Hong Kong Stock Review | Hong Kong Stocks Rally Slightly, Hang Seng Index Breaks Above 26,000 Points, Large Language Model Concept Stocks Hit New Highs

On March 18, the Hong Kong stock market fluctuated within a narrow range, with the Hang Seng Index slightly up.

As of the close, the Hang Seng Index was at 26,025.42 points, up 156.88 points, or 0.61%.

The Hang Seng Tech Index closed at 5,108.30 points, up 0.66 points, or 0.01%.

Market hotspots remained strong in the large model concept. By the close, MINIMAX-W (HK00100) surged 19.85%, hitting an intraday high of over 28%, and reaching a record high of HKD 1,330. The Zhipu (HK02513) also soared 19.47%, setting a new high since listing.

In news, on March 18, MiniMax announced its new generation flagship large model, Agent M2.7, showcasing the “model self-evolution” approach for the first time. The model builds an Agent Harness system, deeply involved in its own training and optimization processes, capable of handling 30%-50% of tasks in some R&D scenarios, and achieving about a 30% improvement on internal evaluation sets.

Open Securities stated that industry monitoring indicates that the surge in AI computing power demand at the start of 2026 is driving the computing power leasing market into a price increase cycle. The proliferation of AI applications and the OpenClaw framework may trigger inference demand, coupled with Nvidia’s limited capacity, rising hardware costs, and gaps in domestic substitution, leading the market into a “seller’s market.” They believe that AIDC, computing power leasing, and CDN, as core components of AI cloud IaaS, are expected to benefit significantly.

Electric power equipment stocks rose in the afternoon, with NE Electric up over 12%, Dongfang Electric over 10%, Harbin Electric over 9%, and Goldwind Technology over 6%.

Dongwu Securities believes that the explosion in AI computing power is accelerating the construction of data centers worldwide. Gas turbines, due to their fast construction and stable power generation, are becoming a key solution for data center power supply. According to Zhiyan Consulting data, from 2024 to 2027, the surge in electricity demand from AI data centers will create an incremental space of about USD 57.8 billion for the gas turbine market, providing strong support for industry development. Additionally, under the “dual carbon” goals, gas turbines, as high-efficiency, low-emission clean energy equipment, are also key to replacing coal power and supporting wind and solar power peak shaving, with domestic installed capacity continuing to grow.

In other areas, the market saw more gains than declines in tech stocks, with Bilibili, Alibaba, and Baidu rising over 2%, Kuaishou falling over 1%, and Tencent Music closing down over 21%.

In terms of funds, by the close, southbound funds had a net purchase of over HKD 1.2 billion in Hong Kong stocks.

Outlook for the future:

Huatai Securities believes that the Hong Kong stock market is already in a bottoming phase, with the index approaching a trough, but a sustained rebound may still need to wait. She pointed out that recent market resilience is due to factors such as earlier corrections in Hong Kong stocks, valuations at mid-to-low levels, a high proportion of high-dividend and cyclical stocks, and the previous accumulation of short positions providing a market floor.

Caituo Securities’ research report notes that, amid the significant adjustment of the Hang Seng Tech Index and continuous net inflows of southbound funds, there are both divergences and opportunities in the sector. From a fundamental perspective, the market maintains strong expectations for earnings growth of the constituent stocks, and these companies are actively transforming into frontier technologies like AI and semiconductors, which constitute the core long-term growth drivers of the sector. Regarding funds, since the beginning of the year, the market has shown a notable “buy more as it falls” characteristic, with southbound funds continuing to net inflow, reflecting recognition of the sector’s long-term value by both institutional and individual investors.

Disclaimer: The content and data in this article are for reference only and do not constitute investment advice. Please verify before use. Operate at your own risk.

Daily Economic News