HALO Assets: The "Shovel Seller" in the AI Era

Ask AI · How Will the Electric Cooperation Policy Reshape the Investment Value of Power Grid Equipment?

Text by Xiè Chángyàn

Edited by Zhāng Jú

HALO is an abbreviation for Heavy Assets (重资产) + Low Obsolescence (低淘汰率), representing a class of assets with physical barriers, difficult to be replaced by AI, and capable of generating stable cash flows continuously. It is not simply a resurgence of “old assets,” but a re-pricing of asset value in the AI era. Against the backdrop of rapid AI technological iteration, capital shifts from “storytelling” light assets to heavy assets, low obsolescence, and stable cash flow physical assets to gain certainty and risk resistance.

Currently, AI is impacting light asset industries such as software and the internet, while HALO assets like energy, power grids, and mines are the “foundation” of AI operation, not only immune to replacement but benefiting from AI expansion. Especially enterprises with physical assets like factories, equipment, and land, even if operations face difficulties, their assets can still be disposed of, providing a “safety cushion” for stock prices; whereas light asset companies, once their valuation collapses, may become worthless. Moreover, industries such as power, non-ferrous metals, and chemicals, with slow technological iteration and stable supply-demand patterns, have strong profit predictability, becoming “safe harbors” for capital amid uncertainty.

Overall, HALO assets feature high dividends, low volatility, heavy assets, and low obsolescence, highlighting their investment value under AI development and market risk aversion. The core investment opportunities mainly focus on traditional energy sectors, new energy sectors, power grid and equipment fields, and strategic resources like copper, aluminum, and gold within non-ferrous metals.

This image may be AI-generated

HALO Assets:

Sharing AI dividends with defensive attributes

[The characteristic of heavy assets is the first moat of HALO assets. The low obsolescence feature ensures the long-term value and cash flow stability of HALO assets.]

The popularity of HALO assets began a few months ago in the US stock market, and has since spread to China’s capital markets. The core logic of these assets is that as AI technology advances toward the “virtual” world, the physical value and strategic importance of energy and infrastructure in the real world become increasingly prominent.

In this regard, Caitong Securities states that, based on current expectations for US tech growth, although high growth continues and recent tech earnings reports are positive, valuations are not significantly high; but from a medium- to long-term perspective, there may be scenarios where capital expenditure cannot be sustained, or AI replaces software and light assets industries, marking inflection points—timing uncertain, possibly this year, next year, or even later. Meanwhile, valuation multiples are at relatively high levels over the long term.

However, HALO assets are currently undervalued, with long-term cost-effectiveness already apparent; from a horizon of more than two years, the probability of outperforming is high. For prudent investors, if they are unsure about the economic cycle of the tech industry or find AI-driven industry substitution hard to counter, choosing HALO assets offers better odds and risk-reward ratios.

Fundamentally, the first moat of HALO assets is their heavy asset characteristic. These assets typically possess extensive physical infrastructure, equipment, natural resources, or logistics networks, with large upfront capital expenditure, long construction cycles, and often strict regulatory constraints. This high-intensity capital investment and complex entry barriers naturally form high industry entry barriers, preventing rapid influx of new competitors, thus protecting existing firms’ market share and pricing power.

From a financial perspective, heavy assets are reflected in high fixed asset ratios and low asset turnover. Wind data shows that independent power producers and energy traders in the energy sector have an average capital fixation ratio of 276.76%.

Meanwhile, the low obsolescence feature ensures the long-term value and cash flow stability of HALO assets. These assets’ core technologies evolve slowly, and their products or services are rigid societal needs, with physical forms and basic functions remaining stable over decades, not easily disrupted by AI or new technologies. For example, regardless of advances in power generation technology, society’s demand for stable electricity remains eternal; regardless of transportation evolution, the fundamental need for oil and gas resources will persist in the foreseeable future.

Therefore, this rigid demand endows HALO assets with the ability to traverse technological and economic cycles, with cash flows that are highly predictable, exhibiting bond-like attributes.

Especially in the AI era, HALO assets will not be replaced; instead, they determine a country’s competitiveness in AI. Future international competition may largely be a contest of computing power and electricity. In this context, CICC likens HALO assets to AI’s “selling shovels.”

