AI and the energy super-cycle support China's growth, but economic divergence and insufficient reflation remain the core constraints

Overview

This report is a macro outlook on China amid the AI and energy capex super-cycle. Its core judgment is that China's economy exhibits a "dual-speed" pattern: exports, advanced manufacturing, AI-related investment, and the energy storage chain provide growth anchors, while consumption, real estate, employment transmission, and broad reflation remain weak. As a result, the aggregate picture is relatively stable, while structural divergence deepens.

Core views

First, Asia's industrial cycle may be entering its strongest phase since the mid-2000s, and China stands to benefit from its advantage in global manufacturing value-added share and export resilience. Second, the spillover from strong exports to the broader economy may be weaker than in the past because the industrial sector is more capital-intensive, more automated, and excess capacity weakens capex transmission. Third, policy will not pivot sharply, but will instead lean more toward budget execution, anti-involution, a unified national market, social welfare reform, and optimization of industrial policy. Fourth, AI's impact on GDP growth is broadly neutral in the short term but positive in the long term, with key opportunities expanding from models and computing power to electricity, energy storage, humanoid robots, autonomous driving, and domestic AI chips.

Analysis framework

The report combines macro cycles, export share, manufacturing value-added, real estate and consumption data, PPI reflation conditions, geopolitical trade relations, AI industry chain bottlenecks, and policy response frameworks to assess China's growth supports, constraints, and structural opportunities.

Methodology notes

• Macro cycle analysisDual-speed economy framework

  Strong exports and advanced manufacturing, weak consumption and housing

  The report uses the dual-speed economy to describe China's stable aggregate growth but diverging internal growth drivers, emphasizing that growth is not spreading evenly.

• Industry cycle analysisAI and energy super-cycle

  AI diffusion drives investment in computing power, electricity, energy storage, power grids, humanoid robots, and autonomous driving

  The report believes AI is moving from the model and training stage toward commercialization and the physical AI stage, with bottlenecks shifting further from computing power toward energy systems and real-world deployment.

• Policy analysisFine-tuning rather than pivoting

  Budget execution, anti-involution, unified market, and social welfare reform

  The report believes export resilience reduces the urgency for additional counter-cyclical easing, making policy more likely to use existing fiscal space and structural reforms to ease divergence.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• China macro assets

  Supported by export resilience, policy fine-tuning, and AI's long-term productivity potential

  Strengths

  Stable aggregate growth, fiscal room still available for execution, strong exports and advanced manufacturing

  Weaknesses

  Weak consumption and housing, high household savings rate, and limited breadth in inflation recovery

  Comparison

  Compared with traditional property-driven cycles, this round relies more on manufacturing, exports, and technology capex

  Risks

  Slowing external demand, insufficient policy support, and continued real estate adjustment

• AI industry chain

  One of the report's core structural opportunities

  Strengths

  Accelerating commercialization, improving domestic stack, and spillover of computing demand into electricity and energy storage

  Weaknesses

  Advanced chips and EDA are still developing, and 2B monetization remains to be validated

  Comparison

  Shifting from past model- and training-led growth toward electricity, energy storage, embodied intelligence, and real-time deployment

  Risks

  Labor disruption, overheated investment, and advanced process constraints

• Energy storage systems

  A beneficiary of the expansion of AI data centers and inference workloads

  Strengths

  Power consumption is shifting from training to inference, increasing demand for energy storage and grid flexibility

  Weaknesses

  Dependent on the pace of data center construction and grid support

  Comparison

  Compared with pure computing hardware, energy systems are becoming the new bottleneck and capex direction

  Risks

  Deployment pace below expectations, power infrastructure constraints, and price competition

• Real estate and consumption

  Macro drags and policy focus areas

  Strengths

  Social welfare reform and income redistribution may unlock household savings

  Weaknesses

  Existing-home sales have weakened again, home prices are still adjusting, and tourism traffic is strong but consumption conversion is weak

  Comparison

  Forms K-shaped divergence versus export manufacturing

  Risks

  Insufficient household confidence, balance sheet pressure, and slower-than-expected consumption recovery

Key data

• Forecast of China's global export market share16.5% by 2030The report expects China's global export market share to rise to 16.5% by 2030.
• Fiscal impulse room in the second half约Rmb2trnThe report says Beijing will focus on budget execution rather than budget expansion.
• Estimated China humanoid robot sales in 202628k unitsMorgan Stanley's China industrial team raised its estimate of 2026 humanoid robot sales to 28k units.
• China L2+ intelligent driving penetration rate32% in 2026; 50%+ in 2030The report cites the China autos team's forecast, with 2025 at 25%.
• Forecast of China's robotaxi fleet size360-400k units by 2030Expected to account for 8% of the total taxi and ride-hailing fleet by 2030.
• Forecast of China's AI chip self-sufficiency rate70% in 2030The report believes China's AI chip self-sufficiency rate could reach 70% by 2030.

Impact & implications

In terms of investment implications, the report supports focusing on structural opportunities in China's export manufacturing, energy storage, AI infrastructure, domestic AI stack, humanoid robots, and intelligent driving; however, it also warns that insufficient macro demand, real estate adjustment, weak consumption, excess capacity, and labor market shocks may constrain earnings diffusion and the sustainability of reflation.

Risks

• Consumption fails to recover sustainably, preventing reflation from spreading from upstream sectors to end demand.
• Real estate sales and home prices continue to adjust, dragging on household wealth effects and local government finances.
• Excess capacity and homogeneous local industrial competition intensify, weakening pricing and profit margins.
• China-US and China-EU trade frictions or supply chain bottlenecks worsen, affecting export resilience.
• AI automation causes labor market disruption; if social protection and retraining are insufficient, consumer confidence may be suppressed.

What to watch

• Whether China's export share and the recovery in non-tech exports remain broad-based.
• Whether month-on-month PPI and downstream profit margins can improve.
• Whether housing sales, home prices, and household propensity to consume can stabilize.
• The pace of implementation and multiplier effects of the roughly Rmb2trn fiscal impulse in the second half.
• The actual deployment speed of AI data centers, electricity, energy storage, humanoid robots, L2+ intelligent driving, and robotaxis.
• The execution strength of anti-involution, a unified national market, social welfare reform, and income redistribution policies.