Morgan Stanley: China's Two-Speed Economy Amidst the AI and Energy Super Cycle

Overview

This research report explores the outlook for China's economy against the backdrop of the artificial intelligence (AI) and energy super cycles. The core conclusion is that China's economy is exhibiting significant 'two-speed' characteristics: on one hand, a robust Asian industrial super cycle and China's dominant position in manufacturing support exports and growth momentum; on the other hand, weak domestic demand recovery, continued real estate market adjustments, and structural pressures in the labor market persist. The report argues that while AI and industrial policies bring structural opportunities, achieving comprehensive rebalancing still relies on deep-seated reforms.

Core views

External demand and the industrial super cycle are the core anchors of current growth. Asia is poised to experience its strongest industrial cycle since the mid-2000s, which is not just a story about AI or technology but encompasses broad capital expenditure. With its dominant position in global manufacturing value-added, China will be a primary beneficiary, with its global export market share expected to rise to 16.5% by 2030. However, the positive spillover effect on the overall economy may be weaker than in the past, as the industrial sector becomes increasingly capital-intensive and automated, reducing its boost to employment. Weak domestic demand and real estate adjustment remain the main drag. Slack in the labor market is viewed as a key factor constraining domestic demand, and the diffusion of AI technology may exacerbate structural replacement risks for youth unemployment and blue-collar services in the short term. Regarding the real estate market, although sales rebounded after the Spring Festival, this is more likely seen as the release of pent-up demand. Most indicators show limited improvement, inventory levels remain high, and price expectations have not fundamentally reversed, necessitating caution regarding whether this is merely another 'false start'. Policy maintains resolve, with structural reform being more important than mere stimulus. Fiscal policy is in 'cruise control,' with the broad fiscal deficit ratio expected to remain at 11.7% of GDP this year, making supplementary budgets unlikely in the second half. On the monetary front, the central bank is unlikely to rely on RMB appreciation to resolve economic imbalances. The report emphasizes that relying solely on smarter industrial policies (such as the 'Hefei Model') cannot narrow supply-demand imbalances, as its success depends on a rare combination of timing, bold bets, and local industrial fit; blind replication could instead exacerbate overcapacity. Resolving legacy issues requires deep-seated reforms in cadre assessment, social welfare, and fiscal/tax systems. AI enters phase 2.0: long-term benefit but short-term friction. China's AI development focus has shifted from infrastructure to profitability realization and application implementation, possessing advantages such as rapid deployment, cost efficiency, and system integration. Surveys show that 51% of enterprises have become AI-enabled or adopters. Looking at historical innovation waves, productivity gains from AI are certain, but labor substitution is transitional and accompanied by boom-bust cycles and inequality risks. Model calculations show that AI's impact on China's GDP is neutral in the short term and positively contributory in the long term. The external environment presents a 'limited truce+' and normalized capital controls. US-China relations are in a 'limited truce+' state, with both sides forming a strategic balance of mutual dependence in chips and rare earths, making substantial escalation unlikely in the short term. Tightening outbound investment controls aims to regulate capital flows rather than close channels, providing short-term support for the RMB exchange rate, but with marginal impact on economic growth. China will continue to adhere to a selective, state-led path of capital account opening to serve the strategy of RMB internationalization.

Analysis framework

The report adopts a 'two-speed economy' analytical framework, breaking down China's economy into two dimensions: external demand/supply side and internal demand/demand side, for independent assessment, avoiding masking structural divergence with aggregate data. When analyzing the impact of AI, it combines patterns from five historical innovation waves with current industry survey data in China, distinguishing between short-term employment frictions and long-term total factor productivity (TFP) enhancement transmission paths. For real estate and policy effects, it employs a combination of high-frequency data tracking and cross-country historical comparisons (such as aligning housing price index peaks) to judge cyclical positions. In analyzing US-China relations, it uses a 'mirror dependence' game theory perspective, quantifying mutual interdependence in key resources (rare earths) and technologies (chips) to assess the boundaries of geopolitical risk.