And in the gold rush driven by this, the real profit often goes to the “shovel sellers.” Computing power, energy infrastructure are the “shovels” for AI. AI needs electricity, which in turn depends on coal and gas; data centers require cooling, which needs copper and aluminum, and these resources are non-renewable.

From an investment perspective, investing in HALO assets not only shares in AI dividends but also offers defensive qualities. Broadly, the investment focus for HALO assets mainly lies in traditional energy, new energy, power equipment, and resource commodities like non-ferrous metals.

Traditional Energy Stocks:

Undergoing a Deep Value Reconfiguration

[For coal, oil, and gas companies, their asset value includes not only commodity attributes but also an added “strategic security” premium.]

Today, energy, as the “blood” of modern society, has an undeniable rigid demand. Even as energy structures shift toward green transformation, it is merely a change in energy form, not demand disappearance. The exponential growth in power consumption by AI data centers further reinforces this demand’s rigidity.

Thus, the energy industry is a natural exemplar of the HALO asset framework, verifiable from asset attributes, financial performance, and market logic.

First, from an asset perspective, energy assets represented by oil and gas exploration, coal mines, power plants, and transmission networks are typical heavy-capital investments, with construction cycles spanning years or even decades, featuring natural or oligopolistic barriers, and extremely high entry thresholds. Second, from a financial standpoint, leading companies in coal, oil, and electricity sectors generally enjoy high gross and net profit margins and stable operating cash flows. More importantly, these listed companies often maintain high dividend payout ratios, providing investors with continuous substantial cash returns—for example, China Shenhua’s dividend ratio in 2025 is as high as 79%, with a dividend yield around 5%.

Recent Middle East conflicts and ongoing issues in the Strait of Hormuz have caused oil and natural gas prices to fluctuate widely after rising. Industrial Securities’ March 28 report, “Impact of US-Iran Conflict on Domestic Energy Chain,” states that due to cross-sourcing and downstream application overlaps, research on energy has long been “mutually influencing and affecting.”

From the era background, compared to the 1970s oil crisis, the increased use of natural gas and new energy sources in the global energy structure offers the possibility of energy renewal. Moreover, the global supply chain has undergone multiple reshuffles, such as Brexit and US-China trade frictions, moving toward greater independence and self-reliance in energy security.

From a market logic perspective, under the triple catalysts of geopolitical conflicts elevating energy prices, national energy security strategies driving capital expenditure, and the AI revolution creating incremental demand, the energy sector shows resilience in profit growth and valuation revaluation. Market funds are also embracing energy companies with abundant domestic resources and autonomous supply chains.

In the A-share market, energy companies aligned with China’s “increase reserves and production” strategy, such as oil & gas and coal firms, have asset values that include an “additional premium for strategic security” beyond commodity attributes. Due to their solid physical barriers, rigid demand base, and robust financial performance, they are experiencing a profound value revaluation.

From an asset structure perspective, oil & gas and coal industries are both typical capital-intensive sectors. For example, by the end of 2024, Huajin Shares in the oil sector maintained fixed assets exceeding 10.3 billion yuan, accounting for over 120% of total market value. Comparing with Gansu Energy in the coal sector, which had fixed assets of 24.7 billion yuan at the end of 2024 and 16.4 billion yuan in 2023. This “huge upfront capital expenditure and long return cycle” industry attribute naturally forms very high entry barriers, aligning with the “buildable, hard to replicate” core features of the HALO framework.

From profit and cash flow quality, despite some companies experiencing performance fluctuations due to industry cycles, the overall sector demonstrates strong profitability resilience and cash flow generation. Take China Shenhua, a leading coal enterprise, with projected 2025 gross profit margin of 35.08% and net profit margin of 21.29%, and return on equity (ROE) of 12%, ranking among the top in the sector. More importantly, its operating cash flow averaged around 92 billion yuan over the past five years, providing a solid foundation for sustained high dividend payouts.

Looking at financial robustness, most traditional energy companies have reasonable debt ratios. For example, China Shenhua’s 2025 debt ratio is only 23.31%, indicating a very healthy financial structure. Even for capital-heavy firms like Huajin Shares, the debt ratio in 2024 is controlled around 56.82%, with overall risk manageable. A solid balance sheet enables these companies to better withstand economic cycles, ensuring asset longevity.

