Core Insights - China has successfully overcome key technologies in 5G communication industrial terminal measurement and testing, establishing a complete autonomous 5G industrial terminal communication quality measurement traceability system and measurement platform, which provides crucial foundational support for the high-quality development of the 5G industry and industrial digitalization upgrades [1] Group 1 - The project team has developed a universal technical indicator standard system for industrial 5G terminals, addressing the lack of communication quality evaluation standards and measurement methods [1] - The team has independently developed ultra-wideband measurement devices based on quantum principles and 4K/8K ultra-high-definition video measurement devices, enabling electromagnetic safety testing of core terminal chip components and objective evaluation of typical 5G signal content quality [1] - The measurement platform and devices have reached internationally advanced levels in precision and capability, supporting key performance verification of 5G terminals in complex industrial environments [1] Group 2 - The series of achievements has been implemented on a large scale in national-level 5G industrial application scenarios for companies such as Shaanxi Coal and China Mobile, effectively supporting the quality evaluation and standardization process of 5G industrial applications [2] - The market regulatory authority plans to continue promoting the industry application and standardization of related technological achievements, enhancing the core competitiveness of China's 5G industry chain and empowering the transformation and upgrading of the manufacturing industry and high-quality development of the digital economy [2]

Group 1: Shenzhen's Capital Landscape - As of December 18, 2025, Shenzhen has 592 listed companies, with 15 new additions in 2025, including 424 A-share companies across various exchanges [1] - Shenzhen's A-share companies have a total market capitalization of 11.34 trillion yuan, accounting for 10.83% of the national total, with a year-on-year growth of 36.82% [1] - The total revenue of Shenzhen's A-share companies reached 3.2955 trillion yuan, a year-on-year increase of 5.96%, contributing to 9.43% of the national revenue [1] Group 2: R&D Investment in Shenzhen - In the first half of 2025, Shenzhen companies disclosed R&D expenses exceeding 102.57 billion yuan, a year-on-year increase of 14.60%, representing 13.9% of the national A-share R&D expenses [2] - The R&D intensity of Shenzhen companies stands at 3.11%, leading other major cities, with manufacturing firms accounting for 90.95% of R&D spending [2] - Over 60% of A-share companies in Shenzhen have overseas revenue totaling 553.33 billion yuan, with significant contributions from manufacturing [2] Group 3: Beijing's Capital Dynamics - By December 2025, Beijing is expected to have over 800 listed companies, leading the nation, with approximately 480 A-share companies and a total market capitalization exceeding 35 trillion yuan [3] - R&D expenses for Beijing's A-share companies are around 120 billion yuan, the highest in the country, with a focus on foundational research and original innovation [3] - The Beijing Stock Exchange plays a crucial role in supporting innovative SMEs, with about 18% of its listed companies being from Beijing [3] Group 4: Shanghai's Capital Environment - Shanghai has around 680 listed companies, with approximately 430 A-share companies and a total market capitalization of about 13 trillion yuan [4] - The Shanghai Stock Exchange's Sci-Tech Innovation Board features 110 companies, accounting for 22% of the national total, with a market capitalization of about 2.8 trillion yuan [4] - Over 50% of Shanghai's Sci-Tech companies have more than 30% of their revenue from overseas, indicating a strong international presence [4] Group 5: Comparative Analysis of the Three Cities - Beijing's capital structure is characterized by state-owned enterprises contributing over 65% of its total market value, while Shanghai has a balanced multi-industry layout [5] - Shenzhen focuses on rapid industrialization and iteration, while Beijing emphasizes foundational research and Shanghai prioritizes application and internationalization [5] - The financial sector in Beijing is heavily regulated, while Shanghai's financial services are market-oriented, and Shenzhen's financial sector is closely integrated with manufacturing [5] Group 6: Complementary Capital Ecosystem - The three cities are forming a complementary capital ecosystem, with Beijing evolving into an "innovation source," Shanghai solidifying its "capital hub" status, and Shenzhen enhancing its role as an "industrial innovation arena" [6] - Investment from Beijing's venture capital firms in Shenzhen's tech companies increased by 31% year-on-year, indicating growing inter-city capital flow [6] - The differences in company numbers and market values reflect the strategic positioning and resource endowments of each city, contributing to a resilient national innovation system [6]

