China's HALO, the physical object is the Ark

Report Overview

We have long emphasized the importance of physical assets, which are now gaining attention through various catalysts and entering the global investors’ view. As the world faces technological challenges to industrial order and regional conflicts challenge globalization, physical assets, often forgotten during periods of order and prosperity, will become systemically important. Chinese assets, being closest to tangible production attributes globally, may reveal that the elusive, non-disruptible HALO assets are widely distributed within the Chinese market.

Summary

1. “HALO” Concept Behind: A Calm Understanding of AI

This week, Nvidia’s earnings exceeded expectations, yet its stock price declined sharply, with a divergence between EPS and stock price since November last year continuing, indicating ongoing market concerns about AI disruption. This phenomenon may be similar to the weakening of domestic new energy assets in 2022: at that time, domestic new energy capital expenditure was still rising with high prosperity, but overall ROE in A-shares continued to decline, and fears of deteriorating supply patterns led the capital market to reprice stable energy assets (coal, thermal power) instead of new energy-related assets. Similar situations are emerging in the US: tech giants’ capital spending continues to rise, but profitability of small and medium enterprises deteriorates, and AI has not increased downstream revenues but mainly reduced costs for large firms and market share battles. This prevents the market from valuing the entire industry chain growth as during the tech bubble era, where revenue growth supported valuation. Investors are now focusing on AI-driven infrastructure and resource sectors, while also worrying about negative impacts on high-value, light-asset industries. The “HALO” concept suggests that fields less susceptible to AI replacement are becoming safe havens. Since the beginning of this year, heavy-asset portfolios in US stocks outperform light-asset ones, with energy, materials, industrials, and utilities sectors showing strong Q4 2025 performance. Demand-side assets benefiting from AI are more resilient.

2. HALO: Chinese Assets

Compared to US stocks, A-shares are concentrated in mining, manufacturing, and other industries less vulnerable to AI disruption. From an industry-neutral perspective, most A-share listed companies have higher tangible asset ratios than their US counterparts. Chinese firms are relatively better equipped to resist AI disruption. From the value-added perspective across society, China’s manufacturing value-added and materials-related industries are also higher than in other major developed economies. Investors may find that the non-disruptible HALO assets they seek are widely present in China. The productive capacity of Chinese assets is irreplaceable, and the long-held view that “productivity equals wealth” is gradually becoming reality. The revaluation of Chinese manufacturing assets has begun, with capital inflows and domestic demand recovery underway.

3. Increased Attention from Overseas Governments on Resource Commodities

The US “Treasury Reserve Plan” and Zimbabwe’s suspension of lithium exports indicate rising overseas government focus on strategic resources. Despite rapid US inventory increases for copper, the inventory-to-annual consumption ratio still has room to grow, especially considering AI + manufacturing recovery will increase the denominator. Three notable features: first, rising government reserve demand in the US; second, resource-exporting countries controlling key mineral supplies, with policies like higher taxes and export controls potentially disrupting supply and raising prices; third, resource-rich countries aiming to develop by extending downstream industries, leveraging long-term demographic and resource advantages, combined with easing monetary policy, to attract capital into emerging markets, benefiting domestic capital goods exports.

4. Middle East Tensions and Oil Prices at Year-End

Recent escalations in Middle East geopolitical conflicts, including US and Israeli strikes on Iran, are likely to cause short-term oil price volatility, impacting US inflation. Over the past three years, oil price changes have had diminishing marginal effects on US CPI, due to declining energy weightings, energy transition, and AI’s impact on employment. Our estimates suggest a 1% monthly increase in oil prices marginally raises US CPI by about 0.14%. If oil prices reach around $90/barrel by year-end, US CPI could shift upward, indicating that the easing cycle and manufacturing cycle are still relatively unaffected.

