2026.03. 20 作者 | 第一财经 李娜 过去一年，围绕大模型与Agent的进展，软件行业经历了一轮罕见的情绪震荡。从Anthropic推出 Claude，到资本市场对SaaS板块的重新定价，"AI会不会杀死软件"成为一个被反复追问的问题。 在刚刚结束的NVIDIA GTC大会上，黄仁勋公开反对"AI会掏空软件"的判断。他表示，Agent必须建立 在企业系统与结构化数据之上。 而在更多的一线软件从业者看来，在AI尤其是Agent体系下，软件的价值开始向更底层转移，但AI并没 有摧毁软件，而是在重塑其呈现方式。但与此同时，他们也承认，AI正在给软件行业带来焦虑。 "在多数企业里，AI代码建议的实际采纳通常处于中下等水平，更多发生在局部代码段而非完全重写， 但在增删改查、脚手架、单测样板等任务上替代效应最明显。"Forrester副总裁兼首席分析师戴鲲日前 在接受第一财经记者采访时表示，而在复杂架构权衡、需求不清与遗留系统排障等关键环节，AI仍然难 以突破。 当软件不再被"打开" 软件行业的焦虑，很大程度上来自一个直观变化：用户正在减少"打开软件"的次数。 本文字数：2725，阅读时长大约4分钟 在传统模 ...

过去二十年，软件行业几乎是资本市场最确定的增长叙事。 这不仅仅是一个关于财务杠杆的问题，更是一个关于商业模式底层逻辑是否依然成立的深刻拷 问。 从 " 资 本 宠 儿 " 到 " 债 务 堰 塞 湖 " ： 软 件 行 业 的 隐 藏 杠 杆 "软件吞噬世界"的逻辑，让 SaaS 企业成为轻资产、高利润、高估值的代名词。投资者曾经坚 信，一旦企业安装了某种软件，极高的迁移成本将锁定未来十年的现金流。这种确定性，使得 软件公司在一级市场和二级市场都享受着近乎无限的宽容。 但在 AI 革命与高利率周期叠加的背景下，这个曾经最具确定性的行业，正悄然进入一场压力 测试。 在过去十多年里，软件行业一直是资本市场最受欢迎的赛道之一。 云计算与 SaaS 商业模式的兴起，彻底改变了企业的成本结构。软件公司拥有极高的资本效 率：不需要像制造业那样投入巨额资金建设厂房，也不需要像零售业那样维护庞大的库存。它 们只需要编写代码，然后通过订阅模式获得持续、可预测的现金流。 【如需和我们交流可扫码添加进社群】 宏观环境的剧变正在剥离软件行业的光环。当巨额债务集中到期与技术范式转移同时出现，一 个过去很少被讨论的问题开始浮出水面： 如 ...

Is ORCL a good stock to buy? We came across a bullish thesis on Oracle Corporation on Compounding Your Wealth’s Substack by Sergey. In this article, we will summarize the bulls’ thesis on ORCL. Oracle Corporation's share was trading at $159.16 as of March 12th. ORCL’s trailing and forward P/E were 30.66 and 20.41 respectively according to Yahoo Finance. Source:Pixabay Oracle Corporation offers products and services that address enterprise information technology environments worldwide. ORCL is demonstrati ...

