证券研究报告 2026 年 1 月 15 日 星期四 【A 股市场大势研判】 三大指数冲高回落 市场表现： | 指数名称 | 收盘点位 | 涨跌幅 | 涨跌 | 上证指数分时图 | | --- | --- | --- | --- | --- | | 上证指数 | 4126.09 | -0.31% | -12.67 | | | 深证成指 | 14248.60 | 0.56% | 79.20 | | | 沪深 300 | 4741.93 | -0.40% | -19.10 | | | 创业板 | 3349.14 | 0.82% | 27.25 | | | 科创 50 | 1500.91 | 2.13% | 31.34 | | | 北证 50 | 1580.87 | 0.98% | 15.29 | | 资料来源：东莞证券研究所，iFinD 数据 板块排名： | 申万行业表现前五 | | | 申万行业表现后五 | 概念板块表现前五 ◼ | | 概念板块表现后五 | | | --- | --- | --- | --- | --- | --- | --- | --- | | 计算机 | 3.42% | 银行 | -1.8 ...

【市场表现】 2026年1月12日（周一），市场全天震荡走强，三大指数均涨超1%，沪深两市成交额超3.64万亿元，创 下历史新高。 截至今日收盘，沪指涨1.09%，深成指涨1.75%，创业板指涨1.82%。今日AI应用大涨，申万一级行业传 媒指数上涨7．80%，计算机指数上涨7．26%，位居涨幅榜前2。 | 序号 代码 | 涨跌幅 ▼ | | --- | --- | | 1 801760 传媒(申万) | 7.80% | | 2 801750 计算机(申万) | 7.26% | | 3 801740 国防军工(申万) | 5.66% | | 4 801210 社会服务(申万) | 3.21% | | 5 801770 通信(申万) | 2.74% | 资料来源：wind，数据截至2026/1/12。 【上涨因素分析】 1、GEO概念发酵。GEO（生成式引擎优化）是一种针对生成式AI平台（如DeepSeek、豆包等）的优化 策略，核心目标是让企业的品牌、产品或服务，能在AI生成式回答中被优先提及、精准推荐。据 Gartner预测，2026年全球传统搜索引擎流量中将有约25%转移至AI工具。在此背景之下，GEO迎 ...

国家知识产权局信息显示，大明电子（重庆）有限公司取得一项名为"一种汽车中控开关组件"的专利， 授权公告号CN223750665U，申请日期为2025年3月。 声明：市场有风险，投资需谨慎。本文为AI基于第三方数据生成，仅供参考，不构成个人投资建议。 财经频道更多独家策划、专家专栏，免费查阅>> 专利摘要显示，本实用新型提供了一种汽车中控开关组件，属于中控开关技术领域。它解决了现有汽车 中控开关组件与汽车屏幕装置连接不方便的问题。汽车中控开关组件，包括开关本体，开关本体内设有 两个相互平行设置的支撑轴，两个支撑轴上均转动配合有旋转块，旋转块上具有由开关本体靠近屏幕装 置的一侧伸出的卡扣部，位于不同支撑轴上的卡扣部的朝向相反，开关本体内设有用于使旋转块向卡扣 部的卡接方向转动的驱动件。通过卡扣部与屏幕装置连接，连接方便，开关本体与屏幕装置的连接稳定 性好，开关本体上的各部件安装方便。 天眼查资料显示，大明电子（重庆）有限公司，成立于2020年，位于重庆市，是一家以从事计算机、通 信和其他电子设备制造业为主的企业。企业注册资本33000万人民币。通过天眼查大数据分析，大明电 子（重庆）有限公司参与招投标项目19次， ...

Core Insights - The A-share IPO review process for 2025 concluded with a high-quality performance, featuring a total of 113 companies reviewed and an overall approval rate of 97.35%, marking a record high in recent years [1] Group 1: Market Performance - The main board and the ChiNext board showed the most stable performance, with the Shanghai main board (19 companies), Shenzhen main board (10 companies), and ChiNext (14 companies) achieving a 100% approval rate, indicating precise support for mature and innovative growth enterprises [1] - The Beijing Stock Exchange emerged as the primary venue for reviewing innovative small and medium-sized enterprises, with 51 companies reviewed and 49 approved, resulting in a 96% approval rate, reflecting sustained market vitality [1] - The Sci-Tech Innovation Board maintained stringent selection criteria for "hard technology," with 19 companies reviewed and 18 approved, yielding a 94.7% approval rate [1] Group 2: Underwriting Institutions - The leading underwriting institutions demonstrated a significant "winner-takes-all" effect, with Guotai Junan Securities leading with 16 approved projects, followed by CITIC Securities (12) and CITIC Jianan Securities (11), solidifying their positions in the top tier [3] - Other institutions such as Dongwu, Guojin, Guotou, Huatai United, and China Merchants Securities each contributed 5 approved projects, forming the backbone of annual IPO underwriting [3] Group 3: Industry Distribution - The capital market's effectiveness in serving the real economy has significantly improved, with resources accelerating towards high-tech manufacturing. The computer, communication, and other electronic equipment manufacturing sectors led with 20 companies reviewed, followed by specialized equipment manufacturing (17 companies), and automotive and electrical machinery manufacturing (10 companies each), highlighting the financing dominance of "hard technology" and high-end manufacturing sectors [5]

