Tag for finding my other articles on this stock: MSFT
About this blog: I have been investing for 25 years, professionally and personally. I look for stocks that have a high probability of compounding at 15% for at least 5 years with limited downside. I write these stocks up on my blog. You can find more about me, my philosophy, my mental models, and my portfolio structure on my site.
I have had great feedback on my 3-part Microsoft review, published in May. But it was too long; I have now edited and tightened it somewhat.
In addition, I felt a summary and update would be useful. These are below, along with links to the edited originals.
Thesis summary
Part 1 described the layers of the AI stack (silicon, infrastructure, LLMs, agents, and distribution) and argued that:
In silicon, Microsoft is behind but catching up with Google and Amazon.
Infrastructure is somewhat likely to commoditise eventually, especially behind the frontier, creating a new utility compute industry and allowing hyperscalers to return to more capital-light ways.
Part 2 argued that:
In LLMs there is a rolling wave of commoditisation, with frontier models never more than a year or two ahead of cheaper models. As more and more use cases can be addressed using cheap models, Microsoft will be the distributor of increasingly powerful commoditised AI to enterprise.
In agents Microsoft has a potential goldmine, with a deep cloud stack allowing customers to develop, orchestrate, monitor, and control agents, plus extremely valuable horizontal and vertical proprietary agents. TAM is every process in every enterprise.
Part 3 argued that:
Distribution is Microsoft’s deepest moat. Microsoft will bundle increasingly powerful intelligence into existing products, wrapping it in trust and pricing it cheaply. It will be hard for IT departments not to buy it.
Combining layers is powerful but Google’s full stack is not an insurmountable advantage.
Thesis update
AI is a remarkably fast-moving space and a lot has happened since I published my review in May. Luckily I think just about all of it is thesis-confirming and Microsoft stock is up over 20%.
AI is becoming much more useful and the debate over ROI on capex seems to be to be quieting down.
More frontier models are appearing at the front of the race - Meta and Grok both seem to have made progress relative to the frontier. More competition is good for distributors like Microsoft.
More importantly, open weight models are not falling behind the frontier. If anything, they are catching up. This is the rolling wave of commoditisation in action.
Copilot is becoming phenomenally more capable. It doesn’t match the frontier, but doesn’t need to: it only needs to get better and better over time so that Microsoft can bundle more and more intelligence into packages that IT departments already buy and workers already use.
Copilot adoption is absolutely humming - 30m paid seats as of June 2026 - and consumption-based charging is only just beginning, so the revenue impact is mostly in the future.
All of this supports the idea that Microsoft is the distributor of commoditised intelligence to enterprise, with a huge opportunity ahead of it as commoditised intelligence becomes more and more capable.
Personal experience with Copilot
On a personal level I find Copilot fascinating. I subscribe to Claude and Copilot through M365. I don’t pretend to be an advanced AI user, but I use it more and more. Claude is definitely ahead of Copilot, but Copilot is improving rapidly in absolute terms (and that’s before the massive upgrade announced last week). As I learn more I use Claude for more and more things. But as Copilot gets better I shift things from Claude to Copilot. Copilot’s advantages are that everything - my data, my work, my bills - stay in one place, and I am not tied to one model provider. A natural outcome of this process is that when Copilot can do everything I need it to, I’ll switch off my Claude subscription.
The main issue with Copilot is that it’s a fixed subscription, so Microsoft has to limit the context window (and therefore performance) in order to control margins. As a result, despite the fact that I have already upgraded my subscription, Copilot is the first product about which I ever remember thinking: I wish I could spend more on this. And lo and behold, Microsoft has just announced a new consumer tier, M365 Pro, at $99/month. That’s 12x M365 Personal and 6x M365 Premium. That’s TAM expansion, right there.
Thanks for reading - if you enjoyed reading this please like and restack, and do get in touch if you have questions.
Pete
Disclaimer: This post is for informational and educational purposes only. Building Arks is not licensed or regulated to provide any financial advisory service and nothing published by Building Arks should be taken as a recommendation to buy or sell securities, relied upon as financial advice, or treated as individual investment advice designed to meet your personal financial needs. You are advised to discuss your personal investment needs and options with qualified financial advisers. Building Arks uses information sources believed to be reliable, but does not guarantee the accuracy of the information in this post. The opinions expressed in this post are those of the publisher and are subject to change without notice. The publisher may or may not hold positions in the securities discussed in this post and may purchase or sell such positions without notice.