Following its Global China Summit, JPMorgan Chase released a strategy report explicitly recommending the use of the big consumer sector as a source of funds for increasing allocations to AI and robotics themes, while maintaining a structurally positive stance on industrial automation.
Client interactions and expert panel discussions during the summit reinforced JPMorgan's core assessment: China's AI capital expenditure still has significant room for growth, but current global and regional investors are primarily accessing AI through U.S., South Korean, and Japanese stocks, with China's AI technology sector not yet fully benefiting from fund inflows. Concurrently, second-order spillover effects from AI capital expenditure are accelerating their diffusion into the industrial ecosystem, providing structural support for factory automation and the robotics sector.
For the consumer sector, significant improvement in domestic macroeconomic consumption data and further earnings-per-share beats in the consumer and internet sectors are necessary prerequisites for expanding allocations to these areas. As these conditions are not yet fully met, the big consumer sector is currently playing more of a funding source role rather than being an object for active accumulation.
China's AI Capital Expenditure: Gap Persists, but Inflection Point Nears China's annual AI capital expenditure remains significantly lower than that of the U.S., at approximately 20% of the latter. This gap stems from both supply and demand factors. On the supply side, access to advanced chips (e.g., NVIDIA's GB200/B300) is constrained, and domestic alternatives still lag in computational throughput, energy efficiency, and high-speed memory capabilities, raising the total cost of ownership for data centers. On the demand side, most Chinese large language models (LLMs) trail the most advanced U.S. systems by about 8 to 12 months in capability, and the domestic ecosystem has not yet reached the self-reinforcing commercial inflection point.
Nevertheless, positive signals are accumulating. DeepSeek V4-Pro is currently on the Pareto frontier in the Code Arena, with its API pricing being about one-fifth the comprehensive price of GLM-5.1/Kimi K2.6, demonstrating the continuous improvement in the cost-effectiveness competitiveness of domestic LLMs. Once past the commercial inflection point, the valuations of China's hyperscale cloud service providers will be more driven by AI, and stocks in the AI infrastructure supply chain and pure LLM companies will continue to outperform with liquidity support.
Addressing market concerns about the AI sector being overly crowded and investor worries about valuation pressures from rising U.S. Treasury yields, the report anticipates the first Federal Reserve rate cut will be delayed until the third quarter of 2027.
AI Spillover Effects Diffuse, Industrial Automation Tailwinds Strengthen In company discussions at the summit, spillover effects from global AI capital expenditure are increasingly benefiting areas beyond the "first-order" beneficiaries in 3C and semiconductors, driving higher-order upswings in adjacent industrial ecosystems—particularly thermal management supply chains, higher-precision machine tools, injection molding, power infrastructure, automotive equipment and production line upgrades, and selected chemicals and metals.
This trend is corroborated by A-share capital expenditure data. In the first quarter of 2026, A-share (excluding finance and real estate) capital expenditure grew 4% year-on-year, with inventory up 8%. Excluding utilities and energy further, the capital expenditure growth rate increased to 7% year-on-year.
Within this, the information technology sector's capital expenditure grew 26% year-on-year, and the industrial sector grew 21%, showing early signs of capital expenditure and inventory upcycle resonance. The improvement in higher-order effects, delivery conversion, and price pass-through prospects may translate into upward revisions of sales and earnings forecasts in the coming quarters.
Consumer Sector: Structural Divergence, Selection Over Broad Allocation Expert panel discussions at the summit indicated that China's consumer sector will offer differentiated structural returns rather than a broad cyclical rebound, with clear divergences in recovery paths across different subsectors and product categories. Speakers advised investors to move away from judging the overall market and instead focus on underlying long-term trends, capturing quality growth assets in niche markets.
From a source of incremental growth perspective, lower-tier cities and township areas are seen as the main engines for incremental consumption growth, with cost-effective factory-direct consumer goods effectively capturing untapped market demand. Evolving consumption habits among younger consumers are driving sustained demand for health-conscious products, evident from sugar-free beverages to premium Chinese-style tea drinks.
Significant improvement in domestic macroeconomic consumption data and further earnings-per-share beats in the consumer and internet sectors are necessary prerequisites for expanding allocations to these areas. Until then, the big consumer sector serves more as a funding source for AI and robotics theme allocations rather than an object for active accumulation.
Comments