Investment Rating - The industry investment rating is "Outperform the Market - A" and the rating is maintained [7] Core Insights - NVIDIA is leading the transition of AI interconnect from scale-up and scale-out to scale-across, introducing Spectrum-XGS Ethernet scalable technology to unify distributed data centers into a gigabit-level AI super factory, establishing a new generation of AI infrastructure [12][13] - The North American cloud service provider (CSP) demand is strong, with an expected acceleration in prosperity from 2026 to 2028, as large-scale CSPs are dominating the construction of AI scale-across backbone networks [3][14] - The construction of cross-data center interconnect infrastructure is expected to benefit first, with long-term positive implications for the upstream optical communication sector [4][15] Summary by Sections Industry Performance - The computer sector has outperformed the Shanghai Composite Index by 1.01% this week, with a weekly increase of 2.89% and a year-to-date increase of 25.74% [16][17] Market Outlook - The report indicates that the interconnection of data centers has become a new competitive direction for North American CSPs, with significant revenue potential from projects like DCOM in collaboration with Meta, expected to reach hundreds of millions of dollars by 2026 [3][14] Key Recommendations - Focus on companies involved in cross-data center interconnect infrastructure, including coherent optical modules (Ciena, Cisco, etc.), switches (Arista, Cisco), and reconfigurable line systems (Ciena) [4][15]

东兴证券发布研报称，计算机行业的投资应围绕"基本面、性价比、吸引力"三维框架展开。人工智能作 为科技周期的核心主线，在政策支持、技术演进与需求释放的三重驱动下，形成了"多维共振"。从基本 面来看，"十五五"规划将"科技自立自强"列为关键目标，为AI、"新质生产力"、信创等领域提供了有力 支持。 投资策略：聚焦AI主线，关注信创景气回升及新兴/未来产业结构机会 主线上全球在AI算力上持续投入，同时应用价值逐步释放。AI算力侧关注国产AI芯片(海光信息、寒武 纪、沐曦股份、摩尔线程等)、AI服务器(中科曙光（603019）、浪潮信息（000977）、华勤技术 （603296）等)、液冷温控(曙光数创、英维克（002837）等)及智算服务(紫光股份（000938）、商汤、 华为产业链等)。AI应用侧重点布局AI Agent与端侧智能，AI Agent正从Copilot向自主决策演进，在金 融、coding、营销等高人力成本行业具有显著降本增效潜力；端侧智能受益于豆包手机等新品催化，AI 眼镜、陪伴机器人等硬件加速放量。推荐金山办公、同花顺（300033）、中科创达（300496）、科大讯 飞（002230）等公司。 ...

对于个人而言，2026年布局科创50ETF需抓住"核心三要素"：优先选择规模50亿元以上、日均成交额活跃的产品，规避迷你基金的流动性风险；重点关注 跟踪误差小、运作稳健的标的，确保精准捕捉指数收益；把握逢跌加仓的节奏，通过长期持有分享硬科技成长红利。 选对优质的科创50ETF标的，就是握住了新质生产力的入场券，剩下的只需静待时间浇灌出成长的玫瑰，在新质生产力成为经济增长新引擎的时代，科创 50ETF正为普通投资者打开一扇低成本参与硬科技浪潮的大门，2026年有望成为其规模与收益双爆发的关键之年。 （来源：淘金ETF） 2025年全年，科创50ETF在A股市场展现出强劲的"吸金"韧性，易方达、华夏等头部产品规模双双突破700亿元，北向资金对指数成分股中半导体、计算机 等硬科技权重的加仓比例攀升至18%，创下年内新高。展望2026年，在政策红利持续加码、产业周期向上突围、估值性价比凸显的三重驱动下，科创 50ETF正从科技赛道的"潜力选手"成长为全民关注的"核心资产"，成为普通投资者捕捉新质生产力红利的最优工具之一。 科创50ETF的底气，源于指数"硬科技纯度拉满"的独特基因。与沪深300的金融、能源权重，中证5 ...

Core Insights - The AI competition is shifting from model capabilities to ecosystem and implementation efficiency [1] - By December 2025, the daily token usage of Doubao's large model is expected to exceed 50 trillion, ranking third globally [1] - Over 100 companies have accumulated token usage exceeding 1 trillion on the Volcano Engine, indicating a transition from technology validation to large-scale implementation in the domestic AI industry [1] Industry Focus Areas - The first area of focus is on general agent development platform vendors, whose toolchain ecosystems will create a sticky barrier [1] - The second area is vertical industry agent solution providers, particularly those with established cases in enterprise services and content creation [1] - The third area involves small and medium-sized enterprises benefiting from cost-reduction policies, where AI application penetration is expected to increase rapidly [1] ETF and Index Information - The Computer ETF (512720) tracks the CS Computer Index (930651), which selects listed companies involved in software development, IT services, and hardware manufacturing from the A-share market [1] - This index reflects the overall performance of listed companies in the computer industry and exhibits significant growth and technology-oriented characteristics [1]