Power grid equipment: The “digital foundation” in the AI era

[Currently, all links in the power industry chain present new investment opportunities, especially the power equipment segment.]

If the HALO value of traditional energy lies in its scarcity and supply rigidity as “strategic resources,” then power grid infrastructure is the “digital foundation” of the AI era.

With AI development, data center electricity consumption is rising exponentially. The IEA forecasts that by 2030, global data center electricity use will increase by 133%–230%. This means that stable, reliable, and sufficient power supply has become a core factor influencing the speed and regional deployment of AI industries. Tech giants like Microsoft building their own power plants exemplify that power supply capacity has become a key competitive advantage, with large, stable generation assets and resilient grids effectively controlling the “lifeblood” of AI industry development.

Driven by explosive growth in AI computing power demand and the national “compute-electricity synergy” strategy, power operation and grid equipment assets are undergoing a valuation redefinition from “public utilities” to “digital infrastructure,” with their HALO attributes gaining new growth connotations.

From an investment perspective, in the A-share market, this field includes 103 power operation companies and 137 power grid equipment firms, whose assets and profit structures all demonstrate heavy assets and stable cash flows. Taking the hydropower leader Yangtze Power as an example, with a market value exceeding 660 billion yuan, its operating cash flows are consistently abundant, supporting a dividend yield of 3.49%, a typical “bond-like” HALO asset.

Additionally, nuclear power operator China General Nuclear, and regional thermal power transformer companies like Shaanxi Energy, also show similar high dividend and stable cash flow features. The core of this profit model is that the heavy asset investments in power generation and transmission networks have natural monopoly or oligopoly characteristics, with high barriers to replication, ensuring long-term cash flow predictability.

In the context of HALO asset popularity, the power equipment sector is expected to gain valuation premiums as an “AI safe haven.” Industrial Securities states that, as a typical HALO asset, the entire industry chain exhibits new investment opportunities, especially in the power equipment segment.

The main reasons are threefold: first, heavy asset barriers. Power equipment (including UHV, transformers, grid automation, power supply devices) relies on heavy assets, long-cycle capacity, and complex engineering networks, with high replication costs, long construction periods, and high entry barriers, difficult for AI to replace. Second, low obsolescence. Power grids and infrastructure are the “physical foundation” of AI computing power, with slow technological iteration, and lifespans of 20–30 years, not easily disrupted by AI, and even driven by increased electricity demand. Third, capital shift. Amid AI anxiety and high interest rates, capital is moving from light, high-obsolescence tech stocks to heavy, low-obsolescence, cash-flow-stable infrastructure sectors.

Beyond upstream equipment, in power generation operations, whether it’s baseload power like thermal, hydro, nuclear, or green energy like wind and solar, under tight supply-demand and market-oriented reforms, they also have broad market space.

Policy-wise, this year’s government work report explicitly states the integration of power and compute as a national strategy, indicating the development direction. Notably, in 2026, “compute-electricity synergy” was listed as a new infrastructure project for the first time, marking the integration of power and compute as a national strategy. Supporting this are unprecedented investment plans: State Grid’s fixed asset investment during the 14th Five-Year Plan is expected to reach 4 trillion yuan, with annual investments exceeding 800 billion yuan, focusing on UHV, smart distribution networks, and flexible DC to enhance grid capacity for distributed compute loads and intelligent dispatch.

Thus, clear policy guidance and massive capital expenditure are expected to bring a long-term industry boom for power grid equipment companies, and also provide a clear path for power operators to upgrade networks and enhance service value.

In the “compute-electricity synergy” context, the business models of power grid companies are undergoing profound changes, further expanding their growth space. Four core investment themes have emerged: first, integrated compute-electricity operators (e.g., Yueneng Holding), directly involved in data center investment and operation, sharing compute value-added benefits; second, regional green power operators (e.g., Jinkai New Energy), providing green electricity to data centers and earning environmental rights (green certificates, carbon sinks); third, grid upgrade and smart dispatch service providers (e.g., Guodian Nari), offering technical solutions for high renewable integration and massive compute load matching; fourth, green base load suppliers (e.g., Wangneng Environment), providing stable clean energy like waste-to-energy.

Overall, this evolution from “power flow” to “power flow + data flow + value flow” endows traditional HALO assets with technological and service growth attributes, combining stable cash flows with growth potential.

