful international franchise, export markets generate roughly one-sixth of total group revenue, rather than the majority of the business.

Myth three: the Ruyuan advantage is captive hydroelectric generation. Company disclosures clarify that HEC's cost structure relies on strategic siting near regional hydro resources⁵ and direct participation in Inner Mongolia's wholesale power market,¹ rather than ownership of hydroelectric dams. Siting and regulatory access deliver real cost benefits, but these arrangements are structurally contestable if competitors secure comparable location-based power contracts, unlike proprietary generation assets.

What HEC does uniquely possess is end-to-end integration across the production chain. Disclosures highlight HEC as the only producer globally operating a fully integrated sequence from high-purity aluminum ingot to plain electronic foil, etched foil, laminated foil, formed foil, and finished aluminum electrolytic capacitors.⁵ Captive electrode foil supply allows the company to manage both input costs and component quality across its downstream capacitor lines.¹ This operational integration is well supported by corporate filings, though it receives less attention in equity market commentary.

Benchmarking the rivals

HEC's competitive landscape is defined by two primary domestic peers, each pursuing a distinct strategy:

Nantong Jianghai expanded downstream into finished capacitors and supercapacitors, focusing on application engineering, custom design-in relationships, and end-market branding. It acts as both a customer purchasing foil from HEC and a direct competitor in finished component markets.

Xinjiang Joinworld expanded upstream in western China, leveraging low-cost coal-fired power in Xinjiang for high-purity aluminum smelting and electronic foil production. While Joinworld's cost advantage derives from metallurgical processing and regional power rates, HEC focused its investment on downstream electrochemical etching and forming processes, where chemical processing know-how accumulates over time.

Ultimately, HEC's competitive standing reflects the combination of low-cost energy access, accumulated chemical process expertise, and long-term customer qualification cycles that create high switching costs.

Where the profit actually comes from

During this growth period, the electrode-foil business generated the operational cash flow that funded group expansion. However, segment profit margins require careful interpretation. In the first half of 2026, the electronic components and materials segment—which includes electrode foil and capacitors—generated RMB 1.90 billion in revenue, up 6.5% year-on-year, at a gross margin of 17.21%.¹¹ While solid, this margin is significantly lower than the 30% gross margins often cited in promotional reports, which reflect peak-cycle pricing for specialized high-voltage grades rather than sustained segment-wide profitability.

Concurrently, the high-end aluminum foil segment—encompassing lower-margin commodity products such as air-conditioner fin stock and industrial foil—recorded a gross margin of 4.31% in the first half of 2026 despite generating the group's largest revenue share.¹¹ For the full year 2025, high-end aluminum foil represented 41.3% of total revenue but yielded a gross margin of just 4.12%, as elevated raw aluminum costs could not be fully passed on to customers.¹⁵

This operational profile reveals a structural divergence: HEC's top-line revenue is dominated by high-volume, low-margin aluminum processing, while its profits depend on a smaller portfolio of specialized chemical and electronic foil products. Evaluating the company solely on consolidated revenue growth risks misinterpreting its underlying earnings power—a structural reality that historically motivated management to seek alternative avenues for growth.

↑ Back to Top

IV. The Great Pharma Detour: Injecting HEC Pharm & The "Kewei" Craze (2018–2020)

In the winter of 2018, long lines formed outside Chinese hospital pharmacies for a bright yellow box of granules. A severe influenza season drove overwhelming demand for a paediatric anti-flu treatment. That product was Kewei (可威)—oseltamivir phosphate, the generic version of Roche's Tamiflu—and it was on its way to becoming the most lucrative line item across the broader HEC group.

The central analytical question is not whether Kewei was an effective medication, but why a listed materials manufacturer acquired a controlling stake in the pharmaceutical business that owned it.

The conglomerate architecture

Understanding the transaction requires examining the group's corporate structure. At the top sat Shenzhen HEC Industrial Co., Ltd. (深圳市东阳光实业发展有限公司), a private holding company. Beneath it were two distinct public vehicles: the Shanghai-listed industrial materials entity (600673.SS) and a Hong Kong-listed pharmaceutical business, HEC Pharm (宜昌东阳光长江药业股份有限公司, then trading under 01558.HK).

This dual-listed structure created inherent incentives. Because the parent entity controlled both vehicles, it could reallocate assets between two corporate entities that traded on different stock exchanges and commanded markedly different valuation multiples.

The 2018 injection

In 2018, 600673.SS acquired a 50.04% controlling stake in HEC Pharm from its parent company. The transaction was valued at RMB 3.221 billion and paid for by issuing new shares at a premium of roughly 1.8 times book value. The China Securities Regulatory Commission's merger and acquisition committee approved the restructuring unconditionally on June 6, 2018, and by July 24, 2018, the domestic equity transfer was complete, establishing HEC Pharm as a consolidated subsidiary of the listed industrial company.⁶¹⁶

The acquisition was backed by explicit performance undertakings: HEC Pharm committed to delivering non-recurring net profits of at least RMB 577 million in 2018, RMB 653 million in 2019, and RMB 689 million in 2020.⁶

The management logic was straightforward: industrial materials businesses suffer from cyclical earnings, heavy capital expenditure requirements, and low valuation multiples, whereas pharmaceutical companies typically enjoy higher gross margins, strong cash generation, and premium market ratings. Combining the two was intended to smooth earnings volatility, boost consolidated margins, and expand the group's equity valuation multiple.

What the asset actually was

HEC's pharmaceutical operation originated during the mid-2000s avian influenza outbreak, when Roche granted licenses to manufacture generic oseltamivir in China. HEC secured manufacturing rights and positioned itself strategically: rather than competing directly in standard capsule forms, it developed a proprietary granule formulation tailored for children, which went on to capture the dominant market share for paediatric influenza prescriptions in Chinese hospitals.

The financial concentration resulting from this strategy was stark. In 2019, Kewei generated RMB 5.933 billion in revenue—representing more than 95% of HEC Pharm's total sales.¹⁷

The target asset was not a broadly diversified pharmaceutical developer. It was essentially a single-product enterprise with a pharmaceutical license, whose financial performance depended almost entirely on an uncontrollable external variable: the annual severity of seasonal influenza.

The analytical verdict on the purchase

Evaluating capital allocation requires judging decisions based on the information available at execution rather than in hindsight. Even in 2018, acquiring HEC Pharm carried substantial operational risk.

Product concentration was evident in public filings, with a single molecule generating over nine-tenths of total revenue. Seasonality introduced further earnings volatility, as influenza drug sales were concentrated heavily in winter months, leaving annual results vulnerable to mild flu seasons. Policy risks were also mounting, as China's volume-based procurement program began systematically lowering generic drug prices across major therapeutic categories.

