Recent insights from Fuguo Fund's Luo Qing, shared at the 2026 Fuguo Fund Strategy Conference, highlight the investment outlook for the AI sector. Luo Qing stated that, from the current perspective, large language models still have room for improvement. As these models continue to advance, applications will also enter a more mature stage. He emphasized that optical modules are indeed one of the most critical sub-sectors within AI investments, and he personally remains highly optimistic about the certainty of their growth. AI hardware is currently in a phase of profit realization and is expected to maintain relatively clear growth over the next two years. Consequently, AI hardware continues to be one of the primary investment themes.
Luo Qing pointed out that he maintains a very optimistic attitude towards the overall AI sector this year. Firstly, from the current state, large models still possess significant potential for advancement. With Nvidia's B-series chips progressively being deployed for training, it is anticipated that more powerful new large models, trained on these B-series cards, will be put into use within the year. Secondly, noticeable improvements can already be observed in accessible AI applications, whether in AI programming or text-to-video generation. As large models continue to progress, applications will correspondingly reach a more sophisticated stage of development.
Luo Qing further indicated that optical modules are truly one of the core sub-sectors in AI investing, and he remains very confident in the certainty of their growth trajectory. This confidence is underpinned by robust capital expenditures from downstream end-users—specifically overseas cloud service providers (CSPs), strong demand from upstream optical chip suppliers, and the aggressive inventory plans of the optical module manufacturers themselves, all signaling that the entire optical module industry remains in a highly prosperous cycle. However, many leading optical module stocks experienced significant gains last year. Therefore, from an investment standpoint, a certain period is indeed needed in the short term to digest the excessive accumulated gains.
Furthermore, the ultimate driver of stock prices relies on improvements in corporate operations, meaning they must be driven by financial performance. Thus, if one adopts a longer-term view and waits for these companies to release their financial statements to validate their high-growth momentum, it is believed that many high-quality companies could still achieve favorable returns this year. Compared to the past two years, Luo Qing is more bullish on AI applications in the current year. He believes that only when foundational large model capabilities truly develop to a sufficient level—where they can tangibly assist in production and daily life—will large-scale commercial applications emerge. He feels the industry is getting closer to this capability threshold, which is the first point.
The second point is that, objectively speaking, the development of AI applications over the past year or two has fallen short of capital market expectations. The capital market is often impatient, hoping to see AI applications rapidly deployed and fundamentally transform enterprise workflows and lifestyles within a year or two. However, in reality, especially for To B (business-to-business) applications, the process of integrating and transforming production and workflows is relatively slow. This is because established enterprise clients have fixed procedural systems, and embedding large models into these existing workflows while ensuring no loss of functionality for the clients requires a considerable amount of time for refinement. Therefore, Luo Qing believes the industry is steadily advancing, albeit at a pace potentially slower than the capital market anticipates.
The capital market sometimes overestimates the speed of AI application rollout, but he remains optimistic in the long run. This is because, often, the long-term impact of a new technology on enterprises can be underestimated. Regarding how to balance investments between AI hardware and AI applications, Luo Qing believes that for an investment portfolio, the most crucial aspect is likely assessing the risk-reward ratio. Each phase has its own investment priorities. At the current stage, AI hardware is恰好 in a period of profit release and still has relatively clear growth prospects for the next two years. Hence, AI hardware remains one of the key investment strands.
Simultaneously, judging from their development trajectory, AI applications are gradually approaching the eve of large-scale commercialization. Consequently, he will also closely monitor the progress of AI applications, seeking out those that can genuinely generate profits and form a sustainable commercial闭环 for investment. On the topic of Hong Kong stocks, Luo Qing mentioned that the Hong Kong market does host some unique AI application companies. However, he believes that regardless of whether a company is listed on the A-share or Hong Kong market, as long as it can develop excellent products, it will inevitably gain market recognition. The key, in any market, lies in a company's ability to create great products, acquire users, and deliver an outstanding user experience. Any company that achieves these objectives is a worthwhile investment target, irrespective of its listing venue.
Comments