Core Viewpoint - Columbus has corrected its revenue growth outlook for 2026 to a range of 0-5%, reflecting a cautious market environment and adjustments in expectations for financial performance [1][12]. Financial Performance - Columbus reported a revenue of DKK 1,576 million in 2025, representing a decline of 5% compared to 2024 [6][7]. - EBITDA decreased by 26% to DKK 113 million, resulting in an EBITDA margin of 7.2% [6][7]. - Profit before tax was DKK 47 million, down 19% from the previous year [6]. - Cash flow from operating activities fell by 43% to DKK 77 million [6]. Service Revenue Breakdown - The service revenue split by business lines showed a decline in Dynamics 365 by 8% to DKK 899,147 thousand and Digital Commerce by 4% to DKK 173,384 thousand [5]. - The Data & AI segment saw a growth of 3% to DKK 90,992 thousand, while total service revenue decreased by 5% to DKK 1,506,353 thousand [5][6]. Market Performance by Region - Revenue performance varied by region, with Sweden and Denmark experiencing declines of 5% and 11% respectively, while the US market grew by 18% [7]. Strategic Outlook - The company aims to return to growth in 2026, with expectations of improved earnings driven by enhanced efficiency and a focus on contract profitability [8]. - The management emphasizes a disciplined approach to execution quality and strategic focus to build a resilient organization capable of delivering long-term value [3].

Core Insights - Columbus experienced a revenue decline of 5% in 2025, amounting to DKK 1,576 million, with EBITDA decreasing by 26% to DKK 113 million, resulting in an EBITDA margin of 7.2% [1][6]. Financial Performance - The company entered 2025 with a solid order book but adjusted revenue expectations to align with 2024 levels due to a cautious investment environment, characterized by longer customer decision-making cycles and delays in larger projects [2]. - The overall financial performance is considered resilient, indicating no fundamental weakening in execution capabilities or customer relationships [2]. - Profit before tax decreased by 19% to DKK 47 million, and cash flow from operating activities fell by 43% to DKK 77 million [6]. Service Revenue Breakdown - Service revenue by business lines showed a decline in Dynamics 365 (-8%), Digital Commerce (-4%), and Other Local Business (-4%), while Data & AI saw a growth of 3% [5]. - Total sales of services decreased by 5% to DKK 1,506 million, while total sales of products increased by 5% to DKK 70 million [5][7]. Market Unit Performance - Revenue performance varied across market units, with Denmark experiencing an 11% decline, while the US saw an 18% increase [7]. Strategic Outlook - The company aims to return to growth in 2026, focusing on improving earnings through enhanced efficiency and contract profitability [8]. - The management emphasizes a disciplined approach to execution quality and strategic focus to build a resilient organization capable of delivering long-term value [3].

Company Overview - Oracle is a global leader in enterprise information technology with a market capitalization of approximately $417.8 billion, known for its Oracle Database and autonomous systems [3] - The company focuses on scalable, secure solutions that support data-driven operations and long-term digital transformation [2] Recent Performance - Oracle's stock peaked at $345.72 on September 10, 2023, after a quarterly report that saw shares soar nearly 36% in a single session [1] - However, since its peak, the stock has declined by 58%, with a 10.8% drop over the past 52 weeks and a 26.36% decline in the last three months [7] Financial Highlights - For fiscal Q2 2026, Oracle reported revenue of $16.06 billion, a 14% year-over-year increase, with cloud revenue climbing 34% to $8 billion [12] - Non-GAAP EPS increased by 54% annually to $2.26, exceeding expectations, while remaining performance obligations (RPO) surged 438% year-over-year to $523 billion [13] - Capital expenditures for Q2 reached approximately $12 billion, contributing to a negative free cash flow of $10 billion for the quarter [14] Debt and Investment - Oracle raised $18 billion in new debt to fund data center construction, pushing total debt above $100 billion [8] - The company is part of the $500 billion Stargate AI project alongside OpenAI and SoftBank, indicating significant investment in infrastructure [8] Future Projections - Management expects fiscal 2026 capital expenditures to reach approximately $50 billion, $15 billion above previous estimates [15] - Cloud revenue growth is projected between 37% and 41% for Q3, with total revenue expected to rise by 16% to 18% [16] - Analysts forecast a 36.6% year-over-year increase in fiscal 2026 EPS to $6.01, followed by a 4.8% rise to $6.30 in fiscal 2027 [16] Analyst Sentiment - Analysts have upgraded Oracle's rating to "Strong Buy," with 31 out of 42 analysts recommending this rating [17] - The consensus price target of $284.02 implies a 91.2% upside potential, with the highest target suggesting a possible 169% increase [18]