Purchasing a single-molecule, season-dependent, regulatory-exposed asset from a controlling shareholder at a 1.8-times book value premium—using newly issued shares that diluted public minority investors—required an optimistic assumption that peak flu demand would persist indefinitely.

Initially, that assumption held up. For two years, group net profits expanded, the company successfully traded as a materials-and-pharma conglomerate, and the strategic rationale appeared validated. Then, in early 2020, the outbreak of the COVID-19 pandemic and widespread adoption of face masks radically altered respiratory disease patterns across China.

↑ Back to Top

V. Pandemic Shocks & The Unwinding: Divesting HEC Pharm Back to Parent (2020–2021)

A rare category of business risk occurs when a company's core product strategy is rendered vulnerable by an unrelated external shock.

When COVID-19 arrived in China in early 2020, the public health response was comprehensive: universal masking, physical distancing, restricted mobility, and hospital protocols that discouraged non-urgent outpatient visits. While designed to suppress a coronavirus, those measures also proved to be an almost total barrier against seasonal influenza, which spreads via identical transmission pathways.

Seasonal flu did not simply experience a mild year; transmission effectively halted.

The collapse

The financial impact on HEC was immediate and severe. Kewei sales dropped from RMB 5.933 billion in 2019 to RMB 2.069 billion in 2020—a 67.2% decline—before falling further to RMB 555 million in 2021.¹⁷

In just two years, roughly 91% of the revenue generated by the flagship asset 600673.SS had acquired in 2018 vanished. The earnings targets attached to the original deal, constructed for a normal epidemiological environment, became unattainable. The listed parent was left holding a controlling stake in an asset burdened by declining revenue, ongoing research expenditure, and zero visibility into when influenza patterns might normalize.

Management could not have anticipated a global pandemic that would suppress secondary respiratory viruses for two consecutive years. However, the core vulnerability exposed by the crisis—an extreme dependence on a single seasonal molecule—was well documented prior to the acquisition. The pandemic did not create the asset's underlying fragility; it triggered it.

The unwinding

In November 2021, HEC announced a complete strategic reversal. The listed company agreed to divest its 51.41% controlling stake in HEC Pharm—comprising 226 million domestic shares transferred to Guangzhou Pharmaceutical Group (广药集团) and 226 million H shares transferred to its wholly owned Hong Kong subsidiary—for a total consideration of RMB 3.723 billion.¹⁸ Official filings explicitly noted that the company would exit pharmaceutical manufacturing and sales entirely to reallocate capital into new-energy materials, including storage systems and electric vehicle applications.¹⁸

The skeptic's read

Examined strictly by the timeline, HEC acquired 50.04% of a flu-drug producer in 2018 for RMB 3.221 billion in newly issued equity at the peak of the product's earnings cycle. It then sold 51.41% in 2021 for RMB 3.723 billion in cash and assumed debt, after segment sales had plummeted more than 90%.⁶¹⁸

On paper, the transaction generated a modest nominal gain on a slightly larger equity block. Yet that surface calculation masks significant economic costs: equity dilution endured by minority shareholders when shares were issued at 2018 valuations, three years of tied-up balance sheet capacity and executive attention, inventory write-downs carried through the downturn, and the reality that both transaction counterparties belonged to the controlling shareholder's broader corporate network.

The structure raises a fundamental corporate governance question: was the listed A-share entity utilized as a balance-sheet buffer for the group, absorbing an asset at peak valuation and returning it during an operational trough? While each transaction was legally executed, their combined sequence created a persistent governance discount for 600673.SS that requires investors to evaluate all subsequent related-party restructurings with rigorous scrutiny.

The management defence — and its residue

Management presented the divestment as a necessary, decisive reset that eliminated pharmaceutical research cash burn, lowered debt levels, and cleared the way for investment in energy materials. The operational timing aligns with that narrative: the company announced the pharma exit in November 2021 and committed to a major battery-foil expansion project within two months.¹⁹

Yet the pharmaceutical chapter left lasting financial artifacts. HEC retained a residual equity stake in the unit, which subsequently listed on the Hong Kong Stock Exchange in 2025.¹² Because this holding is marked to market through the income statement, reported net profit for a company that officially exited pharmaceuticals remains tied to pharmaceutical share price fluctuations. Management has directly attributed ongoing earnings volatility to fair-value markdowns on its HEC Pharm and Lidun equity stakes alongside incentive scheme accruals.¹⁵

The legacy of the pharmaceutical acquisition is not a balance-sheet impairment, but a structural layer of non-operating noise overlaid on reported earnings—making core operational metrics the primary tool for assessing the underlying business.

↑ Back to Top

VI. The Battery Materials Supercycle: PVDF & Ultra-Thin Lithium Battery Foil (2021–2024)

Two months after announcing its exit from pharmaceuticals in late 2021, HEC redirected its capital. In January 2022, the company committed up to RMB 2.71 billion to construct a 100,000-tonne annual capacity plant for high-end battery aluminum foil in Yidu, Hubei. The project was structured in two phases, targeted for completion in 2023 and 2025.¹⁹²⁰

The corporate narrative framed the move with clean symmetry: divest the flu medication to fund the battery expansion. Yet the core strategic insight lay beyond the capital expenditure figure. HEC was not entering an unfamiliar sector; it was deploying two established industrial capabilities toward a new set of downstream customers.

Competency one: rolling aluminium very, very thin

Battery cathode collector foil requires aluminum foil rolled down to 12 microns—and increasingly 9 microns, roughly one-tenth the thickness of a human hair—across a web wider than a meter. The process demands zero pinholes, strict uniform thickness, and precise surface properties to ensure cathode slurry adheres evenly. Removing each micron reduces cell weight and increases energy density, but exponentially increases manufacturing difficulty.

HEC had spent two decades refining aluminum foil rolling for capacitors and heat exchangers. Its metallurgical knowledge, rolling mill infrastructure, and surface-treatment expertise transferred directly to battery applications. To bridge remaining technical gaps, HEC partnered with Japan's UACJ, a global leader in aluminum foil, forming a joint venture to license advanced Japanese process technology.²⁰ By 2024, the group reached approximately 70,000 tonnes of annual battery foil capacity. Crucially, its product passed rigorous qualification audits with major Japanese component manufacturers, including Panasonic and Murata Manufacturing, moving into volume commercial supply.⁵²¹

Securing qualification from Japanese component manufacturers serves as a far more reliable indicator of technical capability than announced capacity figures. This qualification dynamic mirrors the high switching barriers protecting HEC's core electrode foil business, offering external verification of product quality.

