The wave of AI industry transformation is sweeping the globe at an unprecedented pace. Within this industrial trend, Huashang Leading Edge Hybrid Fund (008555) has delivered outstanding results by adhering to a clear investment strategy: focusing on AI as the core theme, prioritizing the overseas computing power supply chain, and allocating to leading domestic computing power enterprises.
As of March 31, 2026, the fund achieved a net value growth rate of 143.08% over the past year, significantly outperforming its benchmark return of 10.94% during the same period, resulting in an excess return of 132.14%. Data from an authoritative fund rating agency shows that Huashang Leading Edge Hybrid Fund ranked 11th among 1,920 peer funds over the past year, placing it in the top 1% of its category.
Fund Manager Peng Wu, who manages Huashang Leading Edge Hybrid Fund, also co-manages Huashang High-End Equipment Manufacturing Stock Fund. Notably, the A-share class of Huashang High-End Equipment Manufacturing Stock Fund ranked first in its category over the past five years, with its one-year and three-year performances both within the top ten, further demonstrating Peng Wu's investment prowess in the technology sector.
Peng Wu graduated from Fudan University with a major in Microelectronics and Solid-State Electronics. She joined Huashang Fund in 2021, serving as an electronics industry researcher and subsequently as a fund manager, accumulating profound industry research expertise. Peng Wu specializes in technology sector research and investment, leveraging solid industry analysis to accurately identify genuine industrial trends and consistently discover high-quality companies with "real growth" potential amidst the advancing technological wave.
Peng Wu believes the key to technology investing lies in making early judgments and accurately identifying true industrial opportunities from numerous tech trends. In her view, technological innovation and investment opportunities over the next five years will likely continue to revolve around AI.
Reflecting on AI developments in 2025, Peng Wu stated that major overseas tech companies have continuously revised their AI capital expenditure upwards since 2025, marking an era of trillion-dollar market caps and hundred-billion-dollar spending. Continuous model iterations, exemplified by the success of Gemini 3.0, demonstrate that large companies will maintain capital expenditure. She also noted that both straightforward scaling of pre-training and gradual post-training methods have proven effective, with improvements in mathematical logic, Agent capabilities, and multimodal abilities. Furthermore, commercialization is showing promise; although a closed-loop return on equity is not yet established, substantial commercial applications have emerged, including products generating billions in revenue (such as coding tools) and companies reaching tens of billions in revenue. Considering AI's boost to core businesses, along with chatbot subscriptions and coding services, AI revenue reached hundreds of billions of US dollars in 2025.
In 2025, domestic open-source models dominated, though computing power faced constraints. Stimulated by Deepseek in the first quarter and increased graphics card shipments, internet data center tenders from major companies declined significantly in the second and third quarters before resuming growth in the fourth quarter. Regarding applications, overseas monetization paths are relatively clearer, with OpenAI's 2030 revenue target of $200 billion appearing achievable. Domestically, B2C business models remain exploratory, with AI phone assistants potentially becoming a key competitive arena in the Chinese B2C market.
Looking ahead to 2026-2027, Peng Wu believes AI technology will continue iterating, with innovations emerging in pre-training, reinforcement learning, synthetic data, multimodality, and other areas, leading to multidimensional model capability enhancements. However, the industry landscape remains fluid, shifting from US leadership to a competitive race between China and the US, and from OpenAI's dominance to a structure of two superpowers and multiple strong players. Anthropic's Annual Recurring Revenue surpassed $30 billion. Beyond coding, intelligent agents are gradually being deployed, while several Cloud Service Providers have raised their capital expenditure guidance. Visibility for the 2027/2028 industry outlook is improving. During the data center upgrade process, optical communication remains the strongest trend. Additionally, orders and guidance from North American manufacturers like Ciena indicate accelerating growth in Data Center Interconnect demand. Currently, Peng Wu remains optimistic about optical modules, optical chips, storage, high-end PCB materials, gas turbines, power supplies for server racks, and liquid cooling. She suggests domestic computing power in 2026 might resemble overseas computing power from about 1.5 years ago. After further tightening of computing resources, domestic computing power is achieving breakthroughs from equipment and fabs to chips, with enhanced advanced process mass production capabilities. She is positive on GPUs & ASICs, fabs, and equipment.
Furthermore, Peng Wu anticipates a new global semiconductor upcycle beginning in 2026, potentially leading to comprehensive price increases across the产业链, from fabs and packaging/testing to chips. During the downturn from late 2021 to 2024, capital expenditure across the产业链 was relatively restrained. AI is率先 driving the computing power-related产业链 into an upcycle, intensifying pressure on non-AI segments. This crowding-out effect is expected to persist throughout 2026, making "price increases" a potential annual theme, especially as the memory cycle might last longer than market expectations. She is optimistic about fabs, packaging/testing, and memory.
Regarding specific portfolio allocation, Peng Wu stated in the fund's latest periodic report that for overseas computing power, the fund优选 optical modules, fiber optic cables, and optical chips, while maintaining some exposure to new technologies like CPO. Additionally, after nearly three years of rapid development in overseas computing power, AI-related automation equipment is scaling up to improve产业链 efficiency, leading to new allocations in this sector. Overall, the fund's investment strategy remains centered on AI as the main theme, focusing on the overseas computing power chain, allocating to leading domestic computing power enterprises, and semiconductor segments expected to see price increases. The approach involves selecting优质 growth stocks based on their industry cycle position and competitive landscape to achieve medium-to-long-term investment returns.
Comments