Core Viewpoint - The low-altitude economy is recognized as a new blue ocean for industrial upgrading, with a focus on technological innovation as the core driver for high-quality development. The Chinese government has included the low-altitude economy in its strategic emerging industry development plan for the 14th Five-Year Plan period, emphasizing the need for a robust industrial ecosystem in this sector [1]. Group 1: Technological Advancements - China Mobile Guangdong has made significant progress in low-altitude communication, transitioning from single-station functionality to network collaboration, and successfully commercializing 5G-A integrated sensing technology in complex urban environments [2]. - The company has completed a year-long large-scale verification test for low-altitude communication in Qingyuan, marking a substantial breakthrough for the application of mobile 5G-A technology in urban settings [2]. Group 2: Industry Standards and Collaboration - Challenges such as inconsistent testing standards and inadequate certification systems in the low-altitude intelligent network sector hinder high-quality industry advancement. China Mobile Guangdong is collaborating with the Guangdong Low Altitude Economic Development Company to establish a testing base focused on developing industry standards and validating equipment [3]. Group 3: Applications of Low-Altitude Technology - The integration of AI algorithms for forest fire prevention and drowning risk management has enhanced emergency response efficiency through automated drone operations, showcasing the potential of low-altitude technology in various applications [4]. - The company has developed a low-altitude capability platform that integrates multiple core technologies to address safety challenges in the low-altitude sector, including real-time data collection and analysis for airspace management [5]. Group 4: Smart Applications and Security - China Mobile Guangdong has implemented a comprehensive intelligent inspection and defense system using 5G-A technology to protect critical power supply lines during major events, addressing vulnerabilities to unauthorized drone incursions [6][7]. - The company has established a rapid response mechanism for unauthorized drone activities, enabling real-time tracking and monitoring of airspace targets, thereby enhancing security measures in low-altitude operations [7]. Group 5: Strategic Development - By focusing on low-altitude, 5G-A, and artificial intelligence technologies, China Mobile Guangdong aims to drive new productive forces and contribute to high-quality development in Guangdong, positioning the region at the forefront of China's modernization efforts [7].

Core Insights - The top three companies by stockholding ratio in the Hong Kong Stock Connect are China Telecom (71.99%), Gree Power Environmental (70.09%), and Da Zhong Public Utilities (68.82%) [1] - Tencent Holdings, Xiaomi Group-W, and Alibaba-W have seen the largest increases in stockholding amounts over the last five trading days, with increases of +1.201 billion, +1.001 billion, and +892 million respectively [1] - The companies with the largest decreases in stockholding amounts over the last five trading days include China Mobile (-2.320 billion), Yingfu Fund (-1.467 billion), and China Petroleum & Chemical Corporation (-364 million) [2] Stockholding Ratio Rankings - China Telecom (00728) holds 9.990 billion shares, representing 71.99% [1] - Gree Power Environmental (01330) holds 0.283 billion shares, representing 70.09% [1] - Da Zhong Public Utilities (01635) holds 0.367 billion shares, representing 68.82% [1] - Other notable companies in the top rankings include Kai Sheng New Energy (67.96%) and Tianjin Chuangye Environmental Protection (67.10%) [1] Recent Increases in Stockholding - Tencent Holdings (00700) saw an increase of +1.201 billion, with a change of +1.9922 million shares [1] - Xiaomi Group-W (01810) increased by +1.001 billion, with a change of +25.5161 million shares [1] - Alibaba-W (09988) increased by +892 million, with a change of +6.1091 million shares [1] - Other companies with significant increases include Hong Kong Exchanges (6.48 billion) and Changfei Optical Fiber (6.09 billion) [1] Recent Decreases in Stockholding - China Mobile (00941) experienced a decrease of -2.320 billion, with a change of -28.0154 million shares [2] - Yingfu Fund (02800) decreased by -1.467 billion, with a change of -56.4370 million shares [2] - China Petroleum & Chemical Corporation (00857) decreased by -364 million, with a change of -44.7216 million shares [2] - Other companies with notable decreases include China Pacific Insurance (-350 million) and Innovent Biologics (-293 million) [2]

来源：光大证券微资讯 热点聚焦 1、隔夜，美股三大指数集体收涨，均录得日线5连涨。道指涨0.6%，标普500涨0.32%，纳指涨0.22%。 其中，标普500、道指均创历史收盘新高。当天盘前，英特尔股价直线跳水，盘中跌幅一度超过5%，收 盘跌0.52%。有报道称，英伟达已暂停对英特尔18A制程工艺的测试。同时，英伟达称并未收购AI芯片 初创公司Groq，仅为技术许可安排。此前报道称，英伟达拟以约200亿美元收购Groq。 2、隔夜，国际贵金属期货收盘涨跌不一。COMEX黄金期货持平，报4505.4美元/盎司；COMEX白银期 货涨1.04%，报71.88美元/盎司。12月25日1:16，现货白银短线拉升，向上触及72美元/盎司，日内上涨 0.8%。此前，黄金、白银在创下历史新高之后，一度跳水。北京时间25日5:59纽约尾盘，离岸人民币兑 美元报7.0074元，较周二纽约尾盘涨119点。 3、据环球时报，美国政府23日宣布，将在2027年对华芯片加征关税，结束了拜登政府发起的对华芯片 贸易调查。美媒分析称，尽管美国政府称中方在芯片产业中的做法"损害美国利益"，但最终决定至少在 18个月内不对华芯片加征额外关税。 ...