However, corporate customer claims require analytical scrutiny. Industry commentary frequently links HEC's battery foil and PVDF lines to top-tier battery manufacturers such as CATL, BYD, and CALB. While such supply arrangements align with HEC's market scale, specific delivery volumes, contract pricing, and wallet share remain undisclosed in public filings. Investors cannot treat named client lists as proxies for profitability. What public disclosures do verify are capacity, audit status, and segment margins. On these metrics, battery foil represents a viable operational line with modest profitability. Industry reports estimate battery foil gross margins at 10% to 15%—higher than commodity air-conditioner foil but well below fluorine chemicals—while contributing less than 30% of high-end aluminum foil segment revenue.¹⁵

Competency two: fluorine chemistry

The second operational pivot proved more strategic because it capitalized on regulatory supply constraints rather than raw manufacturing scale.

Polyvinylidene fluoride, or PVDF, functions as the critical polymer binder holding active cathode materials to collector foil and coating battery separator films. Without it, lithium-ion battery cells delaminate and fail. During the rapid EV battery expansion of 2021 and 2022, PVDF emerged as one of the tightest bottlenecks in the battery supply chain.

The bottleneck originated upstream. PVDF synthesis requires R142b, a hydrochlorofluorocarbon regulated as an ozone-depleting substance under the Montreal Protocol. Because China strictly caps national R142b production, PVDF supply was restricted not by polymer processing capacity, but by legal access to the controlled precursor molecule. Chemical producers with integrated R142b quotas and captive hydrofluoric acid supplies could expand; unintegrated competitors could not obtain raw materials at any price.

HEC possessed that upstream integration. By 2023, the group commanded approximately 25,000 tonnes of PVDF capacity, securing roughly a 25% share of China's lithium-battery PVDF market.⁵ Ownership, however, was defensively structured: HEC recapitalized its fluororesin subsidiary, bringing in battery-materials maker Putailai to hold a 60% controlling stake while HEC's Ruyuan subsidiary retained 40%, alongside joint investments to expand PVDF and captive R142b production.²⁰ As a result, HEC participates in the PVDF boom partly as a minority partner and raw-material supplier rather than as sole proprietor—a distinction that reframes the asset's overall earnings contribution.

The refrigerant quota — the same trick, bigger

The fluorine franchise's largest earnings driver emerged from an adjacent environmental regulation. In late 2023, China's Ministry of Ecology and Environment issued its 2024 hydrofluorocarbon (HFC) quota allocations, capping national third-generation refrigerant production at 745,600 tonnes and distributing permits based on historical baseline output.¹

The policy transformed a commoditized, overcapacitated industry into a regulated oligopoly. Production output was no longer governed by physical plant capacity, but by non-transferable administrative allowances that structurally barred new market entrants. Juhua emerged as the dominant quota holder with a 37.4% share across mainstream refrigerant varieties, followed by Sanmei at 16%, while HEC secured approximately 48,000 tonnes, placing it firmly in the domestic top tier.¹

HEC actively managed its regulatory position. Rather than treating its initial allocation as static, the company aggressively consolidated permits through cross-variety conversions, inter-company quota transfers, and strategic purchases. These transactions expanded its effective quota position to approximately 53,000 tonnes in 2025 and nearly 60,000 tonnes after subsequent adjustments.¹⁵¹

Market pricing responded sharply to the supply cap. Prices for R32, the primary refrigerant used in residential air conditioning, escalated from approximately RMB 17,000 per tonne in early 2024 to nearly RMB 48,000 by late 2025, reaching around RMB 65,000 per tonne by mid-2026.¹⁵¹¹ Applied across a 60,000-tonne quota portfolio, this price surge became the primary driver of HEC's earnings recovery, underpinning the fluorine-chlorine segment's 45.65% gross margin recorded in the first half of 2026.¹¹

The investor conclusion — and the catch

The operational takeaway highlights management's ability to turn environmental regulatory compliance into pricing leverage across two distinct chemical product lines. This strategy reflects effective deployment of existing industrial assets into high-margin niches.

However, two critical qualifications temper this narrative. First, the refrigerant profit expansion represents a regulatory price windfall rather than an operational breakthrough. HEC remains vulnerable to potential quota policy adjustments or price mean-reversion, which would directly compress segment earnings. Second, the competitive dynamics surrounding PVDF deteriorated after 2022. As domestic producers including Dongyue Group and Juhua brought substantial new capacity online, supply shortages eased, stripping battery-grade PVDF of its premium pricing power.

By late 2024, HEC's materials business displayed a mixed profile: a fluorine segment generating elevated profits from regulatory quotas, a high-volume aluminum segment operating near break-even, a stable electrode foil core, and a battery materials division adjusting to increased market competition. Seeking a higher-growth vector, management initiated its next strategic transformation.

↑ Back to Top

VII. The Modern Era: Governance, Management & AI Infrastructure Pivot (2024–2026)

On November 6, 2020, 张中能 Zhang Zhongneng died at the age of 57.²² The founder who had left Zhejiang for a mountain county with cheap water was gone, and control of an industrial group spanning materials, chemicals and pharmaceuticals passed to his family.

The son

张寓帅 Zhang Yushuai was born in 1987. He graduated from 浙江大学 Zhejiang University in 2011 and, rather than being installed in a corner office, joined the group's pharmaceutical research division — first as director of its biology research institute, later heading generic drug research, then rising to vice president. He became a director of the Yichang pharmaceutical entity in 2015.¹²

That path is unusual for a Chinese second-generation heir and worth weighing. He came up through R&D, not sales or finance, in the group's most science-intensive division. Whether that produces better capital allocation is untested; what it does mean is that his instincts were formed around technology platforms rather than trading.

On his father's death he shared ultimate control with his mother, 郭梅兰 Guo Meilan. Then on December 29, 2025, Guo transferred her indirect holdings to him, making him the sole ultimate controller of both 600673.SS and the group's Hong Kong-listed pharmaceutical vehicle, 06887.HK, which had returned to the Hong Kong main board in August 2025.¹² He has framed the resulting agenda as "二次创业" — a second founding.

Day-to-day, the listed company is run by a professional bench rather than by the controlling family directly. 黄晓东 Huang Xiaodong has served as chairman of the board since May 17, 2024, with 刘克军 Liu Kejun as chief financial officer since 2018 and a group of vice presidents appointed across 2022–2024.²³ The group's public voice has emphasised a "链式运作" chain-operation philosophy — describing HEC not as a diversified conglomerate but as a set of interlocking talent, research, product and market chains.⁵ Investors should treat that framing as a hypothesis to test rather than an explanation, because the same word can describe genuine integration or simply a lot of businesses under one roof.

The ownership facts that matter

Two structural facts define minority shareholders' position.

First, control is concentrated. Shenzhen HEC Industrial and its concert parties held 51.33% of the listed company directly and indirectly as of March 2024 — an outright majority, not the roughly 30% often cited.¹ There is no contestability here; the controlling bloc can carry any ordinary resolution.

