Azure Crossed $100 Billion And I Am Still Not Trimming Microsoft

I have owned Microsoft for years and I get asked the same question every time it makes a new high: is this the one you finally sell? No. Let me show you the two numbers from this quarter that decided it for me.

The number that matters

Azure crossed $100 billion in annual revenue for the first time, and it still grew 43% year over year. Total company revenue was $90 billion, up 18%.

Sit with that combination for a second. Growth rates are supposed to decay as the base gets bigger. A business the size of a Fortune 100 company compounding in the low 40s is not supposed to exist. That is the entire thesis in one line, and it is why the stock ran roughly 8% on the print and another 5% on Monday.

What Copilot actually proved

Microsoft 365 Copilot passed 30 million paid seats, with net additions more than doubling from the prior quarter.

The accelerating net adds are the real signal. Every large software company on earth announced an AI product in the last two years. Almost none of them can show you a paid seat count that is inflecting rather than fading after the pilot phase. Microsoft can attach AI to a subscription that finance departments already approved, which is a distribution advantage no startup can buy.

Citi moved its target from $570 to $600 on the back of it.

What I am actually paying for

Here is the honest part. I am not buying Microsoft because it is cheap, because it is not. I am paying a premium for the highest quality recurring revenue base in software, attached to the fastest growing large cloud, run by a management team that has now been right about AI for three straight years.

The risk I take seriously is not competition. It is capital intensity. The whole hyperscaler group is spending enormous sums to build capacity, and the market has already shown this earnings season that it will punish capex without visible revenue attached. Meta traded on that exact question last week.

Microsoft is currently on the right side of that trade because Azure and Copilot are the revenue proof. If that ever stops, the multiple goes with it.

Position

No change. This is a core holding and I am not trimming a compounder for being expensive, because that is the mistake that cost me the most money in my first decade of investing. I add on genuine multiple compression, not on price highs.

Not investment advice.