Highlights

• We see that the AI industry is transitioning from a capacity-driven investment cycle towards an intelligence monetisation cycle, where model differentiation, inference efficiency, enterprise adoption and sustainable gross profit generation will increasingly determine value capture. We remain constructive on China's AI value chain, with a preference for Alibaba and Z.ai, while viewing Tencent as a longer-term beneficiary of AI application monetisation.
   

• Maintain MARKET WEIGHT. Top BUYs: Alibaba, Tencent and Z.ai.

Analysis

• Narrowing the capability gap, accelerating monetisation. We have seen a widening of the capability gap between leading US and Chinese models since mid-26, particularly in complex reasoning, coding and long-horizon agentic tasks. Recent Artificial Analysis benchmarks highlight the divergence, with Claude Opus 5.5 scoring 58 vs 46 for Kimi K3.1 Pro and 32 for GLM-5.3. Despite persistent compute constraints, the stronger real-world data flywheels, increased investment in post-training and recursive self-improvement are likely to help Chinese models narrow the gap by mid-27. China's leading model developers are increasingly shifting their focus from benchmark performance towards real-world adoption and commercial scalability. We see coding, cloud service provider (CSP) partnerships and vertical AI agents as key drivers of the next phase of monetisation.
   

• Anti-distillation could accelerate industry consolidation. While anti-distillation measures raise data acquisition costs rather than fundamentally constrain model scaling, Chinese labs are developing harnesses and synthetic data pipelines to sustain progress. Rising capital, data and engineering requirements should increasingly favour Tier 1 labs with proprietary feedback loops, advanced post-training and efficient inference infrastructure, widening the competitive gap with smaller players.
  
  - Geopolitics and interest rates: Potential tailwinds for China internet valuations. Democratic control of Congress could accelerate a resolution to the US-Iran conflict, ease oil prices and inflation, and potentially pave the way for Fed easing and improved liquidity for China ADRs. We see three potential benefits for China’s internet sector: a) improving foreign inflows into Hong Kong and China ADRs, b) lower AI hardware costs supporting CSP margins and FCF, and c) lower energy prices easing inflationary pressure and supporting consumption. We expect the liquidity-driven valuation uplift to be the most immediate catalyst, particularly for Alibaba and Tencent, while fundamental benefits could materialise more gradually.