Second, and more consequentially, that stake has been heavily encumbered. As of April 13, 2024, the controlling shareholder and concert parties held 1.547 billion shares, of which 80.83% were pledged — all as security for bank loans and other financing at the parent or its subsidiaries. The rating agency reviewing the company flagged this explicitly as a high pledge ratio carrying top-up risk if the share price fell persistently, alongside a heavy debt burden, large impairment losses, and future pressure to absorb new capacity.¹

A pledge ratio above 80% is not a technicality. It means the controlling shareholder's financing is levered to the stock price of the company it controls, which creates an incentive structure that every minority investor should hold in mind when evaluating announcements that move the share price sharply. That is not an allegation about any specific disclosure; it is a description of the structural incentive, and it belongs in the file.

The pivot: from materials to compute

In September 2025, HEC and its controlling shareholder announced a plan to acquire Chindata's China business for RMB 28 billion — the largest M&A transaction in the history of the Chinese data-centre industry.⁷

Chindata's China operation is a genuine asset. It reported 2024 revenue of RMB 6.048 billion and net profit of RMB 1.309 billion, and in the first five months of 2025 revenue of RMB 2.608 billion and net profit of RMB 745 million — a business earning more than four times what HEC itself earned in 2025.⁷³ It operates hyperscale campuses across the Beijing-Tianjin-Hebei ring, the Yangtze River Delta, the Greater Bay Area and the northwest, with total IT capacity in operation and under construction of about 1,640 MW as of the second quarter of 2024.⁷³³

But there is a fact about Chindata that belongs at the centre of any assessment, not in a footnote: it is close to a single-customer business. Revenue from ByteDance accounted for 81.7%, 83.2% and 86.3% of Chindata's revenue across 2020 to 2022, and its largest customer represented roughly 80.23% of sales in 2024 and 83.54% in the first five months of 2025.³³

That concentration cuts in two directions, and honest analysis requires holding both. On the bull side, ByteDance is one of the most aggressive AI infrastructure spenders in China, and being its landlord is a privileged position that HEC could not have built organically. On the bear side, an asset with 80%-plus revenue from one counterparty is not a diversified infrastructure portfolio — it is a leveraged claim on one company's capital-expenditure plans, priced at RMB 28 billion. HEC is therefore proposing to bolt a highly concentrated asset onto a balance sheet that is simultaneously funding a highly concentrated set of anonymous compute contracts. The correlation between those two exposures is not disclosed, and investors should not assume it is low.

The seller side is also worth noting for what it says about price. Bain Capital, which had taken Chindata private, exited into this transaction after roughly two years.³³ Private equity selling an asset at the top of a demand cycle to a listed industrial company is a familiar pattern, and it does not by itself mean the buyer overpaid. It does mean the buyer was not the only sophisticated party at the table, and the more experienced infrastructure owner chose to be the seller.

The structure, however, has drawn serious scrutiny. It ran in three steps. The controlling shareholder first established a vehicle, 东数一号, with nominal capital, keeping the target outside the listed company. That vehicle then acquired Chindata China, bringing in 19 strategic investors — including Yunfeng Capital, Foshan state capital and others — which pushed registered capital to RMB 11.5 billion and left HEC as merely the second-largest shareholder with 30%, with the vehicle formally having no controlling party. The first tranche closed on January 16, 2026, with HEC contributing RMB 3.45 billion for its 30%.²⁴⁸ Finally, from February 2026, the listed company moved to acquire the remaining interest through a share issuance, consolidating full ownership.⁹²⁵

Chinese financial media have questioned whether the sequencing amounts to a 类借壳 quasi-backdoor listing — converting what began as controlling-shareholder assets into third-party jointly held assets, so that the listed company could then appear as a neutral acquirer while achieving the same end state.⁹ The same reporting noted that HEC's shares rose 165.79% between June 3 and September 11, 2025, far outpacing both its sector and the broad market, and raised questions about information flow ahead of the announcement.⁹ The company's board has separately taken the position that the transaction constitutes a major asset restructuring but does not constitute a listing by restructuring.²⁵

For investors, the transaction structure is not a side issue. It determines what minorities are paying, in what currency, and how much dilution accompanies the asset.

The compute contracts

Running in parallel, HEC's subsidiary 东莞东阳光云智算 has signed a sequence of multi-year compute-leasing contracts. On May 5, 2026, it announced a framework contract with "Company A" worth RMB 16–19 billion inclusive of tax over 60 months from acceptance.¹⁰ On June 1, 2026, a second contract with "Company B" for RMB 10–12 billion.²⁶ On July 10, 2026, a third with "Company C" for RMB 13–15 billion.²⁷ Total announced pipeline: RMB 39–46 billion — several times the company's annual revenue.

The commercial model is straightforward: HEC buys and deploys high-performance servers, tests them to the customer's specification, and after acceptance leases the compute and provides full-lifecycle operations, billing monthly.¹⁰

Three things about these announcements deserve an investor's attention.

The counterparties have not been named. All three are identified only by letter. HEC told media that "Company A" is not ByteDance and that the order had taken effect, but it has not disclosed who it is.²⁸ For a contract larger than the company's annual revenue, counterparty identity is not cosmetic — it is the entire credit question. An anonymous counterparty means an investor cannot independently assess ability to pay across a five-year term.

The company itself has flagged that performance is uncertain. After a two-day limit-up run, HEC issued an abnormal share-price movement announcement stating that final fulfilment of the contract and its effect on future performance remained uncertain.²⁹ It has also noted the risk that its own deployment could be delayed if financing is not in place or suppliers cannot deliver.¹⁵

The acceptance terms are asymmetric. Delivery, inspection and acceptance are three separate gates, and if performance specifications are not met or deployment is defective, the customer may unilaterally cancel the order or the contract without liability.³⁰ That is the customer's protection, and it is appropriate — but it means the contracted value is a ceiling, not a floor.

What the first half of 2026 actually showed

The half-year results give the first real evidence. Revenue reached RMB 9.376 billion, up 31.6%, with non-recurring-adjusted net profit of RMB 717 million, up 49.0%. Operating cash flow was RMB 2.512 billion, up 843.7%, driven largely by advance receipts on AI compute services.¹¹

Segment detail matters more than the headline. Fluorine-chlorine chemicals contributed RMB 2.245 billion at 45.65% gross margin. Electronic components delivered RMB 1.902 billion at 17.21%. High-end aluminium foil produced the largest revenue at RMB 3.673 billion but only 4.31% gross margin. And the AI and computing segment generated RMB 289 million — up 1,144% off a near-zero base — at a 45.39% gross margin, of which compute services accounted for RMB 603.4 million of contract revenue recognised against RMB 4.114 billion of server investment.¹¹

Within that, the passive-components thread most connected to HEC's industrial heritage produced RMB 213 million: laminated formed foil and capacitors aimed at low-ESR, high-frequency power supplies for AI servers, with a 20 million square metre first phase running in Ulanqab and a 30 million square metre project under construction in Yidu.¹¹ The liquid-cooling effort has deeper roots than the recent narrative suggests — HEC began working on electronic fluorinated fluids in 2016, claims capability to cool 1,200W chips with 1,400W cold-plate technology in reserve, and has formed a cooling joint venture with optical-module leader 中际旭创 Innolight.³¹ It also invested RMB 90 million in June 2025 into VCSEL chipmaker 纵慧芯光 Vertilite, adding an optical-interconnect position.³¹

The credibility test

Assessing management here requires separating what HEC has proven from what it has promised.

