Microsoft CorpMSFTDON'T BUYNov 06, 2025Stock price when the opinion was issued
As of Jul 24, 2026. Market Open.
Software and spending concerns are problems. What will OpenAI do for MSFT? But if quantum computing happens, MSFT is one of the three hyperscalers who will take advantage. Google and MSFT have produced chips for quantum computing; MSFT's chip is self-cooling, which extends their lifetime. MSFT is volatile, but he wants to add to his position.
Buy it, put it away for 10 years, do very well. Suffering from software outflows from large ETFs. Revenue still growing in teens, even better on EPS. Investing a lot in capacity (but they can afford it), which customers have signed onto for the next several years. Trades at 20x PE, yet nothing's really changed.
Three durable growth engines: Azure, enterprise software, AI monetization. Key is that it keeps turning its installed base into higher-value subscriptions and usage-based revenue, while preserving margins and cash generation. Market's concerned that margins and cash will be pressured as Gen AI gets rolled out through competitors.
Azure remains the clearest growth driver. Key competitive advantage with enterprise software is that one stack bundles infrastructure, security, identity, and data/productivity tools. Raises costs to switch, which provides pricing power. Yield is 0.93%.
The lower MSFT gets, the more he likes it. The valuation keeps falling. He recently bought a position and would add to it now. If it holds, that's a very good technical signal. He loves MSFT, but consider that France will forbid the government using Microsoft Teams. That said, MSFT isn't going anywhere.
Down, but not that much. Expensive stock, so this name would be front-and-centre in a correction. Still, Azure growth was 40% last quarter. Continues to deliver. You can see why they're winning all this cloud business (his office switched to MSFT servers, and now all of a sudden every problem is MSFT's, not John's anymore).
Competitive market. Unknown what ORCL will do as the fourth player. Bigger question for the hyperscalers is how to monetize AI. Owns other names that are cheaper.