Enterprise technology companies have spent two years committing enormous capital to artificial intelligence infrastructure. The outlays have been hard to ignore. What has been harder to establish is whether the spending is converting into the kind of revenue growth that justifies doing it at that scale.
Microsoft (MSFT) has been at the center of that debate. The company has poured tens of billions of dollars into AI-related infrastructure, and its stock has moved with every signal about whether the bet is working.
Shares hit a 52-week low of $349.20 before recovering above $520 this fall. That range reflects how much the market is still working through on the question of AI returns.
Morgan Stanley has not wavered.
The firm shared a note with TheStreet on Nov. 12 laying out why it thinks Microsoft belongs significantly higher than where it is currently trading. The analysis covers what Morgan Stanley describes as an enterprise AI flywheel, a four-part stack the firm believes is starting to gain real momentum across the business.
Azure growth hit 48% and accelerated for the second straight quarter
Microsoft reported fiscal first-quarter 2027 results on Oct. 28. Azure grew 48% year-over-year (YoY) in constant currency, six percentage points faster than the quarter before. Azure’s 48% came in above analyst estimates and outpaced AWS, which grew 39.3% over the same period.
Total revenue for the quarter came in at $92.3 billion, up 19% YoY. Operating margin held near 50%. Capital expenditure doubled to $39 billion, money Microsoft is deploying to build out data center capacity for AI workloads. Microsoft’s Q1 gave investors the AI relief they had been waiting for, and Morgan Stanley’s note landed two weeks later, reinforcing that read.
Also read: Microsoft’s Oracle move raises new question for investors
The firm’s analyst team, led by Adam Wood, identified Azure as the foundation of the investment case. Cloud compute has to keep growing for the rest of Microsoft’s AI products to have somewhere to run. The 48% result came in above what the firm had modeled, and it shifted the near-term picture in Microsoft’s favor.
Morgan Stanley also flagged Microsoft’s Maia custom chip development as a variable worth watching. Azure currently runs on hardware from Nvidia (NVDA) and other third-party suppliers.
Microsoft has been building its own chips to reduce the cost of running AI inference at scale. Progress on Maia economics would add a margin story to what is already a growth story on Azure.
Copilot user numbers climbed, but the revenue proof is still coming
Copilot daily active users rose roughly 50% quarter-over-quarter in fiscal Q1 2027. The number of customers actively consuming Copilot credits climbed more than 75% over the same period. Microsoft disclosed both figures on its earnings call.
Morgan Stanley found those numbers encouraging but was careful about how it framed them. User growth and revenue growth do not always move at the same speed. Many enterprise customers have been running Copilot in select teams or departments. The harder test is whether those limited deployments expand across full organizations.
The firm is specifically watching Microsoft’s E7 enterprise license tier. E7 bundles Copilot with security and compliance tools and carries a higher per-seat price than standard Microsoft 365 plans.
If more large customers migrate toward E7, Microsoft’s average revenue per user rises without the company having to add new accounts. That is a more durable kind of growth.
Microsoft’s M365 Commercial Cloud segment grew 18% in constant currency in fiscal Q1 2027.
Second-quarter guidance pointed to roughly 19% constant currency growth. Morgan Stanley expects Copilot monetization to show up more visibly in those numbers as the year progresses. The user metrics in Q1 suggest the base is there. The question now is conversion.
Global Visual Network / Getty Images
Foundry revenue tripled as enterprises move from testing to building
Foundry is Microsoft’s platform for companies that want to build their own AI applications rather than deploy Copilot as it comes out of the box. Revenue from Foundry more than tripled YoY in fiscal Q1 2027. The number of customers building multi-model agentic systems on Foundry grew five times over the same period.
More than 19,000 companies are now using both Microsoft Fabric and Foundry together, up from more than 17,000 the previous quarter. Fabric handles enterprise data organization. Foundry manages model orchestration.
Morgan Stanley views the combined Fabric-plus-Foundry customer count as one of the cleaner indicators in the whole dataset, because companies using both are deeper in the Microsoft stack and harder to displace.
They also tend to expand Azure spending as their AI workloads grow.
More Microsoft:
- Microsoft’s Windows sales fell 22% as PC prices climb
- Microsoft CEO sees a problem ahead for AI stock investors
- Microsoft just gave Nvidia shareholders a reason to pay attention
A company that has organized its data in Fabric and is running custom models through Foundry is not going to shift infrastructure providers without significant cost and disruption. That stickiness is part of what the flywheel concept is built on.
The Foundry growth also signals a change in what enterprises are actually doing with AI. Early adoption was mostly about giving employees access to Copilot and similar tools. Building custom applications requires real infrastructure decisions, long-term contracts, and internal development resources. That is a different kind of commitment.
What the $650 target and $800 bull case are built on
Morgan Stanley’s base case price target on Microsoft is $650. The stock was trading at $521.13 on Nov. 11, the day before the note was published. The base target sits roughly 25% above that level.
The firm’s bull case goes to $800. That scenario requires the enterprise AI flywheel to broaden significantly, with Azure growth holding above 45%, Copilot monetization expanding across large enterprises, and Foundry establishing itself as a meaningful standalone revenue contributor.
Azure’s 48% Q1 print was the strongest result in several periods, and it is the number Morgan Stanley is using as the foundation for its upside scenario.
The bear case is $300. Morgan Stanley described it as the outcome where large-scale AI infrastructure investment fails to convert at the pace the market is currently pricing in. The firm does not view that as the likely path.
Microsoft’s next earnings report is projected for Jan. 27, 2027, covering fiscal Q2 2027. That gives the company roughly ten weeks to generate data that either supports or complicates the Morgan Stanley thesis.
Morgan Stanley analysts are scheduled to meet with Chief Financial Officer Amy Hood and Microsoft’s investor relations team on Nov. 18. Those conversations will shape how the firm is sizing the gap between the stock’s current price and where it thinks the stock belongs.
Related: JPMorgan set a serious Microsoft stock price target for 2027