Proven: it built a fluorinated-coolant capability starting in 2016, years before it was fashionable, and it has laminated-foil capacity running and supplying more than fifty customers.¹⁵ Those are real industrial assets with a plausible line back to the company's core chemistry.

Unproven: that a materials company can operate a large-scale GPU leasing business at attractive returns, that anonymous counterparties will take delivery across five years, and that the balance sheet can absorb the capex.

On the last point, the numbers are already tightening. Total liabilities rose from RMB 19.815 billion at end-2025 to RMB 22.040 billion by the first quarter of 2026, pushing the asset-liability ratio to 67.00%, a multi-year high. In that quarter, operating cash inflow of RMB 164 million was set against RMB 662 million of investing outflow, and revenue rose 26.95% to RMB 4.249 billion while attributable net profit fell 57.10% to RMB 119 million.¹⁵³²

There is also a useful calibration point on expectations. In September 2025, a major brokerage initiating on the AI thesis modelled 2025 attributable net profit of RMB 1.402 billion.³¹ The company delivered RMB 275 million.³ The gap between the sell-side model and the outcome is a reminder of how wide the error bars are on this transition — and that the 2025 shortfall arrived before the compute business had scaled at all.

The most telling internal number is this: HEC earned RMB 906 million of attributable net profit in the first nine months of 2025 — a record — and finished the year at RMB 275 million.¹²³ The fourth quarter erased two-thirds of the year, driven by the non-operating items management identified.¹⁵ Investors relying on quarterly momentum at this company should hold that fact close.

A note from the auditors

One piece of second-layer diligence deserves attention, because it speaks directly to the questions raised above. 天健会计师事务所 Pan-China Certified Public Accountants issued an unmodified opinion on HEC's 2025 financial statements, with no going-concern emphasis.³⁴ That is the clean outcome, and it should be stated plainly.

What is more informative is which two items the auditors designated as Key Audit Matters. The first was revenue recognition — flagged explicitly on the basis that revenue is one of the company's key performance indicators, creating an inherent risk that management could recognise revenue inappropriately in order to meet a target or expectation. The audit procedures included background investigation of newly added significant customers, obtaining their business registration records and enquiring whether any related-party relationship existed with HEC.³⁴ The second was accounts receivable impairment: gross receivables stood at RMB 3.183 billion at end-2025 against a bad-debt provision of RMB 96.0 million, a balance the auditors judged material and dependent on significant management judgement.³⁴

Neither matter is unusual for a manufacturer of this size, and neither is an allegation. But note what they imply for the year ahead. A company that recognises compute revenue monthly, against contracts with counterparties it will not publicly name, has just been audited under a Key Audit Matter framework that specifically requires the auditor to verify whether new significant customers are related parties. The 2026 audit will be a more consequential document than these usually are, and the receivables line will be the first place a sceptical reader should look.

---

↑ Back to Top

VIII. Acquired Playbook: Business & Investing Lessons

Every long-lived industrial company teaches something. HEC offers six primary lessons, several of which serve as warnings.

1. In electrochemistry, the power contract is the business model. When electricity represents 16% of procurement spend, siting decisions become strategy decisions.¹ HEC's founder understood in the 1990s that a Guangdong mountain county with abundant hydro resources was worth more than a coastal industrial park, and the company re-applied the same formula in Inner Mongolia three decades later. The generalizable lesson: in any process industry, identify inputs that are both large and geographically variable, and evaluate whether a company has structurally cheaper access to them than its peers. If so, that advantage compounds quietly for decades. If the advantage reflects merely a favorable spot price, it does not.

2. Related-party injections deserve a permanently higher burden of proof. The 2018–2021 pharmaceutical round trip was not fraud, and the cash outcome was not disastrous. But a listed company that acquires an asset from its parent at the peak of that asset's earnings, holds it through the trough, and returns it to the parent's orbit is a company whose minority investors bore the operational volatility while the parent retained the financial option. The practical discipline is to treat any transaction where a controlling shareholder sits on both sides as requiring clear evidence of independent price discovery, and to discount the equity accordingly when that evidence is thin.

3. Regulatory scarcity is the most underrated moat in heavy industry — and the most easily mistaken for skill. The HFC quota regime handed China's incumbent refrigerant producers something no amount of capital expenditure could buy: a legally capped supply curve.¹ HEC's second-order move — actively trading and converting quota allocations to expand its market position — demonstrated genuine commercial sophistication.¹⁵ But the underlying economic rent was created by a regulator, not by operational superiority. When quota economics drive most of a company's profit, an investor is underwriting a policy decision and should size the position accordingly.

4. Capability recycling is real, but the further from the core, the weaker the transfer. HEC's fluorine chemistry legitimately spans refrigerants, PVDF binders, and dielectric coolants; its aluminum rolling legitimately spans capacitor foil, air-conditioner fin stock, and battery collector foil. Those are true adjacencies with shared production assets and technical know-how. Leasing GPU capacity is not. There is a plausible narrative bridge — HEC makes coolants and capacitors that go into data centers, so operating one appears logical — but the economics of an asset-heavy leasing business share almost nothing with materials manufacturing. A useful test when evaluating an ecosystem strategy is to ask which physical asset, engineering team, or customer relationship is genuinely shared between the core business and the expansion. Where the answer is simply that the customer operates in the same broad sector, that represents a market adjacency rather than a capability adjacency, conferring no cost or quality advantage over a well-capitalized competitor.

5. Read the segment table before the headline, especially in Chinese industrials. The single most valuable analytical habit this company rewards is looking past consolidated revenue totals. HEC's largest segment by revenue is its weakest by margin, its highest-margin segment is driven by an administratively set regulatory quota, and its fastest-growing segment is expanding off a near-zero base. Three completely different investment theses coexist inside a single income statement. Conglomerate structures of this kind are common among Chinese listed industrials and render screening models based solely on group-level growth or consolidated margins uninformative. The corollary is that when a company emphasizes consolidated revenue growth and operating cash flow — as HEC did in its mid-2026 interim reporting — investors should examine which segment metrics the headline framing obscures.

