（文章来源：每日经济新闻） 每经AI快讯，2月26日，港股收盘，恒生指数收跌1.44%，恒生科技指数收跌2.87%，电力、光模块、光 通信和半导体板块走强，东方电气涨超15%，剑桥科技涨8%，豪威集团涨超6%，长飞光纤涨近3%，兆 易创新涨逾3%。互联网科技股普遍走低，哔哩哔哩跌超4.5%，快手跌4%，阿里巴巴跌逾3.5%，腾讯控 股跌超2%。 ...

每经AI快讯，2月26日，港股收盘，恒生指数收跌1.44%，恒生科技指数收跌2.87%，电力、光模块、光 通信和半导体板块走强，东方电气涨超15%，剑桥科技涨8%，豪威集团涨超6%，长飞光纤涨近3%，兆 易创新涨逾3%。互联网科技股普遍走低，哔哩哔哩跌超4.5%，快手跌4%，阿里巴巴跌逾3.5%，腾讯控 股跌超2%。 ...

Group 1 - The article highlights the top and bottom AH premium rates for various stocks, with Northeast Electric (00042) leading at 527.91% and Ningde Times (03750) at -15.33% [1][2] - The top three stocks with the highest AH premium rates are Northeast Electric (00042), Zhejiang Shibao (01057), and Sinopec Oilfield Service (01033) [1] - The bottom three stocks with the lowest AH premium rates include Ningde Times (03750), Zhaoyi Innovation (03986), and China Merchants Bank (03968) [1] Group 2 - The article provides detailed tables showing the AH premium rates and deviation values for the top and bottom stocks, indicating significant discrepancies between H-shares and A-shares [1][2] - Sinopec Oilfield Service (01033) has the highest deviation value at 25.11%, while Northeast Electric (00042) has the lowest at -270.69% [1][2] - The data is generated based on the Hong Kong Stock Exchange and reflects the current premium rates compared to the average rates over the past 30 days [2]

Core Viewpoint - The data indicates significant outflows of main funds from various stocks, with notable amounts in the energy, materials, and financial sectors [1][2][3] Group 1: Stock Outflows - The stock with the highest outflow is 阳光电源, with a total of -20.02 billion yuan [1][2] - 宁德时代 follows closely with an outflow of -18.87 billion yuan [1][2] - Other significant outflows include 北方稀土 at -14.76 billion yuan and 兆易创新 at -11.52 billion yuan [1][2] Group 2: Sector Analysis - The electric equipment sector shows substantial outflows, with 阳光电源 and 宁德时代 leading [2] - The non-ferrous metals sector is also impacted, with 北方稀土 and 紫金矿业 experiencing notable outflows [2][3] - The financial sector, represented by 中国平安 and 中信证券, also shows significant fund outflows [2][3] Group 3: Additional Notable Stocks - 贵州茅台 has an outflow of -9.42 billion yuan, indicating a decline in investor interest [1][3] - 立讯精密 and 恩捷股份 also report outflows of -8.23 billion yuan and -6.35 billion yuan respectively, reflecting trends in the electronics and energy sectors [1][3] - 比亚迪, a key player in the automotive sector, shows an outflow of -5.95 billion yuan, suggesting potential shifts in market sentiment [1][3]

Investment Rating - The report provides a "Buy" rating for the industry, indicating an expectation of stock performance exceeding the market by more than 10% over the next 12 months [45]. Core Insights - SRAM (Static Random Access Memory) is identified as a high-bandwidth on-chip storage layer that can significantly enhance AI inference speed by reducing latency and jitter compared to external HBM (High Bandwidth Memory) [3][11]. - The architecture of SRAM is gaining mainstream attention, with significant investments and partnerships, such as Nvidia's $20 billion acquisition of Groq's intellectual property and OpenAI's $10 billion contract with Cerebras [3][32]. - The report emphasizes the growing importance of AI memory-related upstream infrastructure, suggesting that investors should focus on key beneficiaries within the industry chain [3][39]. Summary by Sections SRAM as a High-Bandwidth Storage Layer - SRAM is positioned as an essential component in the multi-tier storage architecture, providing high bandwidth but with limited capacity and higher costs [3][11]. SRAM Enhancing AI Inference Speed - SRAM can improve AI inference speed, with examples such as Groq's LPU chip achieving a bandwidth of 80 TB/s and maintaining stable inference speeds of 275-276 tokens/s, outperforming other platforms [3][15][21]. - Cerebras' WSE-3 chip integrates 44GB of SRAM, achieving over 3000 tokens/s in inference tasks, significantly faster than mainstream GPU cloud inference [3][23][39]. SRAM Architecture Gaining Mainstream Attention - The report notes that major companies are investing in SRAM technology, highlighting Groq's partnership with Nvidia and Cerebras' funding round that values the company at $23 billion [3][32][39]. Investment Recommendations - The report suggests that the ongoing expansion of AI memory capabilities will enhance model performance and accelerate the deployment of AI applications, recommending a focus on core beneficiaries in the industry chain [3][39].

