站在2026年的起点重新审视，在2025年这一复杂多变的市场环境中，大众公用为何能够脱颖而出、获得 市场认可？其双轮驱动战略又能否为2026年提前铺路？ 一、2025年终复盘：市场表现亮眼，盈利韧性显现 回顾2025年，大众公用全年表现非常亮眼。 在近期的格隆汇"金格奖"年度卓越公司评选中，大众公用（600635.SH/01635.HK）荣获"年度A+H卓越 企业奖"。 2025年，大众公用A股股价累计涨幅达52.57%，全年最高涨幅一度超过90%；H股表现更为突出，全年 涨幅达86.72%，最高涨幅超过160%。这一表现，反映出市场对公司价值重估的积极反馈。 该奖项从A+H双平台运作能力、业务成长性、财务稳健性、公司治理以及资本市场表现等多个维度， 对上市公司的2025年表现进行综合评价。 图：大众公用A股2025年年内走势 股价背后，是盈利能力的实质性改善。 2025年前三季度，公司实现了净利润同比超两倍的增长，其扣非净利润也达到3.09亿元，同比增长 104.79%，显示主业盈利质量出现明显提升。 与此同时，公司持续优化财务结构。 截至2025年三季度末，公司资产负债率为54.06%，保持在相对合理水平 ...

证券研究报告：计算机|点评报告 | 收盘点位 | | 5425.46 | | --- | --- | --- | | 52 | 周最高 | 5841.52 | | 52 | 周最低 | 3966.07 | 行业相对指数表现（相对值） -5% 0% 5% 10% 15% 20% 25% 30% 35% 40% 2025-01 2025-03 2025-06 2025-08 2025-10 2026-01 计算机 沪深300 资料来源：聚源，中邮证券研究所 研究所 分析师:陈涵泊 SAC 登记编号:S1340525080001 Email:chenhanbo@cnpsec.com 分析师:李佩京 SAC 登记编号:S1340525080003 Email:lipeijing@cnpsec.com 分析师:王思 SAC 登记编号:S1340525080002 Email:wangsi1@cnpsec.com 近期研究报告 发布时间：2026-01-09 行业投资评级 强于大市|维持 行业基本情况 《技术与资本共振，国产大模型护航 AI 应用浪潮》 - 2026.01.05 大厂需求加速，字节 AIDC 和算力链有望 ...

花旗预计，台积电将把2026年资本开支指引设定在460亿至480亿美元区间，英特尔资本开支将趋于稳定， 而三星则可能追随美光的扩产步伐上调投资。三家公司合计占花旗2026年全球WFE支出模型的约59%，其 资本开支动向对整个半导体设备产业链具有风向标意义。 花旗预计半导体三巨头台积电、三星和英特尔即将公布的2026年资本开支指引将呈现积极态势，这将为全 球晶圆厂设备(WFE)市场注入强劲动能。该行维持半导体设备行业正处于第二阶段上行周期的判断，并认 为2026年全球WFE支出正朝着其乐观情景下的1260亿美元迈进，高于基准预测的1150亿美元。 花旗在1月5日发布的研报中指出， 预计台积电将把2026年资本开支指引设定在460亿至480亿美元区间， 英特尔资本开支将趋于稳定，而三星则可能追随美光的扩产步伐上调投资。三家公司合计占花旗2026年全 球WFE支出模型的约59%，其资本开支动向对整个半导体设备产业链具有风向标意义。 花旗分析师Atif Malik在报告中表示，三巨头更高水平的资本开支将与更高的增长机遇相关联。该行基准模 型假设2026年NAND存储芯片设备支出增长30%，DRAM增长12%，前端逻辑 ...

Group 1 - The core viewpoint is that the overseas AI computing power sector is expected to experience strong support due to significant capital expenditures from major cloud providers, with a quarterly capital expenditure exceeding $100 billion and an annualized figure of approximately $400 billion [1][2] - The capital expenditures of North American cloud providers have shown a consistent quarterly growth trend since 2023, leading to a rapid increase in their revenues, despite the absence of standout applications [2][3] - The ongoing investment in AI is expected to empower traditional businesses of cloud providers, resulting in sustained growth in AI-related revenues, which in turn supports ongoing capital expenditures [3][4] Group 2 - Google is emerging as a global AI superpower, excelling in various segments from TPU chips to large models and applications, with its recent Gemini3Pro model setting new performance benchmarks [3][4] - The demand for computing power is rapidly increasing, with Google's AI infrastructure reportedly doubling its computing needs every six months, indicating strong growth potential in both training and inference dimensions [5][6] - The collaboration between departments within Google has improved significantly in 2023, enhancing its capabilities in AI and further supporting the sustained demand for computing power [6]