6. Disclosure choices are data. Naming a counterparty on a contract worth more than a year's revenue is not a courtesy; it is the essential input that makes credit analysis possible. When a company declines to name counterparties — as HEC has on all three compute contracts — the appropriate response is neither to assume bad faith nor to underwrite the projected revenue at face value. Commercial confidentiality is a legitimate constraint, particularly when dealing with major internet platforms that resist public identification. However, the risk of that opacity falls entirely on outside shareholders, who are asked to capitalize a cash-flow stream they cannot independently verify. Where a company repeatedly chooses opacity on the items that move its valuation most, that pattern belongs in the discount rate.

The summary of the playbook is that HEC has demonstrated strong technical execution in specialty chemistry while delivering a far more mixed record in capital allocation. The next section tests whether the underlying chemistry alone constitutes a durable competitive position.

↑ Back to Top

IX. Analysis: Hamilton Helmer's 7 Powers & Porter's 5 Forces

Hamilton Helmer's 7 Powers

Process Power — the strongest claim, and it is real. Etching aluminum into a micro-porous sponge and growing a nanometer-scale oxide layer across every internal surface without making the metal brittle cannot be bought off the shelf. It requires decades of accumulated expertise in bath chemistry, current profiles, temperature control, and yield management. The evidence is commercial rather than promotional: HEC is the only Chinese electrode-foil producer exporting more than half its output, holding direct supplier status with major Japanese passive-component manufacturers.⁵¹⁴ Japanese passive-component makers do not qualify suppliers based on price alone.

Switching Costs — moderate to high, and structurally durable. The multi-month qualification cycle, combined with the requirement to re-validate finished components once a specific foil is designed into an automotive or industrial power bill of materials, creates durable customer stickiness. This lock-in protects HEC's incumbent volume, but it does not automatically translate into pricing power. Passive-component customers are large, sophisticated, and maintain multiple qualified suppliers to preserve purchasing leverage.

Scale Economies — high within the niche. Controlling roughly 30% of global electrode-foil capacity across three production bases gives HEC clear unit-cost advantages in a business with high fixed capital expenditure.⁵ Yet scale economies in capital-intensive processing work in both directions: they yield high operating leverage during upturns, but impose severe margin pressure during cyclical downturns—the precise vulnerability highlighted by rating agencies regarding future capacity-absorption risk.¹

Cornered Resource — moderate, and largely regulatory. Capped hydrofluorocarbon quotas and precursor R142b allocations represent clear regulatory barriers that prospective entrants cannot acquire regardless of capital.¹ Strategic access to local hydro power in Ruyuan and wholesale power trading in Inner Mongolia's Mengxi market offer secondary operational advantages. However, because HEC is not the primary quota holder—Juhua's 37.4% market share significantly exceeds HEC's allocation—this reflects a shared oligopoly rent rather than an exclusive cornered resource.¹

Counter-Positioning, Branding, Network Economies — essentially absent. Industrial electrode foil carries no consumer brand power, generates no network effects, and relies on no counter-positioned business model that incumbents hesitate to replicate.

Porter's Five Forces

Threat of new entrants — low in core materials, high in compute. High capital intensity, strict environmental permitting for industrial electrochemistry, multi-year customer qualification timelines, and capped refrigerant quotas combine to make greenfield entry impractical across HEC's primary materials segments. By contrast, these entry barriers do not apply to the compute-leasing business, where competitive entry depends primarily on capital access and server procurement.

Bargaining power of buyers — moderate to high, and asymmetric across segments. In electrode foil, buyer concentration among tier-one Japanese, Korean, and Chinese capacitor manufacturers is offset by long qualification cycles. In high-end aluminum foil, buyers exert dominant pricing leverage, as evidenced by a 4.31% gross margin that leaves HEC unable to fully pass through rising raw metal costs.¹¹ In compute leasing, the strict inspection and acceptance terms of disclosed contracts assign primary execution and performance risk to HEC.³⁰

Bargaining power of suppliers — moderate in metals, high in compute hardware. High-purity aluminum prices track commodity exchange benchmarks outside HEC's control, though vertical integration partially buffers raw material volatility. In AI compute, supplier power represents a direct operational risk: company disclosures explicitly acknowledge that supplier delivery delays could impede contract fulfillment.¹⁵ In a supply-constrained market for high-performance AI accelerators, hardware vendors retain substantial pricing and delivery leverage.

Threat of substitutes — low to moderate, and structurally stable. Ceramic capacitors continue to gain market share in low-voltage, low-capacitance applications, but fundamental physical constraints preserve the role of aluminum electrolytics in high-voltage, bulk energy storage. Expanding electrification across transport and higher power densities in AI server racks reinforce ongoing demand for high-voltage capacitor foil.

Competitive rivalry — moderate, but intensifying in expansion segments. In electrode foil, high market consolidation maintains disciplined competition. In PVDF, rapid capacity expansion by domestic peers, including Dongyue Group and Juhua, has eroded historical pricing premiums. In refrigerants, administrative quotas suppress price competition by design. In AI infrastructure, HEC operates as a recent entrant competing directly against state telecom carriers, established data center operators, and hyperscale cloud providers backed by significantly larger balance sheets.

The synthesis

Applying both frameworks to underlying segment economics yields a strategic profile that defies simple bullish or bearish characterization. HEC commands a durable, evidence-backed competitive position across a specialized portion of its portfolio—namely medium- and high-voltage electrode foil, alongside regulatory-protected fluorine chemistry. Conversely, its largest single revenue generator in aluminum processing operates with negligible pricing power, while its newly launched compute-leasing venture lacks established structural moats.

For investors, the core strategic question is not whether HEC possesses a moat, but whether current equity valuations reflect its high-margin materials core or price in unproven competitive strength in AI infrastructure.

↑ Back to Top

X. Bear vs. Bull Case & Key KPIs to Watch

The bull case

The profit pool is inflecting for structural reasons, not cyclical ones. The refrigerant quota regime represents a permanent shift in industry supply. HEC's expanded position of roughly 60,000 tonnes generated a 45.65% gross margin in the first half of 2026, supported by an R32 price that has quadrupled from its 2024 trough.¹¹¹⁵ Unlike a standard commodity upcycle, this margin profile is protected from greenfield competition because new production capacity cannot legally operate.

The AI passive-materials business is a genuine core-competence extension with visible traction. Laminated formed foil for AI server power supplies generated RMB 213 million in revenue during the first half of 2026, with a 20 million square meter line in Ulanqab supplying more than fifty customers and a 30 million square meter expansion under construction.¹¹¹⁵ Rising rack power densities increase demand for the low-equivalent series resistance, high-frequency bulk capacitance that HEC's foil provides. This segment represents the portion of the AI narrative built on four decades of physical electrochemistry expertise rather than balance-sheet leverage.

