AI and the energy super-cycle support China's exports and industrial momentum, but domestic demand rebalancing remains the key constraint

Overview

The report focuses on China's growth outlook amid the AI and energy super-cycle, with the core judgment that China's aggregate economy is relatively stable but structural divergence is deepening. Exports, advanced manufacturing, AI-related capital expenditure, and energy security investment continue to support industrial production and growth targets; at the same time, household consumption, real estate, employment absorption, local government debt, and overcapacity continue to suppress the recovery of endogenous demand.

Core views

First, China benefits from Asia's industrial cycle and AI and energy capital expenditure, with exports and advanced manufacturing remaining growth anchors, and its global export market share expected to rise to 16.5% by 2030. Second, policy in the near term is more likely to emphasize faster budget implementation, supply-side measures, and industrial policy rather than large-scale stimulus similar to that of September 2024. Third, the Hefei model shows that local industrial policy can cultivate champions through ecosystem planning, bold early-stage investment, and industrial fit, but it is not easily replicable, and homogeneous imitation by different regions could instead worsen overcapacity. Fourth, high household savings and weak consumption need to be gradually unlocked through income redistribution, social security, and social welfare reform. Fifth, AI diffusion is extending from models and cloud capital expenditure to electricity, energy storage, embodied AI, robotics, intelligent driving, and domestic AI chips, but it also brings risks of labor substitution and cyclical volatility.

Analysis framework

The report uses a framework combining macro aggregates, industry capital expenditure, manufacturing share, policy meeting expectations, geopolitical trade relations, capital flows, and AI supply chain diffusion to assess China's growth drivers, policy reaction function, and cross-asset implications.

Methodology notes

• Macro cycleTwo-speed economy framework

  Strong external demand and industry, weak domestic demand and real estate

  It judges the economy to be stable in aggregate but structurally divergent through export, industrial production, government financing, real estate, and consumption data.

• Industrial policyHefei model

  Ecosystem-planning industrial policy

  It positions local governments as ecosystem planners rather than universal subsidizers, emphasizing early-stage investment, patient capital, and matching with local industries.

• GeoeconomicsDe-risking rather than decoupling

  China-Europe trade pressure is manageable

  The EU simultaneously runs a trade deficit with China and depends on key inputs from China, so trade frictions may rise but the risk of full decoupling is low.

• Technology diffusionAI 2.0 diffusion framework

  From models to electricity and embodied AI

  The bottleneck in AI commercialization is shifting from computing power to electricity, energy storage, the grid, robotics, autonomous driving, and the domestic technology stack.

Asset mapping & comparison

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

• China exports and advanced manufacturing

  Directly benefits from the AI and energy capital expenditure super-cycle

  Strengths

  High share of global manufacturing value added, strong export resilience, and advanced manufacturing remains solid.

  Weaknesses

  Automation and capital intensification weaken transmission to employment and consumption.

  Comparison

  Compared with domestic-demand sectors, external demand and advanced manufacturing are better anchors for near-term growth.

  Risks

  Trade frictions, overcapacity, price deflation, and fluctuations in overseas demand.

• Renminbi and USD/CNY

  Affected by capital flow management and trade settlement conversion

  Strengths

  Short-term capital flow reallocation may support the renminbi, and the report maintains USD/CNY at 6.75 for end-2026.

  Weaknesses

  Excessive tightening may suppress the willingness to convert trade income into renminbi.

  Comparison

  Renminbi appreciation is not the main rebalancing tool; policy is more focused on stabilizing the balance of payments.

  Risks

  U.S. dollar moves, capital outflows, policy communication, and geopolitical shocks.

• AI infrastructure, power, energy storage, and data centers

  AI commercialization is driving the bottleneck to shift from computing power to electricity

  Strengths

  Demand for data center construction, energy storage, real-time deployment, and grid flexibility is increasing.

  Weaknesses

  The pace of commercialization, cost efficiency, and system integration still need to be validated.

  Comparison

  Competition is shifting from pure model competition to competition in physical infrastructure and energy supply capability.

  Risks

  Capital expenditure cycle volatility, power constraints, technology path changes, and overheated valuations.