Core Insights - Tencent Holdings (00700), Alibaba-W (09988), and Hong Kong Exchanges (00388) led the market in net inflows, with amounts of 1.115 billion, 1.055 billion, and 309 million respectively [1] - China Mobile (00941), Luoyang Molybdenum (03993), and CICC (03908) experienced the highest net outflows, with amounts of -1.053 billion, -184 million, and -181 million respectively [1] - In terms of net inflow ratios, China Telecom Services (00552), Yihua Tong (02402), and Sunshine Insurance (06963) topped the list with ratios of 167.31%, 126.18%, and 118.17% respectively [1] Net Inflow Rankings - Tencent Holdings (00700) had a net inflow of 1.115 billion, representing a 13.09% increase, closing at 614.500 [2] - Alibaba-W (09988) saw a net inflow of 1.055 billion, with a 12.23% increase, closing at 146.400 [2] - Hong Kong Exchanges (00388) recorded a net inflow of 309 million, with a 24.77% increase, closing at 407.000 [2] Net Outflow Rankings - China Mobile (00941) faced a net outflow of -1.053 billion, a decrease of -54.21%, closing at 83.700 [2] - Luoyang Molybdenum (03993) had a net outflow of -184 million, down by -29.82%, closing at 19.100 [2] - CICC (03908) experienced a net outflow of -181 million, a decrease of -30.22%, closing at 20.540 [2] Net Inflow Ratio Rankings - China Telecom Services (00552) achieved a net inflow ratio of 167.31%, with a net inflow of 26.5926 million, closing at 4.520 [3] - Yihua Tong (02402) had a net inflow ratio of 126.18%, with a net inflow of 26.5481 million, closing at 23.980 [3] - Sunshine Insurance (06963) recorded a net inflow ratio of 118.17%, with a net inflow of 31.5558 million, closing at 3.940 [3]

本报记者 王云杉 "过马路时，你要注意看车!" "我计划明天去车展看车。" 这两句话里的"看车"是一个意思吗？相信不少人要会心一笑，表面上看是同一个词组，但其含义因语境 不同发生了变化。 这就是中文里常见的"一词多义"现象。人工智能大模型是一种与人类语言密切相关的技术，要让大模型 深刻理解这一现象，离不开中文数据的持续供给。 目前，国内多数模型训练使用的数据，中文数据占比已经超过60%，有的模型达到80%。大模型训练 中，中文数据占比提升有何意义？中文高质量数据为何持续增加？如何进一步增加中文数据的开发与供 给？记者进行了采访。 数据就像大模型的"知识教材" 不同语言的数据对大模型性能有怎样的影响？"数据就像大模型的'知识教材'，教材的语言属性不同， 会对模型的知识体系产生不同影响。"清华大学计算社会科学与国家治理实验室执行主任、教授孟庆国 表示。 从知识来源看，过去我国大模型常面临"数据依赖"风险——英文数据在全球互联网的占比较高，如前沿 科技论文、行业标准、文化典籍等多以英文呈现，全球高质量标注数据也多以英文为主。 "语言类大模型一般需要遵循一定的语言习惯。"工业和信息化部信息通信经济专家委员会委员盘和林 ...

"过马路时，你要注意看车！" "我计划明天去车展看车。" 这两句话里的"看车"是一个意思吗？相信不少人要会心一笑，表面上看是同一个词组，但其含义因语境 不同发生了变化。 这就是中文里常见的"一词多义"现象。人工智能大模型是一种与人类语言密切相关的技术，要让大模型 深刻理解这一现象，离不开中文数据的持续供给。 目前，国内多数模型训练使用的数据，中文数据占比已经超过60%，有的模型达到80%。大模型训练 中，中文数据占比提升有何意义？中文高质量数据为何持续增加？如何进一步增加中文数据的开发与供 给？记者进行了采访。 数据就像大模型的"知识教材" 不同语言的数据对大模型性能有怎样的影响？"数据就像大模型的'知识教材'，教材的语言属性不同， 会对模型的知识体系产生不同影响。"清华大学计算社会科学与国家治理实验室执行主任、教授孟庆国 表示。 从知识来源看，过去我国大模型常面临"数据依赖"风险——英文数据在全球互联网的占比较高，如前沿 科技论文、行业标准、文化典籍等多以英文呈现，全球高质量标注数据也多以英文为主。 "语言类大模型一般需要遵循一定的语言习惯。"工业和信息化部信息通信经济专家委员会委员盘和林认 为，中文数据占 ...