Cash generation has stepped up sharply. First-half operating cash flow reached RMB 2.512 billion—more than eight times the prior-year level and materially above reported profit—driven largely by advance receipts on compute contracts.¹¹ Upfront customer payments provide a tangible signal of demand intensity and help mitigate near-term funding concerns.

If the Chindata consolidation completes, the earnings base changes shape entirely. Chindata earned RMB 1.309 billion in 2024—comfortably exceeding HEC's attributable profit in either 2024 or 2025—from contracted hyperscale capacity with top-tier internet platforms.⁷³ Consolidating an asset of that scale would transform the group's earnings profile and reduce its exposure to commodity aluminum foil.

The bear case

The valuation is discounting execution that has not happened. A market capitalization of roughly RMB 110 billion against 2025 attributable net profit of RMB 275 million leaves no margin for error.²³ The stock price has rallied more than 300% over the past year while underlying earnings declined.⁴ A single quarter—the fourth quarter of 2025, which reduced a record nine-month profit of RMB 906 million down to RMB 275 million for the full year—demonstrates how rapidly reported earnings can deteriorate.¹²³

The leverage is real and rising into the largest capex programme in company history. Total liabilities climbed to RMB 22.040 billion by the first quarter of 2026, pushing the asset-liability ratio to 67.00% while the company concurrently funds RMB 4.114 billion in server investments, a RMB 3.45 billion initial tranche for Chindata, and substantial remaining acquisition payments.¹⁵¹¹⁸ Financial commentators in Chinese media have highlighted heightened liquidity risk and weakened debt-service capacity stemming from these resource-intensive commitments.¹⁵ A tightening credit environment or market disruption would directly threaten this funding model.

Disclosure quality on the compute contracts is weak. Announced contracts totaling RMB 39 billion to RMB 46 billion in potential revenue involve counterparties identified only as A, B, and C—creating a substantial analytical gap.¹⁰²⁶²⁷ Management has acknowledged that contract fulfillment remains uncertain and that customers retain unilateral cancellation rights if technical acceptance criteria are not met.²⁹³⁰ This opacity makes the compute backlog the single most volatile element of the valuation.

The acquisition target is itself a concentration bet. Combining a data center business that relies on a single internet platform for over 80% of its revenue with a compute-leasing arm dependent on undisclosed counterparties leaves group earnings heavily exposed to decisions made outside HEC's control.³³ Should Chinese hyperscale capital expenditure slow due to chip availability, AI monetization constraints, or regulatory shifts, HEC would face dual margin compression across two capital-intensive business units with fixed cost bases.

An activist would find plenty to attack. The corporate governance record presents multiple vulnerabilities: a controlling shareholder holding a majority stake with an historically elevated equity pledge ratio; a round-trip acquisition of an affiliated pharmaceutical asset within the past decade; a three-step data center acquisition structure criticized as a quasi-backdoor listing; major commercial contracts with undisclosed counterparties; a 165.79% share-price rally preceding the transaction announcement; and a core aluminum segment delivering roughly four cents of gross profit on the dollar.¹⁹¹¹ While no single item is decisive, collectively they indicate that minority shareholders possess limited influence over capital allocation and limited visibility to verify execution—a combination that justifies a structural governance discount.

The related-party and structural overhang has not gone away. Financial media continue to question the three-step Chindata acquisition structure and the share-price run-up preceding its public announcement.⁹ Furthermore, the controlling shareholder group has historically pledged more than 80% of its equity to secure group financing.¹ Both factors increase the required rate of return for equity holders.

The core industrial business is not carrying the company. With its largest aluminum segment operating at a gross margin of roughly 4% due to an inability to fully pass through metal costs, and with market competition eroding PVDF margins, the legacy industrial base cannot serve as a reliable earnings floor beneath the AI expansion.¹¹¹⁵

AI infrastructure carries its own technology risk. Server leasing economics depend on the residual value of hardware that risks rapid obsolescence. If accelerator supply normalizes or compute rental rates decline, five-year leasing contracts fixed on current hardware costs will suffer margin compression.

The three KPIs that matter most

1. Electrode foil and laminated foil — gross margin and capacity utilization (化成箔及积层箔毛利率与产能利用率). This serves as the primary health check on the core industrial franchise. It indicates whether HEC's chemical process advantage continues to sustain pricing power, whether export demand from Japanese and Korean passive-component makers remains firm, and whether new laminated-foil capacity in Ulanqab and Yidu is being profitably absorbed. A decline in this metric while the AI narrative expands would signal that valuation is being driven by market sentiment rather than operational strength.

2. Compute contract revenue recognized versus contracted value, and the cash conversion behind it (算力服务已确认收入与合同金额之比及回款). Because compute revenue is recognized upon monthly customer acceptance and billing, tracking recognized revenue converts headline contract figures into an observable operational run-rate.¹⁵ Analysts should benchmark recognized monthly revenue against the announced RMB 39 billion to RMB 46 billion backlog and verify that cash collections keep pace with accounts receivable. Sustained upfront cash collections would confirm customer commitment, whereas an expanding receivables balance would indicate execution friction.

3. Asset-liability ratio and net debt against the capex and acquisition schedule (资产负债率与净负债). The bull and bear cases both depend on balance sheet capacity. HEC is attempting to finance a major data center acquisition and expand its server fleet while operating at a multi-year leverage high.¹⁵ This financial metric will ultimately determine whether the company can complete its strategic pivot regardless of broader market demand.

One deliberate omission: consolidated revenue growth should not be used as a key performance indicator for HEC. Because a low-margin aluminum processing segment dominates total sales while compute revenues are recognized monthly on large capital investments, top-line revenue growth will appear strong without reflecting underlying profit performance.

↑ Back to Top

XI. Epilogue & Conclusion

There is a symmetry to this story that is almost too neat. The company began because a founder from Zhejiang recognized that the primary input to his product was not aluminum, labor, or proximity to customers, but electricity—and that electricity was cheapest where the water fell. Nearly thirty years later, his son is buying data centers in an industry where the defining operational constraint is, once again, access to power.

Between those two moments sits a company that mastered a complex physical chemistry. Learning to etch and anodize aluminum foil to a standard that Japanese capacitor manufacturers would accept took decades. That effort established a commanding market position—controlling roughly 30% of global electrode-foil capacity as the only Chinese producer exporting over half its output—that competitors have yet to dislodge.⁵ The fluorine chemistry that developed alongside it turned an environmental regulatory framework into pricing power. Those operational achievements are real, explaining why this business commands an RMB 110 billion market valuation rather than remaining an obscure regional supplier.