Specifically, segments like power grid equipment, UHV, power sources, cooling, and energy storage are beneficiaries, and the “going abroad” logic for Chinese power equipment is further reinforced. China’s cost advantage in power equipment, along with the “power export/token export” trend, continues to increase export orders for transformers and UHV equipment.

From sub-sector perspectives, Industrial Securities believes that in UHV and grid equipment, companies like State Power Investment, XJ Electric, China Xidian, and Pinggao Electric have the highest barriers and most certain demand; in distribution and transformers, companies like Jinpan Technology and Yigou benefit from rigid demand from AI data centers and grid upgrades; in AI power supplies, liquid cooling, and UPS, companies like Magmet, Yingweike, and Kstar show the greatest resilience and fastest growth; in energy storage and power electronics, Sun Power, Times Electric, and Sifang Co. benefit from new power systems and AI support; in smart meters and distribution automation, Haixing Electric, Samsung Medical, and Dongfang Electronics demonstrate steady growth and good cash flow.

New Energy: Leading Innovation Creates Stronger Barriers

[Massive fixed assets combined with continuous R&D investment form a more solid “innovation-driven” barrier for new energy leaders.]

Beyond traditional energy, wind, photovoltaic, and energy storage industries in HALO assets show a unique duality: they have “heavy assets, low obsolescence, high cash flow” on the asset side; but in components, batteries, and manufacturing, they face potential disruptive technological iteration risks.

Wind data shows that wind and solar operators generally have high fixed asset ratios. For example, Xin Tian Green Energy, Jazeera New Energy, and Jiangsu New Energy all have fixed assets exceeding 50% of total assets. This asset structure indicates extremely high industry entry barriers, requiring huge initial investments and long construction periods.

Meanwhile, these operational assets have lifespans of 20–25 years or more, providing rigid societal demand for power that will not be replaced in the foreseeable future, fitting the “low obsolescence” definition. From a cash flow perspective, although the average return on invested capital (ROIC) in the industry is relatively low (around 2.19%–3.24%), once projects are operational, their operating cash flows are highly stable and predictable. This feature is increasingly attractive amid market-oriented reforms and mechanisms for green value realization.

Moreover, high-quality wind and solar resource areas, especially offshore wind farms, have natural scarcity and exclusivity. Early entrants lock in resources, forming irreplaceable asset moats, with revenue certainty enhanced by market mechanisms.

In 2026, the nationwide unified capacity electricity price policy for energy storage will provide fixed-cost recovery over more than 10 years. As a result, 4–6 hour long-term energy storage projects can achieve internal rates of return (IRR) over 11%, making them economically viable. As the electricity market develops, the pricing mechanism for wind and solar becomes more rational, further stabilizing cash flows.

Since manufacturing faces potential technological disruption, how listed companies in new energy balance long-term cash flow barriers built by heavy assets with the risks of technological evolution is crucial. The current policy environment and market trends are pushing the industry toward this key “HALO re-positioning.”

In heavy asset industries, technological iteration is often “gradual,” because huge sunk costs require new technologies to bring significant economic benefits to drive capacity replacement. Currently, mainstream technologies like TOPCon (tunneling oxide passivated contact solar cells), semi-direct drive, and lithium batteries offer considerable depreciation and profit windows for existing capacity. Leading manufacturers actively invest in R&D to maintain technological advantage, rather than passively waiting for disruption.

Especially for forward-looking, scale-advantaged leaders, technological iteration is not a threat of elimination but a mechanism to oust competitors and strengthen their HALO attributes. For these giants, large fixed assets combined with ongoing R&D form a more robust “innovation-driven” barrier.

In specific investment opportunities, Industrial Securities suggests focusing on: first, wind power—specifically, wind turbine industry chains, with wind turbine profits expected to stabilize and recover, recommending Goldwind; second, offshore wind—due to high barriers and good market structure, with long-term value, recommending Oriental Cable; third, photovoltaics—due to promising space-based PV, with leading companies like Junda, Jinko, and Trina Solar actively deploying; fourth, slurry materials—due to technological progress and potential for intergenerational advantage, recommending Juhua Materials. For energy storage, Sun Power is recommended; for battery storage, CATL.

Non-ferrous resource stocks: “hard currency” of the physical world

[Although Middle East conflicts have paused the sector, related resource stocks are not finished; after short-term adjustments, they are expected to rebound mid-term.]