Group 1 - The core viewpoint of the articles highlights the significant rise in AI-related stocks, particularly in the hardware sector, driven by advancements in AI coding capabilities and the anticipated growth of AI applications by 2026 [1][2] - The AI coding capability is expected to lead to an exponential increase in the volume of effective code globally, indicating a new phase of competition and market dynamics that investors should consider [2] - OpenAI's interview with Peter Steinberger emphasizes the transformative potential of AI in software development, suggesting that the focus will shift from coding to defining intentions, with a prediction of a breakthrough in Agentic Coding by 2026 [1][2] Group 2 - The communication ETF Huaxia (515050) focuses on electronic and communication sectors, with top holdings including Zhongji Xuchuang and Xinyi Sheng, indicating a strong emphasis on hardware related to AI [3] - The entrepreneurial AI ETF Huaxia (159381) has a balanced allocation between hardware and AI software applications, with significant holdings in Zhongji Xuchuang and Xinyi Sheng, and a fund size nearing 2 billion [3] - The cloud computing ETF Huaxia (516630) emphasizes domestic AI software and hardware, with a high combined weight in computer software, cloud services, and computer equipment, showcasing a strong focus on AI capabilities [4]

摘要 ■ 投资逻辑 基于大模型对券商年度展望的全面解析 每年的年底到年初，各券商会密集发布关于总量、行业等年度报告。对投资者而言，难以整合大量研报信息从而提取共识点，此外分歧点也同样缺乏量化 对比。大模型工具的出现解决了这一难点，大模型可以帮助分析师快速完成研报数据预处理、观点精准抓取与分类；大模型通过确定共识点与分歧点的判 定规则，系统整合分散观点，量化呈现各券商对板块不同观点的支持率，清晰核心分歧维度。本次研究以大模型为核心分析工具，全面剖析2026年各大板 块年度展望，梳理宏观、策略、固收及七大细分行业的共识与分歧，为投资者提供高效精准的决策参考。 总量观点总结 对于总量观点，我们分别从宏观经济、年度策略、固定收益三个方面分别剖析了各券商分析师的年度展望。 在宏观经济方面，大多数券商认为随着"十五五"政策发力，2026年GDP目标为5%左右；CPI、PPI从低位逐步修复，国内通货膨胀的压力得到缓解；货币 政策维持适度宽松基调，流动性保持合理充裕；财政政策延续积极取向，赤字率维持合理高位；出口保持稳健，贸易顺差维持历史较高水平；国外环境虽 然出现波动，但整体风险可控。 作者：高智威、许坤圣 在年度策略方面 ...

2026年以来，港股IPO市场延续火热态势，募资总额较去年同期大增10倍。据数据统计，按上市日计 算，2026年1月1日至2月26日，共有24家公司在港交所挂牌上市，较去年同期增加15家；IPO募资额 892.26亿港元，较去年同期大增逾10倍。令人瞩目的是，AI赛道掀起的资本热潮仍在继续，壁仞科技、 智谱、MINIMAX等AI企业均完成港股上市。 据记者不完全统计，2026年以来，已有147家公司向港交所提交IPO招股书。截至目前，处于"待上 市""通过聆讯""处理中"的港交所排队企业391家，其中110家为A股上市公司。据港交所行政总裁陈翊庭 在马年新春开市仪式致辞时透露，目前港交所有488家企业正在排队等候上市，其中超过10家为国际公 司。 刚刚过去的2025年，港股IPO市场迎来强势复苏，全年共有119只新股挂牌上市，IPO募资总额高达 2856.93亿港元，IPO上市数量和募资规模双双攀升。这也是港交所时隔4年IPO募资规模再度突破2000亿 港元整数关，成功夺回全球IPO募资额榜首。 进入2026年，港股IPO市场的强劲势头不减。今年前两个月，已有24家企业成功上市，较去年同期增长 66.67%； ...

半导体板块上涨，中韩半导体ETF今日继续飙涨8%，年内涨幅超66%，成为全市场表现最好的ETF。 海力士在会议中透露了一个关键事实："今年没有任何一家客户能够完全满足其存储需求。"这意味着所 有终端市场的需求满足率都处于低位。从库存角度看，服务器客户的库存已达到健康水平，而PC和移 动客户的库存正呈现下降趋势。对于市场最关注的HBM（高带宽内存），海力士明确表示，2026年的 产能分配已成定局，"2026年的HBM已全部售罄，满足客户需求的生产计划已经分配完毕。" 申万宏源证券认为，本轮涨价潮呈现出全链条传导、海内外共振的鲜明特征，在AI资本开支加速与金 银铜等金属上行共振下，于26年一季度形成"淡季不淡"的全面通胀格局。全球半导体产业链正迎来一轮 覆盖全环节、多品类的系统性涨价潮，从上游晶圆制造、封装测试的产能报价，到中游存储芯片、 MCU、模拟芯片、功率半导体的终端产品定价，再到配套的电阻、电感等被动元件、连接器等核心辅 材，全产业链条同步开启价格重塑。当前电子产业链呈"三线并进"的系统性涨价：景气拉动型（存储、 CPU、ABF载板）、成本传导型（CCL/电子布/铜箔、被动元件、封测/代工）、供给收缩（利 ...

人工智能算力方向延续强势，通信ETF华夏、通信ETF广发、通信ETF银华、通信ETF嘉实、通信ETF、 通信设备ETF涨超2%。 | 证券代码 | 证券简称 | 跟踪指数 | 当日涨幅% | 年内涨幅% | 管理人 | | --- | --- | --- | --- | --- | --- | | 515050.SH | 通信ETF华夏 | 5G通信 | 2.87% | 8.57% | 华夏星美 | | 159507.SZ | 通信ETF广发 | 国证通信 | 2.87% | 10.03% | 广发基金 | | 159994.SZ | 通信ETF银华 | 5G通信 | 2.65% | 8.39% | 银华基金 | | 159695.SZ | 通信ETF嘉实 | 国证通信 | 2.49% | 10.12% | 嘉实基金 | | 515880.SH | 通信ETF | 通信设备 | 2.49% | 8.25% | 国泰基金 | | 159583.SZ | 通信设备ETF | 通信设备主题 40 40 mix, ... J 14 11 10, 30 0 0 0 0 0 0 0 10 10 10 10 10 10 | ...