Yet the corporate record reveals a second, contrasting pattern that investors must evaluate alongside operational success. Twice, HEC has taken cash flows from its industrial core and redeployed them into structurally unrelated ventures—first an affiliated single-molecule pharmaceutical franchise purchased from its parent company at a cyclical peak, and now a data center and GPU-leasing operation funded by a balance sheet at multi-year leverage highs, relying on unnamed counterparties. The first bet ended in an asset round trip that left public minority shareholders absorbing the volatility. The second remains unresolved, and its outcome will not depend on chemical engineering—it will hinge on whether undisclosed counterparties take delivery, whether credit remains accessible, and whether server leasing economics endure over a five-year horizon.

A further strategic asymmetry stands out. The core materials business required three decades to build and would take a competitor years to replicate. By contrast, a compute-leasing business can be entered by any market participant with credit facilities and server vendor access, enabling deployment in a matter of months. Regardless of the ultimate financial result, HEC is exchanging a slow, defensible asset base for exposure to a fast, contestable market—and financing the transition with elevated debt. That strategy is not inherently flawed; market returns across Chinese artificial intelligence infrastructure in 2026 could compensate for the risk. However, HEC has become a fundamentally different enterprise than it was two years ago. Investors evaluating the company for its electrode foil must recognize that materials manufacturing no longer drives the valuation.

Ultimately, HEC illustrates that industrial defensibility and capital allocation discipline are distinct attributes, and a company can possess considerable operational strength alongside inconsistent financial governance. The technical capability underpinning its electrode foil franchise remains unquestioned. What remains uncertain is whether the economic value created by mastering that physical chemistry will ultimately accrue to public shareholders.

References

1. 广东东阳光科技控股股份有限公司主体长期信用评级报告 — 联合资信 / Shanghai Stock Exchange, 2024-07-02 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

2. Guangdong HEC Technology Holding (SHA:600673) Stock Overview — Stock Analysis, 2026-08-18 ↩↩

3. 东阳光(600673.SH)发布2025年度业绩，扣非净利润7.1亿元，同比增长116.86% — 新浪财经, 2026-04-09 ↩↩↩↩↩↩↩

4. 最高190亿元 东阳光锁定算力大单！一年间股价已涨超260% — 央广网, 2026-05-06 ↩↩

5. 从一片铝箔到四大产业 东阳光"链式运作"破解多元化难题 — 证券时报, 2024-12-23 ↩↩↩↩↩↩↩↩↩↩↩↩↩

6. 东阳光并购标的业绩变脸半年亏超5亿 转型医药产业三年"退货"回归老本行 — 新浪财经, 2021-09-03 ↩↩↩↩

7. 东阳光（600673.SH）：点亮AI算力一体化平台之"光" — 国盛证券, 2025-09-15 ↩↩↩↩↩

8. 东阳光关于与关联方共同增资合资公司用于收购秦淮数据中国100%股权暨关联交易交割完成的公告 — 同花顺财经, 2026-01-16 ↩↩↩

9. 东阳光收购秦淮数据"三步走"迷局：是否属于"类借壳"？股价先知先觉 — 新浪财经, 2026-03-10 ↩↩↩↩↩↩

10. 广东东阳光科技控股股份有限公司关于控股子公司签署算力服务采购框架合同的公告 — 上海证券报, 2026-05-06 ↩↩↩↩

11. 东阳光2026半年报：AI业务暴增1144% 战略转型驱动净利增长48.95% — 新浪财经, 2026-08-18 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

12. 张寓帅独掌800亿帝国力推二次创业 东阳光智造升级九个月赚9亿可持续否？ — 新浪财经, 2026-01-05 ↩↩↩↩↩↩

13. 高科技之光照亮中国制造未来——东阳光集团高质量发展启示录 — 新华网, 2025-07-31 ↩↩

14. 东阳光：提早布局高端元器件为战略支点，抢占人工智能新空间 — 中国企业网, 2025 ↩↩↩

15. 一年股价涨317%却持续增收不增利，东阳光高负债押注算力，将"按月确认巨额合同营收" — 新浪财经, 2026-05-11 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

16. 嘉源助力东阳光科（600673.SH）收购东阳光药（01558.HK）重组项目无条件过会 — 嘉源律师事务所, 2018 ↩

17. 集采、竞争"双杀"下，东阳光药磷酸奥司他韦能否重回巅峰？ — 新浪财经, 2023-05-04 ↩↩

18. 东阳光：拟37.23亿元出售东阳光药51.41%股份 将不再从事医药相关业务 — 每日经济新闻, 2021-11-11 ↩↩↩

19. 东阳光作价37亿剥离医药板块 拟投27亿加入电池铝箔扩能竞赛 — 新浪财经, 2022-01-13 ↩↩

20. 27.1亿元！东阳光拟投建年产10万吨低碳高端电池铝箔项目 — 电池网, 2022-01-12 ↩↩↩

21. 东阳光：电池铝箔已通过松下、村田等行业高端日系客户进行认证审核，并开始量产供货 — NE时代 ↩

22. 东阳光创始人张中能病逝 相关股权由其子张寓帅继承 — CPHI制药在线, 2020-11 ↩

23. 东阳光(600673)公司高管 — 新浪财经 ↩

24. 中国IDC史上最大并购案迎后续 东阳光欲谋求控股权 — 央广财经, 2026-02-25 ↩

25. 东阳光筹划发行股份收购东数一号控制权并募集配套资金 — 中国基金报, 2026-02 ↩↩

26. 东阳光：控股子公司签署算力服务采购合同 预计总金额100亿元至120亿元 — 证券时报, 2026-06-01 ↩↩

27. 130亿元至150亿元！东阳光，公告算力大单 — 证券时报, 2026-07-10 ↩↩

28. 独家 | 东阳光回应最高190亿算力合同关键猜疑："A公司"非字节跳动，订单已生效 — 新浪财经, 2026-05-06 ↩

29. 2连板东阳光：百亿算力服务合同最终履约情况存在不确定性 — 证券时报, 2026-05-07 ↩↩

30. 东阳光控股子公司签160亿元至190亿元算力服务框架合同 — 中国基金报, 2026-05-05 ↩↩↩

31. 东阳光（600673.SH）：点亮AI算力一体化平台之"光"（技术引擎与投资建议章节） — 国盛证券, 2025-09-15 ↩↩↩

32. 东阳光2026年一季报发布：核心主业提质增效 新兴业务布局稳步落地 — 证券日报网, 2026-04-28 ↩

33. 字节核心算力供应商秦淮数据300亿卖身东阳光，贝恩投资两年净赚70亿 — 新浪财经, 2025-09-11 ↩↩↩↩

34. 广东东阳光科技控股股份有限公司2025年度年报审计报告（天健审〔2026〕11-348号） — 天健会计师事务所, 2026-04-10 ↩↩↩