In the macro context of 2026, resources are no longer just industrial “raw materials.” Resource commodities have mandatory demand; energy transition (green power) and compute revolution (AI) both heavily depend on copper, aluminum, lithium, and high-heat energy.

Currently, amid ongoing geopolitical frictions and intensifying competition in AI and new energy industries, national security boundaries are expanding to include economic security, supply chain security, and key resource guarantees.

Resource commodities will influence industrial competitiveness and security, with their strategic importance rising in major economies’ strategic layouts. Competition and geopolitical games around strategic minerals are becoming more intense, turning resource stocks into leverage in great power geopolitics. This catalyst has triggered structural reshaping of global supply chains and revaluation of strategic resources, leading to increased market attention on strategic resource sectors. Accordingly, Guosen Securities indicates that related sectors may present long-term investment opportunities.

Data shows that since 2025, prices of copper, aluminum, lithium, cobalt, and other resources have risen sharply. According to IMF data, aluminum, copper, cobalt, and rare earth prices increased by 20.6%, 45.4%, 129.9%, and 115.8%, respectively. Correspondingly, resource sector stocks performed strongly: non-ferrous metals up 98.9%, chemicals up 42.5%.

CICC’s February 8 report on resource stocks’ investment value notes that commodities are beneficiaries of diversified global capital flows. Currently, valuations and costs of energy and chemical commodities are near bottom ranges. Despite short-term volatility, driven by AI compute expansion and energy transition rigid demand, and with some supply-demand structural gaps unchanged, resource stocks are not finished; after short-term corrections, they are expected to rebound mid-term.

On one hand, regarding resource stocks, CICC suggests focusing on: in non-ferrous metals, for gold, pay attention to companies with clear earnings releases, accelerated project commissioning, or overseas gold mine acquisitions; for copper, focus on mines with high self-sufficiency, strong reserve expansion, and acquisition potential.

It also favors aluminum, expecting supply contraction from capacity cuts during losses, leading to price recovery. Upgrading and cost improvements in electrolytic aluminum are also expected to reprice profits and valuations.

On the other hand, in chemical stocks, with limited new capacity and rapidly growing energy storage demand, focus on companies with price elasticity and bottomed prices, with clearer supply-demand improvements, especially those benefiting from refining chain recovery.

Gold stocks remain a focus, with significant gains since 2025. Under geopolitical risk and USD depreciation, combined with HALO assets, some listed companies in this sector show high profit growth in 2025.

For example, Zijin Mining’s 2025 net profit attributable to parent reached 51.78 billion yuan, up 61.55%; core net profit was 50.72 billion, up 60.05%, hitting a new high. In 2025, gold production reached 90 tons, copper 1.09 million tons, and lithium carbonate equivalent 25.5k tons, with gold prices significantly rising, boosting profits.

Guotai Haitong notes that Zijin’s Q4 2025 net profit slightly declined quarter-on-quarter, but core profitability improved, with fluctuations mainly due to investment income decline, non-operating expenses, and management costs, not operational weakness.

Looking ahead, Zijin’s 2026 gold target is raised to 105 tons, with a plan of 130–140 tons by 2028. Key projects like Akyem, Reko Diq, and sea-area gold mines will continue to contribute. The company also plans to acquire Allied Gold and a 25.85% stake in Chifeng Gold. Copper targets are 1.2 million tons in 2026 and 1.5–1.6 million tons in 2028, with major contributions from the Long Dron copper mine phase II, Serbia copper mine expansion, and Morro do Ouro restart, with strong growth certainty in the coming years.

Additionally, the company’s lithium business is entering a volume-expansion phase. With the first phase of salt lake project in Q3 and Laoguo Cuo project ramping up, the lithium business is moving from early layout to scale realization, with strong growth potential beyond 2026. Longer-term, projects like the second phase of Q3 Salt Lake, Laoguo Cuo phase II, Xiangyuan hard rock lithium, and Manao northeast lithium will contribute new capacity. Besides lithium, molybdenum and silver also maintain steady growth, enhancing Zijin’s multi-metal synergy and overall resilience.

(This article was published in the April 4 issue of Securities Market Weekly. The stocks mentioned are for illustrative purposes only and do not constitute investment advice.)