“The trend in the AI industry has not changed; what has changed is the market's calibration of pricing logic.” On the 21st, Zhang Yiming, managing director of CICC and director of the Global Equities Business Management Committee, said in an exclusive interview with Zhongxin Jingwei during the “Jointly Building a New Ecosystem to Serve the Small and Medium Enterprises Capital Market” sub-forum of the 2026 Financial Street Forum Annual Conference. Zhang Yiming said that in the first half of the year, parts of the AI industry chain rose rapidly and transactions were crowded. When the pace of overseas AI capital expenditure and model commercialization progress fluctuated in stages, the return of valuation to fundamentals was a normal function of the market mechanism. This kind of calibration will help quality companies with real technical barriers and order fulfillment capabilities in the long run. At the same time, he mentioned that if the market also classifies the valuation of leading targets with performance support as a “bubble,” it clearly ignores the solid fundamentals of China's AI industry chain. The core link represented by optical modules and computing power infrastructure is clear. The pace of sales of 1.6T high-speed optical module products is clear, and the revenue and profit growth rate of leading companies continues to verify the industrial boom. The AI industry chain will become fragmented, and leading companies with global competitiveness will still have medium- to long-term allocation value even after their valuations are digested.

“The trend in the AI industry has not changed; what has changed is the market's calibration of pricing logic.” On the 21st, Zhang Yiming, managing director of CICC and director of the Global Equities Business Management Committee, said in an exclusive interview with Zhongxin Jingwei during the 2026 Financial Street Forum Annual Conference “Jointly Building a New Ecosystem to Serve the Small and Medium Enterprises Capital Market” sub-forum. Zhang Yiming said that in the first half of the year, parts of the AI industry chain rose rapidly and transactions were crowded. When the pace of overseas AI capital expenditure and model commercialization progress fluctuated in stages, the return of valuation to fundamentals was a normal function of the market mechanism. This kind of calibration will help quality companies with real technical barriers and order fulfillment capabilities in the long run. At the same time, he mentioned that if the market also classifies the valuation of leading targets with performance support as a “bubble,” it clearly ignores the solid fundamentals of China's AI industry chain. The core link represented by optical modules and computing power infrastructure is clear. The pace of sales of 1.6T high-speed optical module products is clear, and the revenue and profit growth rate of leading companies continues to verify the industrial boom. The AI industry chain will become fragmented, and leading companies with global competitiveness will still have medium- to long-term allocation value even after their valuations are digested.