微软价值 6250 亿美元的人工智能飞轮，不改长期价值

Core Viewpoint - Microsoft is not only experiencing continuous growth but is also achieving compound expansion through a substantial reserve of unfulfilled contracts (RPO), which secures future revenue. The integration of Azure, Foundry, and Copilot enhances customer stickiness and the company is actively investing in building a long-term AI infrastructure. Notably, despite a stock price correction, the long-term investment logic remains unchanged [1][9]. Financial Highlights - The balance of unfulfilled contracts (RPO) reached $625 billion, a year-on-year increase of 110%, with commercial bookings up 230%, indicating high visibility for future revenue [5]. - Approximately 25% of the $625 billion RPO is expected to convert to revenue within the next 12 months, with a year-on-year increase of 39% in short-term revenue conversion. The remaining long-term portion grew by 156%, securing long-term cash flow against short-term market fluctuations [5]. - In Q2 of FY2026, Microsoft's cloud business revenue reached $51.5 billion, a 26% year-on-year increase, with Azure and other cloud services growing by 39%, significantly outpacing the industry median revenue growth rate of 9.22% [2][5]. AI and Cloud Services - Microsoft 365 Copilot seat adoption increased by 160% quarter-on-quarter, with the number of large customers exceeding 35,000, tripling. Daily active users grew threefold, driving an increase in average revenue per user (ARPU) without needing to acquire new customers, with paid seats reaching 15 million [5]. - The integration of Foundry with models from OpenAI, Anthropic, Mistral, and Cohere has created a platform lock-in effect, with over 1,500 customers using both Anthropic and OpenAI models on Azure [5]. Azure Growth and Infrastructure - Azure's growth is driven by the scale of data processing, with the Fabric product achieving an annualized revenue of over $2 billion and over 31,000 customers. The product's aggregation effect reduces customer reluctance to migrate to the cloud, enhancing Azure's retention rate [6]. - The company is vertically integrating its self-developed chips, such as the Maia 200 AI accelerator and Cobalt 200 CPU, to control costs, particularly for OpenAI workloads, reducing reliance on third-party hardware [6]. Capital Expenditure and Market Dynamics - Microsoft is facing pressure from capital expenditures (CapEx), with Q2 CapEx reaching $37.5 billion, primarily for GPUs and CPUs, which constitute two-thirds of total expenditures. This has led to a significant negative impact on free cash flow [7]. - Supply chain issues have delayed hardware deployment, preventing Azure from meeting all AI computing demands, which could result in lost revenue opportunities and market share [7]. - The market is disappointed with the return curve on AI investments, as substantial short-term investments have not yet translated into immediate revenue boosts, contributing to a 10% drop in stock price following Q2 earnings [8]. Conclusion - In summary, Microsoft is not merely riding the AI wave but is establishing itself as a "toll road" in the AI era. The $625 billion in unfulfilled contracts, 39% growth in Azure, and explosive penetration of Copilot provide rare revenue visibility and compound monetization capabilities. Despite short-term pressures from capital expenditures, these investments are essential for building a long-term AI infrastructure moat, which does not alter the company's long-term value [9].