证券研究报告：传媒 | 公司点评报告 股票投资评级 资料来源：聚源，中邮证券研究所 公司基本情况 | 最新收盘价（元） | 41.90 | | --- | --- | | 总股本/流通股本（亿股）12.55 / 12.54 | | | 总市值/流通市值（亿元）526 / 526 | | | 52 周内最高/最低价 | 49.94 / 25.70 | | 资产负债率(%) | 17.9% | | 市盈率 | -32.23 | | 第一大股东 | 新余盈瑞世纪软件研发 | | 中心(有限合伙) | | 研究所 分析师:王晓萱 SAC 登记编号:S1340522080005 Email:wangxiaoxuan@cnpsec.com 昆仑万维(300418) 收入高增延续，关注 AI 算力芯片领域布局 ⚫ 事件回顾 2025 年 8 月 22 日，公司发布 2025 年半年报，2025 上半年，公 司实现营业收入 37.33 亿元，同比增长 49.23%；归母净利润-8.56 亿 元，同比减少 119.86%；扣非后归母净利润-8.59 亿元，同比减少 110.91%。2025Q2，公司实现营业收入 19.69 亿 ...

Industry Investment Rating - The investment rating for the computer industry is "Outperform the Market" and is maintained [1][8] Core Viewpoints - The report highlights that China Mobile plans to procure approximately 7,058 AI general computing devices (inference type) for 2025-2026, with a capital expenditure of 37.3 billion yuan for AI computing resources [4][5] - The procurement project by China Mobile has a total value exceeding 5 billion yuan, divided into six packages, with significant portions allocated to CUDA and CANN ecosystem devices [4][5] - Huawei's Ascend 910C is noted to have performance comparable to NVIDIA's H100, indicating that domestic computing power can meet the inference needs of large AI models [5] Summary by Relevant Sections Industry Basic Situation - The closing index for the computer industry is 5,804.0, with a 52-week high of 5,804.0 and a low of 2,805.53 [1] Recent Research Reports - The report mentions the acceleration of AI deployment by China Mobile, with a focus on general computing devices for inference [4] - The report provides details on the winning bidders for the procurement, with ZTE winning the largest share at 885 million yuan [5] Investment Recommendations - The report suggests focusing on the domestic computing power supply chain, highlighting companies associated with Huawei and other ecosystems [6]

证券研究报告：农林牧渔 | 公司点评报告 股票投资评级 资料来源：聚源，中邮证券研究所 公司基本情况 | 最新收盘价（元） | 22.89 | | --- | --- | | 总股本/流通股本（亿股）4.65 / 3.35 | | | 总市值/流通市值（亿元）106 / 77 | | | 52 周内最高/最低价 26.80 / 11.29 | | | 资产负债率(%) | 37.1% | | 市盈率 | 35.04 | | 第一大股东 | 梁武 | 研究所 分析师:王琦 SAC 登记编号:S1340522100001 Email:wangqi2022@cnpsec.com 瑞普生物(300119) 业绩高增，多业务协同发力 ⚫ 事件： 公司公布 2025 年中报，实现营业收入 17.08 亿元，同比增长 20.53%，归母净利润 2.57 亿，同比增长 57.59%。公司聚焦生物制品 主业，同时战略布局宠物新业态，实现了收入与利润的双增长。 ⚫ 点评：经济动物板块稳健，宠物板块三箭齐发 经济动物动保板块：公司聚焦大客户战略，与头部养殖集团建立 战略合作，凭借品牌、产品质量及研发实力等优势持续扩大市场份额。 分 ...

l 投资要点 25H1 营收稳健增长，数据中心及服务器收入规模进一步提升。 公司不断丰富产品矩阵，推出一系列满足数据中心及服务器应用需求 的高性能 PCB 产品，在数据中心及服务器领域的收入规模进一步提 升，服务器 CPU 主板、存储板、散热板等产品稳定批量交付，GPU 板 组也在部分客户中取得实质性进展。上半年实现营业收入 25.65 亿 元，yoy+19.43%，实现归母净利润 1.96 亿元，yoy-11.96%，综合毛 利率 22.34%，yoy-0.34pct；单季度来看，Q2 实现营收 14.01 亿元， yoy+19.65%，qoq+20.33%，归母净利润 0.84 亿元，yoy-24.61%，qoq- 25.56%，综合毛利率 20.91%，yoy-2.95pct，qoq-3.15pct。 持续加大研发投入，服务器/AIPC/汽车多点开花。在服务器领域， 公司围绕小间距 BGA 夹线、SI 数据库、背钻 stub 及对准度等关键技 术升级；针对下一代服务器OAK平台开发混压工艺体系，通过Pin-lam 设备提升对准度控制水平，部分订单控制到 4mil 的对准度偏差，并 展开 Pcle6.0/ ...