Core Viewpoint - The Hang Seng Index (HSI) is currently experiencing a consolidation phase, with mixed signals from various sectors, indicating potential investment opportunities and risks in the market [1][3]. Market Overview - On the previous day (23rd), the HSI closed at 25,774.14 points, a slight decrease of 0.11%, with a total turnover of HKD 157.13 billion, reflecting a general market consolidation [1]. - The short-term support levels for the HSI are at 25,448 points and 25,116 points, while resistance levels are at 26,147 points and 26,432 points, with a 53% probability of an upward movement [1]. Technical Analysis - The HSI is currently oscillating between the 10-day moving average (MA10) of 25,614.57 points and the 30-day moving average (MA30) of 25,878.94 points, indicating significant pressure from the upper moving averages [1]. - Various technical indicators are showing neutral signals, with the RSI at 52, while the VR ratio indicates a "sell" signal and the MACD shows a "buy" signal, suggesting a stalemate between bulls and bears [1]. Sector Performance - **Technology Stocks**: Tencent (00700) closed at HKD 602, down 2.03%, with technical indicators showing a "buy" signal but facing short-term resistance. Alibaba (09988) rose 0.55% to HKD 147.20, while JD.com (09618) increased by 0.45% to HKD 112.80, with resistance at HKD 116.80 [1][2]. - **Financial Stocks**: HSBC Holdings (00005) rose 0.82% to HKD 122.30 but is in the overbought zone with an RSI of 78, indicating a "strong sell" signal. AIA Group (01299) remained flat at HKD 82.65, needing to break through HKD 83.50 to confirm an upward trend [2]. - **Consumer Stocks**: Meituan (03690) rose 0.39% to HKD 103.20 but shows a "strong sell" signal, while Xiaomi (01810) fell 1.51% to HKD 39.20, indicating an oversold state [2]. - **Traditional Blue Chips**: China Mobile (00941) fell 1.02% to HKD 82.85, with a weak "buy" signal, while CNOOC (00883) decreased by 1.15% to HKD 20.56, needing to hold above HKD 20.10 [2]. Investment Opportunities - Recent performance of HSI bull certificates shows significant elasticity, with Bank of China bull certificate (68350) rising 6% and (68194) rising 8% within two days, while the HSI only increased by 0.33% [3]. - Selected high-cost performance products include UBS call warrants (19853) and Bank of China call warrants (19861), both with a strike price of 27,135 points and high leverage, suitable for investors expecting a breakout [3][4]. Bearish Instruments - For bearish strategies, JPMorgan put warrants (20889) and Bank of China put warrants (20721) offer high leverage and low premiums, suitable for short-term traders expecting a decline in the HSI [4][5].

Core Insights - The 2025 Computing Power Internet Conference was held in Chengdu, Sichuan, where the national computing power internet service platform's cross-domain system was officially launched, along with the Sichuan hub service node [1][3]. Group 1: National Computing Power Internet Service Platform - The official launch of the national computing power internet service platform's cross-domain system marks the preliminary validation of the "1+M+N" architecture, which includes one national service platform, multiple regional platforms, and various industry platforms [3]. - The platform aims to manage the interconnectivity of computing power across different regions and industries, with significant participation from major telecom operators like China Telecom, China Mobile, and China Unicom [3]. Group 2: Sichuan Hub Service Node - The Sichuan hub service node is the first in the country to establish a cross-service provider and cross-architecture supply-demand matching service system, integrating resources from various local telecom operators and service providers [5]. - It features the first intelligent service "supermarket" for computing power products in China, promoting a shift from traditional infrastructure leasing to a service-oriented supply model [5]. - The platform also boasts the first minute-level monitoring capability for computing resource status, enhancing the usability of computing power in Sichuan [5]. Group 3: Importance of Computing Power - Computing power is recognized as a core engine driving the digital transformation of industries and a key indicator of national competitiveness [6]. - The China Academy of Information and Communications Technology emphasizes the need to implement the "1+M+N" multi-level platform system to support the development of computing power interconnectivity and contribute to the growth of the artificial intelligence industry and high-quality digital economy [6].