Today, JPMorgan released a major research report, systematically expanding its coverage of China's AI value chain for the first time and giving the sector a "constructive" view.

The bank forecasts that between 2026 and 2030, the cumulative operating profit of China's AI value chain will reach US$544 billion, with the model and application layer contributing one-third of industry profits and accounting for nearly two-thirds of profit growth during the period.

JPMorgan holds a highly optimistic view on the long-term growth of China's AI industry. Its core forecast data includes: Token demand is expected to grow approximately 60-fold between 2026 and 2030, with enterprise demand accounting for 69% of total consumption by 2030 and consumer demand accounting for 31%.

Revenue growth: hardware, cloud, and model-plus-application revenue are projected to grow 1.8-fold, 12.5-fold, and 26.9-fold, respectively. Operating profit: China's AI industry operating profit will surge from US$27 billion in 2026 to US$243 billion in 2030.

As early as its Global China Summit in June this year, JPMorgan clearly pointed out that over the next 12-18 months, AI revenue will be mainly driven by B2B/API, while consumer-facing mass-market AI has no short-term monetization path.

In this coverage initiation, JPMorgan provided a clear top-pick list. Alibaba: the core logic lies in its vertically integrated capabilities across chips, cloud, and models. Alibaba recently unveiled its self-developed AI chip Zhenwu V900 at the Yunqi Conference, with computing power increased to three times that of the previous generation, and it clearly set a goal for Alibaba Cloud's global data center scale to exceed 20GW by 2032, further strengthening its infrastructure moat.

Tencent: JPMorgan believes Tencent possesses scarce AI application assets that have not yet been fully reflected in its valuation. Beta testing of the WeChat AI Agent has already begun, and its value creation path has shifted from "an optional project with no clear timetable" to "a phased rollout project with observable milestones," and the risk premium is expected to continue declining.

Z.AI: JPMorgan remains bullish on Z.AI's differentiated intelligence capabilities and pricing power. Since the release of the GLM-5 series, Z.AI has repeatedly raised API prices while still maintaining rapid business growth, validating market recognition of its model value.

Just two days before JPMorgan released its China AI research report (September 22), JPMorgan CEO Jamie Dimon also said that total AI investment in the hyperscaler ecosystem could reach US$1 trillion next year, an increase of more than 40% from about US$700 billion this year. This global trend provides macro support for the continued growth of China's AI industry chain.

For ordinary investors, after the earlier pullback, valuations in the A-share and Hong Kong AI sectors have returned to a reasonable range. JPMorgan's judgment maintaining its year-end target of 100 for the MSCI China Index (implying 37% upside) and 5,200 for the CSI 300 also indicates that institutions' long-term investment logic for China's AI ecosystem has not changed.

However, the market is still in a correction phase, and the clear path to breaking this correction depends on the earnings reports starting next month. As long as the earnings data are good, there is a chance to break through the current valuation ceiling. In addition, macro factors such as interest rates and the risk-free rate are also important sources of pressure, but this pressure factor may be difficult to break in the short term. Therefore, it is still important to focus on the earnings reports starting next month and return to fundamentals, which is also consistent with the current AI investment style that emphasizes performance delivery.