Summary of Key Points from the Conference Call Industry Overview - The conference call focuses on the **computer industry** and its transformation driven by **AI** technology, highlighting the significant recovery in profitability but noting that valuations are at historical highs, specifically at the 80th percentile [1][2]. Core Insights and Arguments - **AI as a Growth Driver**: AI is identified as the key growth point for the computer sector, with major demand expected from large enterprises in 2026, which will drive the development of cloud and public computing resources [3][4]. - **Market Dynamics**: The call discusses the competitive landscape, indicating that overseas models are beginning to consume SaaS applications, a trend expected to manifest in the domestic market with a one-year delay [4]. - **Data-Driven Transformation**: The software moat is shifting from process-driven to data-driven, with companies that possess exclusive data and demonstrate effective data-driven transformations likely to excel in various sectors such as content creation, customer service, e-commerce recruitment, taxation, and multi-modal justice [5]. Financial Performance - **2025 Recovery Indicators**: The computer industry is experiencing a weak recovery, with contract liabilities increasing by 9.6% year-on-year, revenue growth of 5.1%, and a net profit of 12.41 billion yuan, reflecting a 184% increase compared to the same period in the previous year [6]. - **Valuation Concerns**: Despite the recovery, the stock prices have already reflected these expectations, with current valuations at the 86th percentile since 2016 [6][7]. Market Configuration - **Low Allocation in Computer Sector**: As of Q3 2025, the allocation ratio for the computer industry is at a historical low of 2.3%, indicating market concerns regarding the slow penetration of AI due to weak digital infrastructure [7][8]. Future Trends and Directions - **AI Technology Evolution**: The development of AI technology is shifting from AGI to operating systems, indicating a restructuring of the tech ecosystem rather than a significant increase in productivity [9]. - **Investment Strategies of Tech Giants**: Overseas tech giants are focusing on transformative investments rather than immediate economic returns, with significant capital allocated to enhance future competitiveness [11]. - **Domestic vs. Overseas AI Development**: There is a one-year lag in domestic AI development compared to overseas, with domestic companies having advantages in engineering and localization [12]. Investment Opportunities - **Focus on Technical Resources**: Investment opportunities are primarily concentrated on the technical resources side, particularly in high-complexity scenarios such as autonomous driving, where end-to-end algorithms are becoming crucial [17]. - **Emerging Independent Applications**: In vertical applications like healthcare, taxation, and industrial sectors, independent third-party application companies are expected to emerge, requiring strong industry understanding and data foundations [18][20]. OpenAI's Growth Projections - **Revenue Growth Forecast**: OpenAI's revenue is projected to grow from $4 billion in 2024 to $200 billion by 2030, with a compound annual growth rate of approximately 92% [19]. Conclusion - The conference call highlights the transformative impact of AI on the computer industry, the financial recovery indicators, and the competitive landscape, while also identifying significant investment opportunities in emerging technologies and applications.

Core Viewpoint - Huawei Terminal Co., Ltd. has undergone significant management changes, with Yu Chengdong appointed as the new chairman, succeeding Guo Ping, alongside other executive changes [1]. Group 1: Management Changes - On December 12, Huawei Terminal Co., Ltd. experienced a corporate change where Guo Ping stepped down as chairman, and Yu Chengdong took over the role [1]. - Other executives, including Meng Wanzhou and Xu Zhijun, also resigned from their positions [1]. Group 2: Company Background - Huawei Terminal Co., Ltd. was established in November 2012, with a registered and paid-in capital of 606 million RMB, primarily engaged in the manufacturing of computers, communications, and other electronic devices [1]. - The company is wholly owned by Huawei Terminal (Shenzhen) Co., Ltd. [1]. Group 3: Yu Chengdong's Career - Yu Chengdong, born in 1969, has a long history with Huawei, joining in 1993 as a technician and progressing through various roles, including 3G product director and CEO of the consumer business group [5]. - He has been instrumental in Huawei's smartphone business growth since taking over the terminal business in 2012, focusing on Huawei's own brand rather than OEM products [5]. Group 4: Automotive Market Entry - Since 2019, Yu Chengdong has led Huawei's entry into the automotive market, establishing the Intelligent Automotive Solutions Business Unit (BU) [7]. - Huawei has adopted various collaboration models with car manufacturers, including traditional parts supplier, HI model, and the "Smart Selection" model, resulting in partnerships with several automotive companies [8]. Group 5: Sales Performance - Under Yu Chengdong's leadership, the sales of Huawei's automotive products have been notable, with the "Hongmeng Smart Travel" series achieving over 1 million deliveries in just 43 months [8]. - The top three models in sales for the first eleven months of this year were the AITO M8, AITO M9, and AITO M7, with cumulative sales of 133,300, 108,900, and 89,100 units, respectively [9].