5. Physical Assets as the Ark in AI Disruption

We have long called for market attention to physical assets, which are now gaining momentum through various catalysts. In a world challenged by technological disruptions and regional conflicts, physical assets will be systemically important. Chinese assets, with their close ties to tangible production, are being revalued. Recommendations include: (1) assets less replaceable by AI benefiting from AI development and resource attention—copper, aluminum, tin, crude oil, oil shipping, rare earths, gold; (2) Chinese export chains with global advantages at cycle bottoms—power grid equipment, energy storage, engineering machinery, wafer manufacturing, and domestic manufacturing turnaround sectors—petrochemicals, dyeing, coal chemicals, pesticides, polyurethane, titanium dioxide; (3) capital inflow, easing balance sheet pressures, and inbound personnel trends—aviation, duty-free, hotels, food and beverages; (4) non-bank financials benefiting from market expansion and bottomed long-term returns.

Main Body

1. Overseas Markets: Concerns over AI Disruption and Opportunities Beyond AI

1.1 Ongoing Concerns about AI Disruption

This week, Nvidia announced its FY2026 Q4 results, with EPS exceeding consensus by 5.5%, yet its stock fell over 8% in three trading days. While such divergence between earnings and stock price is not uncommon, the current reaction is the largest in three years, significantly below historical averages. Since November last year, Nvidia’s stock and EPS have diverged: forward EPS continues to rise while stock prices marginally weaken. Historically, Nvidia’s stock has shown a tendency to revert to forward EPS. From a fundamental perspective, the AI sector showed no significant deterioration last quarter: US enterprise AI adoption rates are rising, with large firms approaching 35%, a record high; US tech giants’ capital expenditure remains robust, with data center activities guiding higher earnings. Market consensus expects Amazon, Google, Microsoft, Meta, and Oracle to spend about $670 billion in 2026, with over 60% YoY growth.

In fact, Nvidia’s recent stock decline and the divergence from fundamentals may resemble the weakness in domestic new energy assets in 2022: despite rising capital expenditure, overall fundamentals weakened, with declining ROE in A-shares; concerns about sustainability of capital spending and profits grew, leading to significant asset price adjustments. Similar patterns are emerging in US stocks: despite ongoing capital expenditure growth by tech giants, overall fundamentals worsen, with profits concentrating in a few large firms. The ROE of the “Mag7” (top 7 tech giants) approaches 32%, while the Russell 2000’s ROE hovers near zero; EPS divergence is evident. AI’s potential to reduce costs motivates deployment, but the lack of downstream revenue growth and the risk of disruption threaten the sustainability of profits, contrasting sharply with the internet bubble era, where rapid expansion of internet companies’ revenues justified valuations. Currently, concerns over AI’s impact on capital expenditure and profitability are causing anomalies in US tech stocks.

1.2 Opportunities Beyond AI Are Emerging

Amid fears of AI disruption, investors are shifting focus to assets less susceptible to AI replacement. Generally, industries vulnerable to AI are information-intensive, creating value by reducing information asymmetry, characterized by light assets and human resources as core assets, with low tangible assets per employee. Conversely, industries closely linked to physical production, with heavy assets and high tangible asset ratios, are less vulnerable. Comparing the top 10% and bottom 10% of US Russell 3000 stocks by fixed asset ratio since 2026, the heavy-asset group has outperformed the light-asset group. Industries like utilities, oil & chemicals, non-ferrous metals, and coal, with high per-employee tangible assets, are less AI-replaceable.

Interestingly, these industries performed well in US Q4 2025 earnings: utility and energy sectors exceeded revenue expectations by over 5%, industrials and materials by over 15%. Outside AI, US industrial capex increased in Q4 2025, with utilities maintaining high growth. Sub-sectors like electrical equipment, transportation, and durable goods saw rapid capex growth.

Industries less vulnerable to AI can be divided into those benefiting from AI (e.g., metals like copper, power) and those immune (e.g., steel, petrochemicals, durable manufacturing). The former benefit from AI infrastructure growth, while the latter are less related to AI and more resilient. Under ongoing AI disruption fears and sustained capital expenditure growth by tech giants, AI-benefiting industries are more aggressive; if AI disruption spreads and capex slows, AI-immune sectors may offer better defense.