Investment Rating - The report maintains a "Buy" rating for the company, indicating an expected relative increase in stock price of over 20% compared to the benchmark index within the next six months [2][11]. Core Insights - The company reported a revenue of 797 million yuan for the first half of 2025, representing a year-on-year growth of 34.45%. The net profit attributable to shareholders reached 145 million yuan, up 29.08% year-on-year [4]. - The growth in performance is attributed to the gradual release of production capacity from fundraising projects, which improved product structure and sales volume. The main business segments, tantalum and niobium products, generated revenues of 784 million yuan, with a gross profit of 147 million yuan, reflecting a growth of 35.50% and 31.38% respectively [5]. - The company is focusing on high-end tantalum materials, with key products like high-purity tantalum targets and powders driving revenue growth. The ultra-high purity tantalum target with a purity of 5N9 meets the demands of advanced chip manufacturing [6]. Financial Projections - The company is expected to achieve revenues of 1.969 billion yuan, 2.522 billion yuan, and 2.822 billion yuan for the years 2025, 2026, and 2027 respectively, with year-on-year growth rates of 53.72%, 28.12%, and 11.88% [6][9]. - The net profit attributable to shareholders is projected to be 291 million yuan, 371 million yuan, and 417 million yuan for the same years, with growth rates of 36.57%, 27.38%, and 12.36% respectively [6][9]. - The price-to-earnings (P/E) ratios for 2025, 2026, and 2027 are estimated to be 36.43, 28.60, and 25.46 respectively [6][9].

Investment Rating - The report upgrades the investment rating of the company to "Buy" based on expected revenue and performance growth [9]. Core Views - The company reported a robust revenue growth of 24.0% year-on-year for the first half of 2025, achieving a revenue of 213 million yuan and a net profit of 85 million yuan, which is a 15.1% increase year-on-year [5]. - The growth is primarily driven by the ramp-up of the intelligent connected vehicle testing facility in the Yangtze River Delta, which began operations in July 2024, leading to increased demand from automotive companies for testing and certification [6]. - The company plans to acquire 100% equity of a winter testing facility in Hulunbuir for 111 million yuan, enhancing its testing capabilities across all seasons and potentially boosting long-term performance [8]. Summary by Sections Company Overview - Latest closing price: 6.59 yuan - Total shares: 1.324 billion, Market capitalization: 8.7 billion yuan - 52-week high/low: 7.06/4.92 yuan - Debt-to-asset ratio: 16.5%, P/E ratio: 50.69 [4]. Financial Performance - In Q2 2025, the company achieved a revenue of 125 million yuan, a 20.1% increase year-on-year, and a net profit of 56 million yuan, up 11.0% year-on-year [5]. - The gross margin for the first half of 2025 was 66.7%, reflecting a decrease due to increased depreciation costs associated with the new testing facilities [7]. Revenue and Profit Forecast - Revenue projections for 2025-2027 are 622 million, 737 million, and 841 million yuan respectively, with net profits expected to be 220 million, 273 million, and 335 million yuan [9][11]. - The expected EPS for the same period is 0.17, 0.21, and 0.25 yuan per share, with corresponding P/E ratios of 39.7, 32.0, and 26.1 [9][11].

