Microsoft is currently underperforming the market with a roughly \~20% drawdown year to date. They are also the only company that is facing all three of the "major threats" that have overshadowed the 2026 market space.
Those are:
* Fear of Artificial Intelligence "eating" software
* Fear of OpenAI's commitments
* Fear of CapEx overspend
# VALUATION:
The most simple valuation metric to look at for Microsoft is PE. Current PE is 22.5, forward PE at \~19.50. Future revenue growth is expected around 13-18% over the next 5 years. However, that's not enough to look at.
Looking at JUST the Intelligent Cloud segment, we see that for Q3 2026 (Microsoft Timetables are strange): Intelligent Cloud revenue was $34.7 billion and growth was 30%.
For FY 2026, the Intelligent Cloud segment is projected to have an annual revenue of \~$135.2 billion, (97.1B from Q1-Q3 actuals + Q4 guidance).
Now, let's consider operating margin at \~44%. A proper multiple for revenue growing at 30% would be 30x EV/EBIT, factoring in near-term capital "drag" from CapEx spending.
That gives a valuation of just the Intelligent Cloud segment at $1.785T.
* Azure is also expected to "moderately reaccelerate" in the second half of calendar year 2026, past 40%+ revenue growth, which is not factored into Microsoft's 2026 timetables.
Microsoft currently trades at a market cap of $2.82T. So, for the remaining \~1.04T you'd get:
* Microsoft 365 Commercial (Excel, Word, PowerPoint, Teams, CoPilot)
* Dynamics 365
* LinkedIn
* Windows OS / OEM licensing
* Xbox + Activision Blizzard
* Bing / Ads
* Surface devices
* Net cash (\~$26B)
* OpenAI stake (\~$150B - 300B)
* Anthropic stake (\~$10–20B)
Just for clarity, Intelligent Cloud made: $34.7 in revenue vs everything else at $48.2B. Though, everything else should be given a lower multiple due to lower growth.
The "software" segment's Microsoft owns, are effectively being priced at very low multiples already, as if they are already being priced for failure even before any impact has been shown.
# OpenAI:
Microsoft has been scrutinized this year for their reliance on OpenAI for future growth. However, their RPO is $344B, excluding OpenAI's $281B portion. If OpenAI even make good for half of their commitments, that is an RPO of $485B.
However, I believe that in the long-run OpenAI will be an AI-winner (or a closer second). I do not believe Anthropic will be able to significantly surpass them.
Reasons:
* OpenAI has investments from NVIDIA, Microsoft, and Amazon.
* OpenAI "wants it" more than Anthropic.
* Anthropic tries to be ethical. They refused deals from the government due to ethical concerns (even if it meant them being blacklisted), Dario Amodei going on TV and saying Anthropic should be taxed more, being careful around releases to not disrupt corporations, etc.
* OpenAI controls the majority of non-enterprise use and is taking share from Claude in enterprise via Codex.
* OpenAI is rolling out ads and trying to generate revenues in ways Anthropic would not even consider.
Also, many assume that Microsoft's rival is Anthropic. However, when Anthropic was labeled a supply chain risk, guess who stepped up? Not Google, not Amazon, but Microsoft:
["Microsoft backs AI firm Anthropic in legal battle against Pentagon."](https://www.theguardian.com/technology/2026/mar/12/microsoft-amicus-brief-anthropic-pentagon)
* Microsoft has thrown its weight behind Anthropic’s legal challenge against the Pentagon, filing a court brief in support of the AI company’s effort to overturn an aggressive designation that effectively bars it from government work.
It may seem that Anthropic succeeding hurts their investments in OpenAI and would hurt their software suite. However, what matters more than that, is that Microsoft wants to distribute Claude models throughout their platforms. They want Anthropic to spend on Azure. They also want Anthropic to use Microsoft's [chips](https://www.cnbc.com/2026/05/21/anthropic-microsoft-maia-200-ai-chip.html).
# Fear of CapEx Overspend
This I believe is the most "silly" point. Microsoft is not Meta. Microsoft has a higher credit rating than the United States government at AAA. They have held this rating since 2008 and it is higher than all of big techs. They know how to PROPERLY spend.
What I have noticed, in the market, is the recent trend is Microsoft, Meta (and to some extent Amazon and Google) have been trading inverse to AI infrastructure (Marvel, MU, AMD, INTC).
The reasoning is as follows:
* The AI spend from Microsoft (and others) does not show proper profitability.
* The AI spend from Microsoft (and others) show that AI spending will continue and AI infrastructure suppliers will benefit.
This assumes, in the long run, that Microsoft (and others) will continue to spend on AI WITHOUT strong profitability. That Microsoft (and others) will spend just to spend. It is obvious, that in the long run, if AI does NOT show substantial returns for Microsoft (and others) they will slow their spending. Then, AI infrastructure suppliers will crumble.
As such, there is no way for you to be BULLISH on AI infrastructure suppliers and bearish on AI spenders.
Microsoft's AI spend, like Google and Amazon is MOSTLY concentrated on their cloud businesses. Which are commitments from Anthropic, OpenAI, etc that they will spend on compute and training. If these commitments slow down, so will their spending. So if you're bullish on AI infrastructure, you have to be bullish on compute: Microsoft, Amazon, and Google. Similarly, you have to be bullish on Anthropic/OpenAI.
# Bonus Point: CoPilot Sucks
CoPilot, is not the best. However, as a Microsoft shareholder, I'd prefer if they did not invest into building their own models as a priority. In Q2 2026 Azure was capacity constrained and Microsoft missed Azure growth by 1% as they dedicated those resources to training their own AI models instead of leasing out compute. I believe this is not the way to go.
Microsoft benefits more by:
* Receiving money from Anthropic/OpenAI/etc for compute
* Revenue sharing by distributing these models through enterprise
By doing so, Microsoft would show clear profitability. Investing in AI models, won't. AI models are now becoming commoditized, meaning they can easily be swapped for another. Why would Microsoft spend resources developing their own models with unpredictable outcomes rather than collecting cash for their compute/distribution scale?
However, CoPilot needs to be "good enough" to properly distribute models or else that becomes problematic. They do not, however, need to train their own models.
# POSITION:
* 300 shares at \~393.50 average