• Humanoid robots and intelligent driving

  Key applications of embodied AI diffusion

  Strengths

  2026 humanoid robot sales are estimated at 28,000 units, and L2+ intelligent driving penetration is expected to continue rising.

  Weaknesses

  Degrees of freedom, reliability, cost, and B2B monetization are still developing.

  Comparison

  Compared with pure software AI, embodied AI relies more on manufacturing, supply chains, and systems engineering capabilities.

  Risks

  Demand realization below expectations, regulation, labor substitution controversies, and mass-production bottlenecks.

• Real estate and household consumption

  The main drag on China's domestic demand recovery and the key to rebalancing

  Strengths

  Some first-tier cities are showing signs of stabilization, and there is a positive correlation between higher social security spending and consumption share.

  Weaknesses

  The broader real estate market is still declining, household savings rates remain high, and consumer confidence is insufficient.

  Comparison

  Compared with exports and industrial supply chains, recovery in real estate and consumption is slower and more dependent on institutional reform.

  Risks

  Further home price adjustment, local government debt constraints, employment pressure, and insufficient policy implementation.

Key data

• China global export share forecast16.5% by 2030The report expects China's global export market share to rise to 16.5% by 2030.
• Available fiscal and quasi-fiscal impulse in 2H约Rmb2trnThe report mentions that around RMB 2 trillion of budgeted fiscal and quasi-fiscal support remains unused in the second half of the year.
• USD/CNY forecast6.75 by end-2026，短期或接近6.70The report maintains its end-2026 USD/CNY forecast of 6.75 and believes there is room for it to move toward 6.70 in the short term.
• China humanoid robot sales estimate28k units for 2026The China industrial team raised its estimate for 2026 humanoid robot sales to 28,000 units.
• L2+ intelligent driving penetration32% in 2026；50%+ in 2030The China autos team expects L2+ intelligent driving penetration to rise from 25% in 2025 to 32% in 2026 and exceed 50% in 2030.
• Robotaxi fleet size360-400k units by 2030，约占8%The report expects China's robotaxi fleet to reach 360,000 to 400,000 vehicles by 2030, accounting for about 8% of the total taxi and ride-hailing fleet.
• China AI chip self-sufficiency rate70% in 2030The report expects China's AI chip self-sufficiency rate could reach 70% by 2030.
• Policy meeting expectationAccelerate budget implementation rather than expand the budgetThe preview of the July Politburo meeting shows that policy focus may still lean toward the supply side, AI+, and energy security.

Impact & implications

For investing, the report favors structural allocation rather than a simple aggregate recovery trade. Beneficiary areas include export supply chains, advanced manufacturing, AI infrastructure, power and energy storage, intelligent driving, robotics, and some domestic substitution; pressured areas include the real estate supply chain, low-end overcapacity manufacturing, labor-intensive services, and sectors dependent on a spontaneous recovery in household consumption. The renminbi may be supported in the short term, but its policy role is more about maintaining capital flow and balance-of-payments stability rather than resolving supply-demand imbalances through significant appreciation.

Risks

• Although the real estate market has stabilized in some areas, it remains in a downward trend at the national level.
• Overcapacity and price deflation may weaken the macro transmission from exports and industrial profits.
• A widening China-Europe trade deficit may trigger more trade measures.
• Although China-U.S. relations show characteristics of a limited truce, export controls, arms sales to Taiwan, and the Middle East situation may still create disruptions.
• AI diffusion brings risks of labor substitution, income inequality, and boom-bust cycles.
• If economic activity or policy implementation does not improve in July and August, pressure for additional easing after September will rise.

What to watch

• Economic activity data for July and August and the pace of policy implementation.
• How the Politburo meeting frames budget execution, AI+, energy security, and supply-side policy.
• Whether PPI, core CPI, and industrial profits can continue to improve.
• Whether signs of real estate stabilization in first-tier cities spread nationwide.
• Policy moves between China and Europe around key inputs, trade deficits, and competitiveness protection.
• Further developments between China and the U.S. on tariffs, export controls, agricultural purchases, Boeing cooperation, and the balance around rare earths/chips.
• Whether AI-related capital expenditure, data center energy storage demand, humanoid robot sales, and intelligent driving penetration deliver as expected.