Report Overview - The report is a fixed - income research report released on August 26, 2025, aiming to analyze the bond market adjustment and provide investment suggestions through historical review, institutional cost assessment, and market sentiment analysis [1][10] Industry Investment Rating - No industry investment rating is provided in the report Core Viewpoints - During the equity bull market from 2014 - 2025, the performance of the bond market was not unilaterally opposed to the stock market but depended on the dynamic balance of risk preference and capital flow [3] - The current bond market correction is more like a "topping - out period" rather than a "peaking period". The upward space and time of interest rates are constrained by factors such as the return of allocation, policy soft constraints, and marginal improvement in supply and demand [4] - It is recommended to adhere to the "bottom - line thinking", focusing on the upper - bound constraint of interest rates and entry opportunities. If the risk preference is extremely priced and the 10Y - 1Y spread reaches 50 - 60BP, the corresponding 10 - year Treasury bond yield of 1.85% - 1.95% is the bottom - line range [4] Summary by Directory 1. 2014 - 2025, Review of Bond Market Trends in Previous Stock Bull Markets 1.1 2014–2015: Bull and Divergence of Stocks and Bonds Driven by Loose Pattern and Expectation Divergence - In the early stage (June - November 2014), with the promotion of reform expectations and loose signals, the stock and bond markets showed a short - term "double - bull" pattern [11] - From November to December 2014, after the central bank's interest rate cut and price mechanism reform, the stock market accelerated, while the bond market showed "profit - taking", presenting a "stock - up, bond - down" situation [11] - From December 2014 to February 2015, during the stock market consolidation period, the bond market had a repair opportunity, and the 10 - year Treasury bond yield dropped by more than 45bp [11] - From March to April 2015, with the reduction of reserve requirements and interest rates and the expansion of leveraged funds, the stock market accelerated, and the bond market showed an "N - shaped" shock [11] - From May to June 2015, with the influx of leveraged funds into the stock market, the bond market was under pressure, showing a typical "strong - stock, weak - bond" situation [11] 1.2 2016 - 2017: "First Bull, Then Bear" in the Bond Market under the Background of Supply - side Reform and De - leveraging - From June to August 2016, due to the supply - side reform and loose monetary policy, the stock and bond markets both rose [14] - From September 2016 to February 2017, the stock market continued to rise, while the bond market was under pressure due to the expectation of economic stabilization and inflation recovery, presenting a "stock - up, bond - down" situation [14] - From March to May 2017, due to domestic de - leveraging, tightened monetary policy, and external shocks, the stock and bond markets both declined [14] - From the second half of 2017 to the end of 2017, the stock market was strong, and the bond market was weak, with the 10 - year Treasury bond yield approaching 4.0% [14][16] 1.3 2019–2021: Deduction of the Stock - Bond Seesaw and Structural Bull Market - In 2019, during the GEM bull market, the stock market was strong, and the bond market was stable with a narrow - range fluctuation of the 10 - year Treasury bond yield around 3.1% - 3.2% [17] - In 2020, affected by the epidemic, the bond market first entered a bull market, and then the stock market became strong again after the economic recovery, showing a seesaw effect [19] - In 2021, with the weakening of growth momentum, the bond market returned to a bull market, and the stock market still had structural opportunities, showing a phased resonance [19] 2. In the Assumption of an Equity Bull Market, What Stage is the Current Bond Market Correction in? 2.1 In This Round of Bond Market Correction, the Interest Rates of Some Varieties are Close to the Holding - Cost Lines of Product Accounts - For wealth management products, as of the week of August 24, the 1 - year cost yield of inter - bank certificates of deposit was 3.28BP higher than the average interest rate of certificates of deposit, and the 1 - year cost yield of 0 - 1Y policy - financial bonds was 1.86BP lower than the 1 - year CDB bond yield [22] - For funds, as of the week of August 24, the 1.5 - year cost yield of 7 - 10Y policy - financial bonds was 16.14BP higher than the 10 - year CDB bond average, and the 1.5 - year cost yield of 10Y+ Treasury bonds was 23.36BP lower than the 30 - year Treasury bond yield [22] - In terms of institutional trading behavior, insurance institutions increased their net positions in ultra - long - term bonds, rural financial institutions adjusted their positions, and wealth management products shifted from the interest - rate style to the credit style [23] 2.2 Analyzing the Market's Deduction Space from Micro - sentiment Indicators - The stock - bond seesaw effect is still significant, but the upper bound of long - term yields may be gradually clear. The stock market is hot, while the bond market sentiment is controllable [30] - The stock - bond yield spread shows that the cost - performance of bond assets has increased, attracting the return of some allocation - type institutions [30] - The scale of wealth management products is under pressure, with an increase weaker than the seasonal level, but it remains relatively stable. There is no large - scale redemption of fixed - income funds [31] 2.3 Bond Market Outlook: Adjustment May Have Intervals. Pay Attention to the "Topping - out - Returning" Rhythm with Bottom - line Thinking - The current bond market correction is relatively moderate, and the upward space and time of interest rates are constrained. Wealth management products and bond funds still have safety cushions and profit margins [34] - The central bank has increased liquidity injection. If interest rates over - adjust, the probability of the central bank's intervention will increase [34] - The supply - demand relationship may improve marginally. The peak of bond issuance has passed, and the return of allocation demand will help balance supply and demand [35] - The economic growth and inflation are in a moderate range, and the bond market pricing will return to the center determined by the fundamentals and policy interest rates. It is recommended to use bottom - line thinking for long - term interest - rate bond allocation [36]