Core Insights - The current software sector pullback is driven by a debate over long-term value and whether AI will erode existing profit pools and competitive advantages [1][2] - Goldman Sachs analysts have identified seven bearish arguments regarding software companies, assessing their risks and potential impacts on various segments [1][2] Group 1: Market Concerns - The focus has shifted from short-term growth to concerns about whether AI will diminish software companies' competitive moats [2] - The report categorizes bearish arguments into a structured analysis, assigning risk scores to each argument to evaluate what can sustain long-term value [2] Group 2: System of Record (SoR) Risks - The risk of SoR being replaced is considered low (risk score 1), as generative AI is more suited for analysis rather than transactional processes [3] - However, there is a potential risk of value migrating from SoR to an "agentic operating system/orchestration layer" (risk score 4), which could weaken traditional competitive advantages [5] Group 3: Data Boundaries and Value Migration - If companies keep their data advantages confined within existing applications, the stability of SoR will be maintained, but profit pools may be siphoned off by new layers [4] - The orchestration layer could become more valuable as it enables cross-system reasoning and workflow automation, potentially undermining the traditional user interface and process dependencies of SoR [5] Group 4: Vertical vs. Horizontal Software - Vertical software is currently more resilient but may face challenges from horizontal platforms that allow users to create industry workflows using AI tools (risk score 2) [6] - The report highlights that established vertical software companies have significant barriers to entry due to proprietary data and deep integration into workflows [6] Group 5: Development Costs and Competition - The decline in coding costs due to AI tools will lead to increased competition, but the risk is rated as moderate (risk score 2) since software engineering involves more than just coding [8] - Efficiency gains from AI tools may shift bottlenecks to new areas, particularly in enterprise-level delivery where security and integration remain critical [8] Group 6: Customization Trends - Companies may increasingly prefer to build custom solutions, particularly in scenarios where existing software does not meet their needs (risk score 3) [9] - Palantir is cited as an example of a company successfully leveraging customization to create quantifiable ROI for clients [9] Group 7: Profit Margin Pressures - The industry is expected to experience moderate margin pressures over the next 12-24 months as companies absorb costs related to AI adoption [12] - The shift towards consumption-based pricing models may alter traditional SaaS economics, with some AI-native companies reporting lower margins compared to established SaaS firms [12] Group 8: Technological Uncertainty - The rapid pace of technological advancement presents the highest risk, making it difficult to predict long-term outcomes (risk score 5) [13] - The report notes that the unpredictability of technology evolution can lead to lower valuation multiples due to increased uncertainty [14] Group 9: Stability Signals - Key signals to watch for stability include whether software companies can demonstrate that domain expertise leads to higher quality outcomes and whether financial fundamentals can stabilize or improve [15]

Core Viewpoint - The human software sector is currently trading at 10-year lows in terms of multiples, indicating a significant market selloff that has broadly affected many companies, suggesting an exaggerated reduction in expected growth rates [1][5][12] Valuation Metrics - The sector is trading at an average of slightly above 10 times EBITDA, a stark contrast to the historical average of about 25 times, indicating strong support levels for depressed valuations [5] - The one-year forward PE ratio for the sector is currently at 17 times, projected to decrease to 14 times in 2027, aligning more closely with the broader equity market despite the sector's faster growth [5] Growth Expectations - Companies in Europe are expected to achieve around 10% revenue growth, but the market is pricing in a much lower growth rate for the future [6][12] - The anticipated growth for many companies in Europe remains healthy, with no downgrades expected through 2026, although acceleration in revenue growth is not evident [12] Company-Specific Insights - Companies with strong customer bases and data modes, such as SAP, are viewed as more insulated from market risks, despite not holding shares in these companies [3][9] - The integration level of software solutions, particularly ERP systems, makes it challenging for companies to switch providers, providing a competitive advantage to established players [7][9] Market Dynamics - The current market environment shows a lack of differentiation among software companies, creating potential investment opportunities in firms with low churn and high data modes [2][8] - The divergence in performance between hardware and software sectors is notable, with hardware companies like ASML reporting record earnings while software firms like SAP face investor disappointment due to slower growth in their cloud businesses [10][11]

