In early trading on the 5th, the optical module CPO sector launched a strong counterattack. Eoptolink Technology Inc.,Ltd. led the gains, surging over 9%, while Zhongji Innolight rose more than 6%. Companies such as UNIC Technology, Zhaolong Interconnect, and Ruijie Networks saw increases exceeding 3%. Among popular ETFs, the ChiNext Artificial Intelligence ETF (159363), which focuses heavily on optical modules, climbed 3% in on-market trading, with transaction volumes surpassing 200 million yuan.
A recent Goldman Sachs report highlighted that China's AI sector is forming an independent investment theme separate from global tech stocks, with its potential economic benefits being severely underestimated. Allocating to Chinese AI assets has become a necessary strategy for hedging risks in traditional industries. Unlike the US-dominated semiconductor and model layers, China possesses significant global comparative advantages in areas such as power, infrastructure, and physical AI. Currently, global funds have an extremely low allocation to Chinese AI assets. Once capital markets begin to correct this imbalance, substantial fund inflows are anticipated.
Regarding computing power, Wanlian Securities suggested continuing to focus on the AIDC and optical communication sectors. Strengthened capital expenditure and growth in token usage are expected to continue driving demand for AIDC construction. Breakthroughs in CPO (Co-Packaged Optics) technology and its increasing adoption within cabinets are anticipated to further boost demand and penetration rates for 800G and 1.6T products. From a medium to long-term perspective, it remains crucial to capture investment opportunities along the core AI computing power industry chain.
As AI development shifts from computing infrastructure construction to application deployment, the ChiNext Artificial Intelligence ETF (159363) and its off-market feeder funds (Class A: 023407, Class C: 023408), which provide one-click exposure to both "Computing Power + AI Applications," stand to benefit more directly from the growth dividends of the AI technology commercialization boom. In terms of portfolio composition, approximately 60% of the ChiNext Artificial Intelligence ETF is allocated to computing power (including leading optical module and IDC companies), while about 40% is allocated to AI applications, making it not only a core "computing power" play but also a genuine representative of "AI applications."
Comments