Investment Rating - The investment rating for the company is "Buy" and is maintained [1] Core Views - The company reported a revenue of 2.473 billion yuan for the first half of 2025, representing a year-on-year increase of 12.50%. However, the net profit attributable to shareholders decreased by 9.93% to 796 million yuan [4] - The decline in performance is primarily attributed to the suspension of operations at the Yinman Mine and losses at the Yubang Mine, along with reduced profits from Rongguan and Qianjinda mines [4][5] - The company expects a recovery in operations in the second half of the year, supported by a significant increase in silver and tin prices in Q3 [6] Company Overview - The latest closing price is 42.38 yuan, with a total market capitalization of 66.5 billion yuan [3] - The company has a total share capital of 1.57 billion shares, with an asset-liability ratio of 8.2% and a price-to-earnings ratio of 25.84 [3] Financial Performance - In the first half of 2025, the company produced 131.32 tons of silver (up 4.57% year-on-year) and 3589.82 tons of tin (down 20.64% year-on-year) [6] - The average prices for silver and tin increased by 20.31% and 9.23% year-on-year, respectively [6] - The company forecasts revenues of 5.092 billion yuan, 6.184 billion yuan, and 7.140 billion yuan for 2025, 2026, and 2027, respectively, with corresponding net profits of 1.950 billion yuan, 2.303 billion yuan, and 2.782 billion yuan [7][8] Earnings Forecast - The expected earnings per share (EPS) for 2025, 2026, and 2027 are 1.10 yuan, 1.30 yuan, and 1.57 yuan, respectively [7] - The projected price-to-earnings (P/E) ratios for the same years are 18.72, 15.85, and 13.12 [7]

