MSFT
Microsoft
Summary
What they do:
Operates Azure (the second-largest cloud platform globally), builds and deploys AI infrastructure at hyperscale ($100B+ annual capex run-rate), and monetizes AI through Copilot across the world's dominant enterprise software ecosystem (Microsoft 365, Teams, GitHub, Dynamics).
Why they matter:
Microsoft is THE CUSTOMER for the entire AI supply chain. When Microsoft commits $80B+ in AI capex, every upstream company's order book activates — NVIDIA ships GPUs, TSMC runs fabs, Arista ships switches, Vertiv ships cooling. Microsoft's capex guidance is the single most-watched demand signal in AI infrastructure.
Recent performance:
Q2 FY2026 (Jan 2026) revenue $81.3B (+17% YoY), Azure +39% CC, Microsoft Cloud $51.5B (+26% YoY). Capex $37.5B in Q2 alone (+66% YoY). Stock ~$386, down ~19% YTD despite earnings beats. Market cap ~$2.9T.
Our Verdict
The enterprise AI distribution machine — Azure growth sustaining at 37-39%, Copilot reaching 15M seats, and $100B+ annual capex confirming multi-year infrastructure commitment — but the stock's 19% YTD decline reflects growing market skepticism about AI capex returns that creates a rare entry point for the most durable AI business model.
Structural trends
Structural
77
/ 100
Moat
9/10
Enterprise lock-in
AI Exp.AI Exposure
High~65% AI
Play Type
ConsensusAI Growth
~35-40%
Rel. Value
51
ATTRACTIVEPriceLIVE
$393.11
+2.27%
Live via Yahoo Finance · refreshes every 5 min
Market Cap
$2.9T
P/E Ratio
24.6
P/S Ratio
9.6x
52W High
$555.45
52W Low
$355.67
52W Chg
10.5%
Beta
1.11
Microsoft is simultaneously the most important customer and the most important distributor in the AI economy. As a customer, Microsoft's $100B+ annual capex run-rate drives demand across every layer of the AI supply chain — from NVIDIA GPUs to Corning fiber optic cable to Eaton electrical equipment. As a distributor, Microsoft's enterprise relationships (95%+ of Fortune 500 companies use Microsoft products) provide a channel for AI monetization that no startup or competitor can replicate.
The company operates in three segments. Intelligent Cloud ($32.9B Q2 revenue, +29% YoY) houses Azure, the growth engine. Azure revenue grew 39% in constant currency in Q2, with AI services contributing an estimated 13 percentage points of that growth. This segment is where the AI capex gets deployed and where AI revenue gets recognized. Productivity & Business Processes ($29.4B Q2 revenue) includes Microsoft 365, LinkedIn, and Dynamics. This is where Copilot monetization lives — 15M commercial seats as of Q2, up from essentially zero two years ago. More Personal Computing ($14.7B Q2 revenue) covers Windows, Xbox, and Surface — relevant to the AI thesis primarily through Windows AI features.
The OpenAI partnership remains Microsoft's strategic crown jewel. Microsoft holds exclusive cloud rights to OpenAI's models, has invested $13B+ cumulatively, and is the technology partner for the $500B Stargate project (alongside SoftBank and Oracle). This gives Microsoft a structural advantage in access to frontier AI models that feeds directly into Azure AI services and Copilot products.
Microsoft's data center footprint exceeds 800 facilities across 60+ regions. The newest campuses are purpose-built for AI, housing NVIDIA Blackwell GPU racks, connected by Arista switches, cooled by liquid cooling systems, and powered by utility-scale power including the Three Mile Island nuclear restart agreement with Constellation Energy (1,000MW dedicated clean power). First-half FY2026 capex reached $72.4B, putting the company on track for ~$100B annually — approximately two-thirds directed to short-lived assets (GPUs, CPUs).
Microsoft also designs custom silicon. Project Maia (AI accelerator) and Cobalt (ARM-based CPU) are manufactured by TSMC and deployed in Azure, gradually reducing dependence on NVIDIA for specific workloads while maintaining NVIDIA as the primary GPU supplier for the foreseeable future.
Supply Chain Dependencies
Upstream Suppliers
gpu_supplier · weight 0.8
networking_supplier · weight 0.6
server_supplier · weight 0.5
optical_supplier · weight 0.5
networking_customer · weight 0.5
asic_customer · weight 0.5
ip_customer · weight 0.4
hyperscaler_customer · weight 0.5
chip_customer · weight 0.4
optical_customer · weight 0.5
optical_customer · weight 0.5
optical_supplier · weight 0.6
networking_customer · weight 0.5
storage_customer · weight 0.5
storage_customer · weight 0.4
server_customer · weight 0.5
server_customer · weight 0.4
partner · weight 0.3
manufacturing_customer · weight 0.4
cloud_customer · weight 0.5
site_selection_customer · weight 0.3
power_customer · weight 0.5
power_customer · weight 0.5
power_customer · weight 0.3
power_customer · weight 0.5
construction_customer · weight 0.4
electrical_customer · weight 0.5
fiber_customer · weight 0.5
cooling_customer · weight 0.6
modular_customer · weight 0.4
colocation_customer · weight 0.6
colocation_customer · weight 0.5
storage_customer · weight 0.3
tower_customer · weight 0.3
fiber_customer · weight 0.3
cloud_customer · weight 0.8
Downstream Customers
The Catch
Microsoft is spending $100B+ annually on AI infrastructure — more than the GDP of most countries — and the return on that investment is still a faith proposition. Azure AI services contribute 13 percentage points to Azure's 39% growth, but whether that justifies $37.5B per quarter in capex remains unproven. Copilot has 15M seats but needs 50M+ to meaningfully move the financial needle. The stock's 19% YTD decline despite consecutive earnings beats suggests the market is losing patience with the "spend now, monetize later" narrative. If AI adoption curves disappoint — if enterprises decide Copilot is $30/month they can live without — the most capital-intensive bet in technology history will need to be unwound, and Microsoft's margins will compress before the revenue comes.
If They Win
If Azure AI services accelerate to 20pp+ of Azure growth, Copilot reaches 50M+ enterprise seats generating $18B+ in annual recurring revenue, custom silicon reduces AI serving costs by 40%+, and the Stargate project establishes Microsoft-OpenAI as the dominant AI platform, then Microsoft becomes the operating system of the AI economy — not just the cloud provider but the enterprise AI distribution layer that every company on earth runs through. Revenue compounds to $400B+ by 2028, free cash flow margins recover as AI utilization improves, and the market re-rates Microsoft from a "capex-heavy cloud company" to "the enterprise AI monopoly." At that trajectory, the current $2.9T market cap looks cheap against a $4-5T potential as AI generates the same kind of durable enterprise revenue that Windows and Office did for three decades.
Others in Run the Workload
Not financial advice. All scores generated via AI algorithms using public data.