Quick follow-up for anyone who saw my earlier [posts on Microsoft and its AI spending](https://www.reddit.com/r/ValueInvesting/comments/1uio0by/six_months_ago_we_called_microsoft_overvalued_at/). They reported Q4 FY26 tonight, the stock jumped 8%, and it's worth a look at what the market rewarded.
The quarter was genuinely strong. Revenue of $90.0B (up 18%), non-GAAP EPS of $4.74 (up 23%), operating income of $40.6B. Azure grew 43% and crossed $100B in annual revenue for the first time, with Copilot past 30 million paid seats. This is still a solid and growing business.
But capex didn't fall - it's rising. FY26 cash capex hit $115.9B, 35% of revenue (up from 23%), and FY27 is guided higher still ($255-260B, Q1 alone above $50B). The reported number only looks milder ($175B for 2026, down from $190B) because Microsoft stretched data-centre depreciation from 15 to 25 years and shifted some leases off the capex line. So, Microsoft is spending more but reporting less.
However, what the market liked is real: Azure's acceleration shows the capex is earning a return, and management guided to stay cash flow positive through FY27. This contrasts with Google, which has already turned cash flow negative.
So, Microsoft finally gave the market proof that its AI spending is paying off, but the spending itself is still accelerating, and the restraint is largely an accounting change, not a real slowdown in AI spending.
Not investment advice. For educational purposes only. The author and Stockoscope may hold positions in the securities mentioned, so always do your own research.