Methodology notes

• Macroeconomic framework

  Two-Speed Economy Analysis Framework

  Splits the economy into two departments with divergent performance (e.g., strong exports/manufacturing vs. weak domestic demand/real estate) for separate analysis, avoiding masking structural contradictions with averages, helping investors identify divergence opportunities and risks.

• Industry/Industrial Analysis FrameworkPenetration S-curve

  Judgment of AI Technology Diffusion and S-Curve Stage

  Through enterprise survey data (51% adoption rate) and changes in the application layer (from infrastructure to profitability), the report judges that China's AI is at an inflection point where the S-curve accelerates from infrastructure construction to commercial application penetration, which is a key basis for assessing the switching logic of valuations related to AI assets.

• Cycle and Prosperity FrameworkCapacity/Equipment Cycle (Juglar)

  Asian Industrial Super Cycle

  The report defines the surge in investment in high-growth sectors in Asia currently as the strongest industrial/equipment cycle since the mid-2000s, which belongs to typical Juglar cycle analysis, used to judge the sustainability of prosperity in capital goods, intermediate goods, and upstream resources.

• Event Game and Behavioral FinanceGame Theory Analysis

  Mirror Dependence and Strategic Balance

  In the analysis of US-China relations, the report points out that the 'mirror dependence' structure, where the US relies on Chinese rare earths and China relies on US chips, constitutes a Nash equilibrium. This mutual check-and-balance reduces the probability of extreme decoupling or conflict in the short term, providing a game-theoretic foundation for risk assessment.

• Company Fundamentals and Financial Framework

  Augmented Fiscal Deficit

  Unlike the official deficit ratio, this indicator includes quasi-fiscal activities such as local government special bonds, LGFV bond financing, and policy bank bonds, reflecting more truthfully the actual fiscal support provided by the Chinese government to the economy and its sustainability.

Key data

• Forecast of China's Global Export Market Share in 203016.5%Further increase from current levels under the base case scenario
• Forecast for Broad Fiscal Deficit Ratio in 202611.7%Flat compared to 2025, with no supplementary budget expected in the second half
• Proportion of Chinese Enterprises Empowered/Adopting AI51%Based on AlphaWise surveys, showing the breadth of AI diffusion
• Cumulative Decline in Second-Hand Home Prices in Tier-1 Cities37%As of April 2026, widening further from 27% in March 2025
• Household Savings Rate in China34.7%2024 data, still at a global high, constraining consumption release
• USD/CNY Exchange Rate Forecast6.75 (4Q26E) / 6.80 (2Q27E)RMB expected to appreciate moderately against the USD

Impact & implications

For the Chinese market, the 'two-speed economy' implies highly differentiated investment logic. Export chains, advanced manufacturing, and AI computing power and application-layer enterprises benefiting from the Asian industrial super cycle and AI application implementation are expected to receive sustained earnings support; sectors highly correlated with domestic demand and real estate still face valuation suppression, waiting for deeper catalysts such as social welfare reform or re-anchoring of inflation expectations. The US-China 'limited truce+' pattern provides a phased window for cross-border investment and supply chain layout, but normalized capital controls require investors to pay closer attention to compliant channels and liquidity management. In the long run, whether reforms can release excess household savings and alleviate employment frictions caused by AI will be the key variable determining whether China's economy can return from 'two-speed' to 'balanced'.

Risks

• AI technology diffusion speed falls short of expectations or commercial monetization is hindered
• Labor market frictions exacerbated by AI suppress short-term consumer confidence
• The real estate market experiences another bottom-dive after a 'false start', dragging down domestic demand recovery
• Local governments blindly replicating the 'Hefei Model' leads to a new round of overcapacity
• US-China relations break through the 'limited truce+' framework, causing geopolitical risks to heat up again
• Unexpected tightening of capital control measures affects the efficiency of cross-border capital flows

What to watch

• Substantial changes in monthly real estate sales data and inventory destocking cycles
• Specific policy implementation progress in social welfare and fiscal/tax system reforms
• Financial performance of AI application-layer enterprises and B-end monetization data
• Follow-up results of US-China trade and investment negotiations and tariff adjustment dynamics
• Signs of excess household savings migrating to consumption or equity assets
• Actual execution rates of capital expenditure in high-growth Asian sectors