Report published September 23, 2026
China AI Value Chain: The Expanding Pie and Race for Value – J.P. Morgan Sector Report
J.P. Morgan · September 23, 2026
J.P. Morgan initiates broader coverage of the China AI value chain with a constructive view, forecasting US$544bn of cumulative operating profit over 2026E-2030E driven by a 60x surge in token consumption. While hardware dominates near-term earnings, downstream models and applications are expected to capture 54% of operating profit by 2030E, with Alibaba, Tencent, and Zhipu identified as preferred stock exposures.
Read the remaining 4 pages
Finvaulta summarizes J.P. Morgan's analysis. Attribute opinions, forecasts, time-sensitive values, and chronology to the issuer and report date; do not treat this page as an independent verification or a current market-data source.
Key Takeaways
- 1.China AI value chain operating profit is forecast to expand from US$27bn in 2026E to US$243bn in 2030E, generating US$544bn in cumulative operating profit over 2026E-2030E.
- 2.Profit pools will shift structurally downstream: hardware's share of operating profit will drop from 93% in 2026E to 30% by 2030E, while models and applications will move from losses to 54% of industry profit by 2030E.
- 3.China token consumption is projected to grow approximately 60x from 2026E to 2030E, driven predominantly by enterprise recurring workloads (c.69% of 2030E consumption).
Table of Contents
- Executive summary
- Exponential intelligence consumption growth creates a larger market and growing earnings across the AI value chain
- Downstream contributes one-third of five-year profits and nearly two-thirds of earnings growth, creating room for valuations to rise
- Opportunity × Scarcity × Value Capture favors Alibaba, Tencent and Zhipu alongside nine initiations/ assumptions
- Three preferred exposures and nine initiations/ assumptions apply these judgments
- AI value chain: A growing end-demand pool with earnings shifting toward models + applications
- Intelligence consumption: About 60x token growth over 2026E-30E, roughly two-thirds by enterprise
- >US$1.4tn provides the end-demand anchor, with further upside from AI-driven value creation
- Value is concentrated upstream today, and that matches where revenue currently sits
- How we build the forecast: One dollar of investment, earned three times
- AI value chain layer analysis: Where each layer earns, and what remains scarce
- Investment framework: Opportunity × Scarcity × Value Capture
- Opportunity: Start with the economic pool around the workload
- Scarcity: Identify what remains difficult to replicate or substitute
- Value Capture: Scarcity has to translate into company economics
- Valuation: Attractive economics still need to offer stock upside
- Preferred exposures: Alibaba, Tencent and Zhipu
Report data
Table 2: 2026E-30E model output: Revenue grows across every layer; hardware earns first, downstream (AI models + applications) earns the growth — as of September 22, 2026.
| Metric | Estimate | Context |
|---|---|---|
| China cumulative AI value chain operating profit | 543.8 US$ bn | Cumulative operating profit pool across hardware, cloud, and downstream models and applications. |
| China annual AI value chain operating profit | 243.3 US$ bn | Annual industry operating profit across hardware, cloud, and downstream. |
| Downstream share of annual AI operating profit | 54% | Downstream models and applications share of total annual operating profit. |
| China hyperscaler AI capex | 777.6 US$ bn | Total headline capex deployed by hyperscalers over the five-year period. |
| Downstream operating profit inflection | 6.2 US$ bn | Downstream layer turns profitable in 2028E as gross margins reach 40%. |
Access the full report
And others from the leading tier 1 banks.
Free account: read previews and a daily allowance of full reports. Pro removes the daily limit.