- DeepSeek-V3.1 model is an upgraded version of the DeepSeek language model, featuring a hybrid inference architecture that supports both "thinking mode" and "non-thinking mode" for different task complexities[12][13][14] - The model's construction involves dynamic activation of different attention heads and the use of chain-of-thought compression training to reduce redundant token output during inference[13] - The context window length has been expanded from 64K to 128K, allowing the model to handle longer documents and complex dialogues[15] - The model's performance in various benchmarks shows significant improvements, such as a 71.2 score in xbench-DeepSearch and 93.4 in SimpleQA[17] - The model's evaluation highlights its advancements in hybrid inference, long-context processing, and tool usage, although it still faces challenges in complex reasoning tasks[21] - Seed-OSS model by ByteDance features 36 billion parameters and a native 512K long-context window, emphasizing research friendliness and commercial practicality[22][23] - The model uses a dense architecture with 64 layers and integrates grouped-query attention (GQA) and rotary position encoding (RoPE) to balance computational efficiency and inference accuracy[23] - The "thinking budget" mechanism allows dynamic control of inference depth, achieving high scores in various benchmarks like 91.7% accuracy in AIME24 math competition[24] - The model's evaluation notes its strong performance in long-context and reasoning tasks, though its large parameter size poses challenges for edge device deployment[25] - WebWatcher by Alibaba is a multimodal research agent capable of synchronously parsing image and text information and autonomously using various toolchains for multi-step tasks[26][27] - The model's construction involves a four-stage training framework, including data synthesis and reinforcement learning to optimize long-term reasoning capabilities[27] - WebWatcher excels in benchmarks like BrowseComp-VL and MMSearch, achieving scores of 13.6% and 55.3% respectively, surpassing top closed-source models like GPT-4o[28] - The model's evaluation highlights its breakthrough in multimodal AI research, enabling complex task handling and pushing the boundaries of open-source AI capabilities[29] - AutoGLM 2.0 by Zhipu AI is the first mobile general-purpose agent, utilizing a cloud-based architecture to decouple task execution from local device capabilities[32][33] - The model employs GLM-4.5 and GLM-4.5V for task planning and visual execution, using an asynchronous reinforcement learning framework for end-to-end task completion[34] - AutoGLM 2.0 demonstrates high efficiency in various tasks, such as achieving a 75.8% success rate in AndroidWorld and 87.7% in WebVoyager[35] - The model's evaluation notes its significant advancements in mobile agent technology, though it still requires optimization for cross-application stability and scenario generalization[37] - WeChat-YATT by Tencent is a large model training library designed to address scalability and efficiency bottlenecks in multimodal and reinforcement learning tasks[39][40] - The library introduces parallel controller mechanisms and partial colocation strategies to enhance system scalability and resource utilization[40][42] - WeChat-YATT shows a 60% reduction in overall training time compared to the VeRL framework, with each training stage being over 50% faster[45] - The model's evaluation highlights its effectiveness in large-scale RLHF tasks and its potential to drive innovation in multimodal and reinforcement learning fields[46] - Qwen-Image-Edit by Alibaba's Tongyi Qianwen team is an image editing model that integrates dual encoding mechanisms and multimodal diffusion Transformer architecture for semantic and appearance editing[47][48] - The model's construction involves dual-path input design and chain editing mechanisms to maintain high visual fidelity and iterative interaction capabilities[48][49] - Qwen-Image-Edit achieves SOTA scores in multiple benchmarks, with comprehensive scores of 7.56 and 7.52 in English and Chinese scenarios respectively[50] - The model's evaluation notes its transformative impact on design workflows, enabling automated handling of rule-based editing tasks and lowering the barrier for visual creation[52] Model Backtest Results - DeepSeek-V3.1: Browsecomp 30.0, Browsecomp_zh 49.2, HLE 29.8, xbench-DeepSearch 71.2, Frames 83.7, SimpleQA 93.4, Seal0 42.6[17] - Seed-OSS: AIME24 math competition 91.7%, LiveCodeBench v6 67.4, RULER (128K) 94.6, MATH task 81.7[24] - WebWatcher: BrowseComp-VL 13.6%, MMSearch 55.3%, Humanity's Last Exam-VL 13.6%[28] - AutoGLM 2.0: AndroidWorld 75.8%, WebVoyager 87.7%[35] - Qwen-Image-Edit: English scenario 7.56, Chinese scenario 7.52[50]

证券研究报告：宏观报告 发布时间：2025-08-26 分析师:李起 SAC 登记编号:S1340524110001 Email:liqi2@cnpsec.com 研究助理:高晓洁 SAC 登记编号:S1340124020001 Email:gaoxiaojie@cnpsec.com 近期研究报告 《出口韧性、"反内卷"，引领 PPI 边 际改善》 - 2025.08.25 宏观观点 海外宏观周报：美联储降息预期升温 ⚫ 核心观点： 杰克逊霍尔全球央行年会上鲍威尔意外放鸽，表示"风险平衡正 在发生转变"，强调就业风险。鲍威尔指出，虽然劳动力市场似乎处 于平衡状态，但这一平衡是供给和需求同时放缓的结果，这种不寻常 的情况表明，就业的下行风险正在上升。通胀方面，鲍威尔认为关税 对通胀的影响相对短暂，仅仅是价格水平的一次性变化。此外，鲍威 尔还宣布完成了联储第二次公开货币政策框架评估，废除 2020 年以 来的平均通胀目标制，恢复到 2%通胀目标的框架，但仍保留就业方面 的目标，将减轻与最大就业率的"短缺"而不是"偏差"，即更重视 就业的下行风险。在鲍威尔明确传递降息预期后，交易员将 9 月降息 概率推升至约 90 ...