For the A-share market in 2025, technology was the dominant trend, leading as the core theme throughout the year. After a full year of strong performance, can the technology sector continue its upward trajectory in 2026? Today, we will analyze this from both valuation and performance perspectives.

The market has shifted from speculating on expectations to verifying actual results. Looking back from 2025 to the present, especially since July 2025, the technology sector has shown significant structural gains. Within subsectors, AI computing power and semiconductors saw increases of 69.76% and 65.31%, respectively. The primary drivers behind this rise include strong policy support, an explosion in AI computing demand, and the deepening of domestic substitution. From the current standpoint, the technology rally has largely concluded its phase of valuation repair and sentiment-driven consensus gains. The static price-to-earnings ratio for the A-share technology sector is already relatively high; for instance, the STAR 50 Index's P/E ratio has reached 175 times, near its highest level in over 20 years, placing it above the 90th historical percentile. In contrast, the Hong Kong technology sector's valuation remains at a historical low, with the Hang Seng Tech Index's P/E ratio at 21 times, situated around the 15th percentile since 2021. At this juncture, we believe market sentiment has fully priced in optimism towards technology. Sentiment indicators, such as the proportion of trading volume, margin buying amounts, and the allocation ratio of technology stocks in public funds, have reached阶段性 highs. When the market is widely discussing the AI revolution and domestic substitution, it indicates that sentiment has been fully expressed. Consequently, the market will inevitably place greater emphasis on the delivery of actual performance. We anticipate that the technology sector will transition from speculating on expectations to a phase of performance verification.

Screening for genuinely growing technology companies requires a multi-dimensional assessment. Taking semiconductors as an example, according to the latest industry chain research, order growth in the equipment segment has remained above 25% for three consecutive quarters, reflecting solid demand. However, within the computing power sector, differentiation is extreme: leading optical module companies have seen revenue double year-over-year, yet net profit margins range from 30% down to 10%. Why such disparity? Because some firms are merely assemblers, while others possess core chip technology barriers. In practice, identifying technology companies with strong performance necessitates a comprehensive evaluation across multiple dimensions: profit growth, R&D investment, competitive barriers, and industry景气度. 1) Focus on sustained high profit growth; for instance, with the AI demand surge, assess whether a company's revenue and net profit have exceeded expectations for multiple consecutive quarters. 2) Evaluate R&D investment intensity; in the semiconductor industry, the average R&D expense ratio is around 15%, with top enterprises potentially exceeding 20%. Also consider the stability of the core technical team and metrics like patent count relative to R&D expenses. 3) Assess competitive barriers: Does the company hold a leading market share in its niche, potentially with an improving trend? Is it able to form strong customer relationships, avoiding over-reliance on a single client? Does it possess proprietary technology, such as self-developed chip architectures or patent portfolios, creating a technological gap with competitors? Can it maintain a gross margin at least 10 percentage points above the industry average? 4) Industry景气度 determines the growth ceiling. This includes the state of AI infrastructure development and demand growth within various semiconductor subsectors. For example, global AI server demand is projected to increase by 83% year-over-year in 2025, driving growth in areas like liquid cooling servers and optical modules. Finally, industry景气度 should also be validated against政策支持 for specific sectors.

Based on the latest industry chain developments, AI applications are expected to gradually become clearer, but their translation into corporate earnings still requires time. In contrast, orders for computing chips, optical modules, and servers represent tangible demand. Currently, capital expenditure from major overseas and domestic players continues to compound at growth rates exceeding 20%, providing the most certain support for near-term performance. Another key area is the 'tipping point' for domestic substitution, particularly in upstream semiconductor equipment and materials. When the localization rate突破 from 10% to 30%, related enterprises often experience non-linear, explosive revenue growth. Investors may consider technology-focused products within selection pools, which include offerings such as a technology innovation fund and a stable return fund.

The information contained herein is based on publicly available data and research interviews. No guarantee is made regarding the timeliness, accuracy, or completeness of this information, nor is there any assurance that the content will not change. The content is for reference only and does not represent any definitive judgment. This document and its contents do not constitute investment advice and have not been tailored to the specific investment objectives, financial situations, or needs of any individual client. Institutions or individuals making investment decisions based on this document shall bear the investment risks themselves. All rights to this document are reserved. Reproduction, copying, publication, citation, redistribution, or any other form of infringement is prohibited without prior written permission.

*Risk Warning: Investing involves risks, and decisions should be made cautiously. The views expressed do not represent platform investment advice, and the content is for informational purposes only and should not be considered as any investment or application recommendation. Unauthorized reproduction, citation, mirroring, or linking to the content and views expressed herein is prohibited, and all legal rights are reserved.