在此之前，赣锋锂业（002460）、天齐锂业（002466）也已经先后向业内释放出了乐观预期。两家全球性行业龙头普遍认为，受益于可再生能 源并网、商用重卡等电动化设备等领域的需求增长，2026年全球锂需求有望增长至200万吨左右，基本达到供需平衡。 截至12月22日，国内电池级碳酸锂市场均价升至9.9万元/吨，创出2024年6月以来的新高。 期货价格更高。即将交割的LC2601合约最新价达到11.25万元/吨，反映明年价格预期的LC2605等远月合约则全面逼近12万元/吨。 天齐锂业 赣锋锂业 天齐锂业 赣锋锂业 分时图 日K线 周K线 月K线 55.78 1.86 3.45% 4.36% 2.91% 1.45% 0.00% 1.45% 2.91% 4.36% 51.57 52.35 53.14 53.92 54.70 55.49 56.27 09:30 10:30 11:30/13:00 14:00 15:00 0 59万 119万 178万 四季度以来，锂行业内部的乐观情绪迅速蔓延。 五矿证券等机构普遍认为，锂盐价格经过2023-2025连续三年下跌后，行业资本开支大幅降低，2026到2028年锂行业供给 ...

Core Viewpoints - The fragility of capital expenditure will manifest through deteriorating liquidity, with potential financial risks arising from interconnected transactions and off-balance-sheet financing. The "political-liquidity-narrative" framework is identified as a key source of external volatility [2] Group 1: AI Investment Bubble - Many believe that there is no bubble in the AI investment sector, citing the healthy revenue and cash flow of tech giants compared to the dot-com bubble era. However, this comparison overlooks fundamental differences in scale and concentration of AI investments today [7] - The value of AI in enhancing productivity across industries will take a long time to materialize, as organizational and process changes lag behind technological advancements. AI currently serves more as a predictive tool rather than a decision-making replacement [9] - Despite the long-term nature of AI's impact on productivity, investment in AI has become a market consensus, driven by various stakeholders including tech companies, financial institutions, and media [10] Group 2: Capital Expenditure Vulnerability - From Q3 2025, capital expenditures among major tech firms investing heavily in AI reached $105.77 billion, a 72.9% year-on-year increase. This surge raises concerns about cash flow sustainability, with the average Capex/CFO ratio rising by 29.7 percentage points to 75.2% [24] - Projections indicate that by Q2 2027, the average Capex/CFO ratio for these firms could reach 95.9%, nearing the peak levels seen during the dot-com bubble [25] - The potential for negative free cash flow could deepen vulnerabilities, particularly for firms like Meta, which may face a cash flow crisis by Q4 2026 [32] Group 3: Financial Risks from High Leverage and Off-Balance-Sheet Financing - In the first 11 months of the year, the total issuance of corporate bonds by hyperscaler companies reached $103.8 billion, significantly exceeding previous years. This surge has led to increased bond spreads and heightened financial risk [39] - Companies like Meta are employing off-balance-sheet financing strategies to manage massive capital needs while maintaining favorable financial statements. This approach poses significant risks, especially if technology bubbles burst or market conditions shift [42][43] Group 4: Political Uncertainty and Liquidity Risks - The sustainability of the AI narrative is closely tied to liquidity conditions, which have been bolstered by recent interest rate cuts. However, political uncertainties, particularly surrounding upcoming elections, could tighten liquidity and impact market sentiment [44][48] - The interplay between political decisions and liquidity will likely lead to increased volatility in the stock market, particularly for AI-related investments [50]

宏观经济 | 动态跟踪 报告发布日期 2025 年 12 月 10 日 | 孙国翔 | 执业证书编号：S0860523080009 | | --- | --- | | | sunguoxiang@orientsec.com.cn | | | 021-63326320 | | 黄汝南 | 执业证书编号：S0860525120004 | | | huangrunan@orientsec.com.cn | | | 010-66210535 | | 孙金霞 | 执业证书编号：S0860515070001 | | | sunjinxia@orientsec.com.cn | | | 021-63326320 | | 王仲尧 | 执业证书编号：S0860518050001 | | | 香港证监会牌照：BQJ932 | | | wangzhongyao1@orientsec.com.cn | | | 021-63326320 | | 陈至奕 | 执业证书编号：S0860519090001 | | | 香港证监会牌照：BUK982 | | | chenzhiyi@orientsec.com.cn | | | 021-6332 ...