Core Insights - The industrial software industry is at a critical juncture, driven by the need for innovation and the urgency of development, particularly in the context of China's economic transformation and the push for self-sufficiency in core technologies [1][4][17] - The market for industrial software in China is projected to approach 300 billion by 2024, indicating robust growth despite challenges such as a hollowing out of core technologies and imbalanced industrial structures [1][17] - The evolution of industrial software is characterized by a shift from tools to systems, platforms, and eventually to a genetic level, focusing on data value and efficiency [2][48] Industry Dynamics - The industrial software market is large, with significant opportunities for companies to target head, mid, and long-tail customers, each with distinct needs and potential for revenue generation [2][50] - The core evolution path of industrial software is from tools to systems, then to platforms, and finally to genetic integration, emphasizing the importance of data flow and value efficiency [48][49] - The industry faces systemic challenges, including a lack of foundational technologies and difficulties in integrating into supply chains, which hinder the development of domestic industrial software [26][17] Product Development Trends - Current industrial software primarily focuses on product sales, but there is a shift towards selling "intelligence" as data assets are accumulated and utilized effectively [3][52] - The integration of AI and large models is expected to enhance the capabilities of industrial software, particularly in areas such as code generation and human-computer interaction [43][52] - Future products are anticipated to evolve into "digital engineers," capable of autonomous task execution and intelligent interaction, moving beyond traditional software tools [52] Market Characteristics - The industrial software market is characterized by a high degree of fragmentation, with varying levels of domestic replacement and integration needs across different customer segments [14][50] - The demand for industrial software is driven by practical applications in enterprises, government initiatives, and the integration of research institutions, each with unique procurement focuses [14][16] - The market is currently experiencing a transition from subsidy-driven growth to a more market-oriented approach, emphasizing the importance of innovation and self-sufficiency [19][12] Challenges and Opportunities - The industry is grappling with significant challenges, including a lack of core technologies in research and design software, which is critical for engineering optimization [23][17] - Companies are encouraged to leverage policy incentives and market opportunities to enhance their technological capabilities and address the "bottleneck" issues in core components [17][26] - The evolution of industrial software is expected to create new revenue streams through data value services, as companies adapt to the changing landscape of technology and market demands [30][52]

Core Viewpoint - The article critiques the current state of enterprise AI agents, highlighting their perceived ineffectiveness compared to general AI agents, which continue to attract attention and investment despite their limitations in practical business applications [1][3][17]. Group 1: Perception of AI Agents - Enterprise AI agents are viewed as underappreciated by management, who often favor general AI agents due to their flashy capabilities and market appeal [4][5]. - Employees express frustration with general AI agents, which fail to address specific operational issues, while enterprise AI agents integrate seamlessly into existing workflows and automate repetitive tasks [5][7]. Group 2: Limitations of Enterprise AI Agents - The article identifies three main shortcomings of enterprise AI agents: a sense of unrecognized talent, a disconnect between expectations and reality, and poor cost-effectiveness [8][12]. - Enterprise AI agents struggle to demonstrate their value compared to general AI agents, which are perceived as more capable of delivering tangible results and creative solutions [7][9]. Group 3: Cost and Value Concerns - The development and integration of enterprise AI agents require significant investment in terms of time and resources, leading to concerns about their return on investment [12][15]. - Unlike general AI agents, which can quickly adapt to various business scenarios, enterprise AI agents often fail to provide sufficient value, making them less appealing to cost-conscious businesses [15][17].