2. Chinese Assets Are More Resilient Amid AI Disruption

Compared to US stocks, Chinese non-financial assets’ revenue is more concentrated in mining and manufacturing, closely linked to the real economy and less vulnerable to AI. US stocks’ revenue is more in services like healthcare, retail, media, which face higher AI risk. Industry-neutral comparisons show Chinese firms have similar per-employee fixed assets but higher tangible asset ratios than US firms, indicating stronger resilience. From the value-added perspective, China’s manufacturing and materials sectors are also higher than other major economies. As concerns over AI disruption grow, focusing on Chinese assets remains a prudent choice.

3. Overseas Governments Increasing Focus on Resource Commodities

This year, overseas governments’ attention to strategic resources has intensified: in February, Trump announced the “Treasury Reserve Plan” for USGS-listed minerals; Zimbabwe recently suspended lithium exports to extend downstream processing. Three features stand out:

(1) Demand side: US reserve needs are rising, with critical mineral supply chain security becoming a national strategic issue, involving not just access but maintaining competitiveness. The US, Japan, and Europe are expanding critical mineral lists—US’s list grew from 35 to 60 minerals between 2018 and 2025—indicating strategic importance. The overlap among these lists suggests increasing strategic value of certain metals, with geopolitical and industrial implications. The global low inventory regime is challenged as strategic positioning shifts.

(2) Supply side: resource-rich countries control significant supply; when they implement higher taxes or export controls, supply disruptions and price increases are likely. Africa, Latin America, and some ASEAN countries dominate global critical mineral supplies, giving them greater influence and potential for price volatility.

(3) Long-term trends: resource nationalism has evolved from oil and gas to critical minerals, with resource countries seeking development through downstream industries, leveraging demographic and resource advantages. Short-term easing cycles may attract capital into emerging markets, boosting exports of domestic capital goods.

4. Middle East Tensions and Oil Prices at Year-End

Recent escalations in Middle East geopolitical conflicts, including US and Israeli strikes on Iran, are likely to cause short-term oil price volatility, impacting US inflation. Over the past three years, oil prices’ marginal impact on US CPI has diminished, due to declining energy weightings, energy transition, and AI’s impact on employment. Our estimates suggest a 1% monthly oil price increase marginally raises US CPI by about 0.14%. If oil prices reach around $90/barrel by year-end, US CPI could turn upward, indicating that the easing cycle and manufacturing cycle are still relatively unaffected.

5. Strategies in the Face of AI Disruption

Market concerns about AI revolve around whether AI can drive downstream revenue growth and sustain capital expenditure. The path to ultimate winners in tech remains uncertain, but opportunities outside AI are emerging, with investments spreading into real sectors—upstream resources, midstream manufacturing—favoring Chinese assets, which are more resilient under AI disruption.

Specific allocation suggestions include: (1) upstream resources benefiting from AI and resource attention—copper, aluminum, tin, crude oil, oil shipping, rare earths, gold; (2) Chinese export chains with global advantages at cycle bottoms—power grid, energy storage, engineering machinery, wafer manufacturing, and domestic manufacturing turnaround sectors—petrochemicals, dyeing, coal chemicals, pesticides, polyurethane, titanium dioxide; (3) capital inflow, easing balance sheet pressures, inbound personnel—aviation, duty-free, hotels, food & beverages; (4) non-bank financials benefiting from market expansion and bottomed long-term returns.

Source: Guojin Strategy Team

Risk Warning and Disclaimer

Market risks exist; investments should be cautious. This article does not constitute personal investment advice and does not consider individual user’s specific investment goals, financial situation, or needs. Users should evaluate whether the opinions, views, or conclusions herein are suitable for their circumstances. Investment is at their own risk.