Zheng Quan Shi Bao·2025-10-28 23:15Core Insights - The AI computing sector has seen significant returns for heavily invested funds this year, leading to a divergence in investment strategies among fund managers [1][2] - Notable fund managers are adjusting their portfolios within the AI computing chain, with some reducing holdings in high-performing optical modules while increasing positions in PCB and AI application sectors [2][3] Fund Manager Adjustments - Fund managers like Jin Zicai from Caitong Fund have significantly increased their holdings in PCB-related companies such as Shenzhen South Circuit and Shengyi Technology, while reducing positions in leading optical module firms like NewEase and Zhongji Xuchuang [2][4] - The China Europe Digital Economy Fund, managed by Feng Ludan, has made more substantial adjustments, reducing exposure to optical modules and PCB while increasing investments in AI application stocks like Alibaba and Tencent [3][4] Long-term Outlook on AI Infrastructure - Despite the adjustments, there remains a consensus among fund managers regarding the long-term potential of AI infrastructure, with expectations of increased demand for computing power in the coming years [4][5] - Jin Zicai emphasized that the growth certainty of overseas AI has improved, predicting faster growth in computing demand for 2026 and 2027 [4] Investment Risks and Valuation Concerns - While acknowledging the long-term value of the AI computing sector, fund managers have also highlighted the risks associated with high valuations following significant price increases [7] - Concerns have been raised about the sustainability of past performance as the AI sector's overall valuation is no longer at a low level, with some stocks reflecting overly optimistic growth expectations [7]