Core Insights - The report indicates that major overseas tech companies have exceeded market expectations in their Q3 2025 earnings, showcasing strong revenue growth and robust cloud business performance [1][11][33] - Capital expenditures (CapEx) for these companies continue to rise significantly, driven by investments in AI and cloud infrastructure, raising concerns about profit margins and return on investment [2][58][61] Company Summaries Microsoft - Microsoft reported Q1 FY26 revenue of $77.67 billion, a year-on-year increase of 18%, slightly above market expectations [11] - The intelligent cloud segment generated $30.9 billion in revenue, growing 28% year-on-year, with Azure cloud services seeing a remarkable 40% growth [12][15] - Capital expenditures reached $34.9 billion, a 74.5% increase year-on-year, primarily focused on AI and infrastructure investments [14][15] Meta - Meta's Q3 2025 revenue was $51.24 billion, up 26.25% year-on-year, surpassing both the upper limit of guidance and market expectations [19] - The company faced a significant drop in net profit due to a one-time tax asset impairment, but adjusted net profit was $18.6 billion [19][21] - Capital expenditures totaled $19.37 billion, primarily for servers and data centers, exceeding market expectations [23][26] Google - Google achieved Q3 2025 revenue of $102.35 billion, a 15.95% increase year-on-year, with net profit rising 32.99% [33][36] - The Google Cloud segment reported revenue of $15.16 billion, growing 33.51% year-on-year, driven by strong demand for AI products [36][42] - Capital expenditures for Q3 were $23.95 billion, with a forecasted increase for the full year to $91-93 billion [42] Amazon - Amazon's Q3 2025 revenue reached $180.17 billion, a 13% year-on-year increase, with net profit rising 38% [43][46] - AWS revenue was $33.01 billion, marking a 20% year-on-year growth, the highest quarterly growth rate in 2023 [46][52] - Capital expenditures were $34.2 billion, a 61% increase year-on-year, with expectations for continued growth in FY2026 [52][56] Market Trends - The report highlights a trend of increasing capital expenditures across major tech firms, indicating a competitive "arms race" in AI and cloud infrastructure [58][61] - Investors are advised to monitor financial performance indicators such as revenue growth, gross margins, and changes in CapEx guidance, as these will impact long-term cash flow and profitability [3][62]

Core Insights - Company reported a revenue growth of 11.9% year-on-year to 7.36 billion yuan for the fiscal year 2025, which is in line with expectations; adjusted EBITDA increased by 10.5% to 4.17 billion yuan, exceeding expectations due to better-than-expected cost control [1] - The number of full-time students increased by approximately 5% year-on-year to 282,000 as of August 31, 2025, but growth in student numbers for the 2025/26 academic year showed a slowdown with only a 0.2% increase [1] Financial Performance - The gross margin for the fiscal year 2025 was 53.3%, a decrease of 2.1 percentage points year-on-year; adjusted EBITDA margin was 56.6%, down 0.7 percentage points year-on-year [2] - The company recognized goodwill impairment totaling 1.706 billion yuan for three of its schools during the period [2] - Capital expenditures decreased by 45.2% year-on-year to 2.66 billion yuan, indicating a turning point in capital spending [2] - As of August 31, 2025, the company's interest-bearing debt ratio was approximately 26.0%, slightly down from 26.4% in the previous year, indicating a healthy level of debt [2] Future Outlook - The company has adjusted its revenue forecast for fiscal year 2026 down by 4% to 7.77 billion yuan and reduced the adjusted EBITDA forecast by 1% to 4.19 billion yuan; it has introduced a revenue forecast for fiscal year 2027 of 8.16 billion yuan and an adjusted EBITDA forecast of 4.42 billion yuan [2] - The target price has been lowered by 30% to 3.5 HKD, corresponding to 3.4 times the estimated adjusted EV/EBITDA for 2026 [2] - The company is currently trading at 2.8 times the estimated adjusted EV/EBITDA for 2026, indicating a potential upside of 22% [2]

Investment Rating - The report maintains a rating of "Outperform" for the computer sector [5] Core Insights - Alibaba is leveraging the Qwen large model foundation to comprehensively reshape its business and is accelerating the construction of B-end ecological barriers through an open-source strategy and strong performance [1][2] - The company is increasing capital expenditure to meet the surging demand for computing power, with cloud revenue continuing to grow significantly, validating the "infrastructure investment - technology iteration - commercial monetization" closed-loop logic [1][4] Summary by Sections Section 1: Alibaba's AI Layout - Alibaba has transitioned from a cloud-based service model to a comprehensive AI-driven approach, utilizing its self-developed "Flying" system to support internal and external business operations [10][11] - The company has established a solid technical foundation through its research institutions, including the Damo Academy and Tongyi Laboratory, to drive AI development [10][33] Section 2: Model Performance and Market Position - The Qwen series flagship model has rapidly iterated and is now among the top tier globally, with performance approaching that of leading overseas closed-source models [3][46] - Alibaba's open-source strategy has positioned it to capture market share in the domestic B-end market, building a strong user moat and ecological barrier [3][37] Section 3: Cloud Infrastructure and Revenue Growth - To address the explosive demand for AI computing power, Alibaba is increasing its infrastructure investments, achieving rapid revenue growth in AI-related products [4][76] - The company has developed self-research chips and a global data center network to ensure soft and hard resource coordination [4][76] Section 4: Investment Recommendations - The report suggests focusing on various players within the Alibaba ecosystem, particularly those involved in Pre-AI revenue generation and specific vertical AI applications [1] - Recommended companies include Fengwei Network, Kingdee International, and others in the AI coding and multimodal sectors [1]