Microsoft Stock in 2026: Azure Grew 39%, but AI Capex Grew 60%

So the stock debate is not whether Microsoft has demand. It clearly does. The debate is whether that demand stays large enough, profitable enough, and durable enough to keep the AI infrastructure bill from becoming the main story.

Source Evidence Snapshot

Source capture: Microsoft FY26 Q2 Investor Metrics, captured 2026-04-08 from the investor metrics page showing commercial remaining performance obligation, Microsoft Cloud revenue, gross margin, and Azure growth rows.

The investor metrics table is the fastest way to see the quarter's core tension. It shows commercial remaining performance obligation at $625 billion, Microsoft Cloud revenue at $51.5 billion, Microsoft Cloud gross margin at 67%, and Azure and other cloud services revenue growth at 39% year over year. Demand is clearly large. Margins are still high. But the margin line is no longer moving up.

Source capture: Microsoft FY26 Q2 Segment Results, captured 2026-04-08 from the segment results page.

The segment table gives the operating frame. Intelligent Cloud revenue was $32.9 billion in the quarter and $63.8 billion over the first six months of FY26, while segment operating income was $13.9 billion in the quarter and $27.3 billion over the first half. In other words, the cloud engine is still large enough to throw off real profit, not just fast top-line growth.

Source capture: Microsoft FY26 Q2 Cash Flows, captured 2026-04-08 from the cash-flows statement page.

This is the line investors cannot skip. Microsoft reported first-half operating cash flow of $80.8 billion, common-stock repurchases of $13.1 billion, cash dividends of $12.9 billion, and additions to property and equipment of $49.3 billion. That means the company is still producing huge internal cash, but it is also reinvesting at a rate that would have looked extreme for Microsoft only a few years ago.

Azure Demand Is Still Running Well Ahead of Normal Big-Tech Growth

The easiest bullish read is also the simplest one. Azure and other cloud services revenue grew 39% in the quarter, and Microsoft Cloud revenue reached 51.5 billion dollars. That is not the profile of a business running out of commercial demand.

The backlog line makes the same point from a different angle. Commercial remaining performance obligation rose to $625 billion. That number matters because it represents contracted but unrecognized revenue still sitting ahead of the income statement. It does not all convert at once, but it tells you demand is not being scraped together quarter by quarter.

Microsoft's own performance page reinforces the message. It says revenue increased by $11.6 billion, driven mainly by Intelligent Cloud and Productivity and Business Processes, and that Microsoft Cloud gross margin percentage slipped to 67% because of continued investments in AI infrastructure and growing AI product usage. That last clause matters. Margin pressure is not showing up because demand vanished. It is showing up because demand is expensive to serve.

That is a much healthier problem than weak bookings. Still, it is the problem investors have to watch.

The Important Cloud Number Is Profit, Not Just Revenue

Some AI stories get trapped at the top line. Microsoft's filings are more useful because the segment table still shows strong operating income inside Intelligent Cloud.

Quarterly Intelligent Cloud revenue reached $32.9 billion, up from $25.5 billion a year earlier. Operating income for the segment reached $13.9 billion, up from $10.9 billion. Over the first six months, the segment produced $27.3 billion of operating income on $63.8 billion of revenue.

That is why the Microsoft case is stronger than a generic "AI infrastructure is expensive" story. The cloud business is not just growing fast enough to absorb the capex conversation. It is still growing profit dollars as well.

But the gap between growth and spending is now narrow enough to matter. If Azure growth slows materially before the current investment cycle matures, investors will notice quickly because the fixed-cost base is rising so aggressively.

AI Capex Is No Longer a Background Detail

The cash-flow statement is where the quarter really becomes a stock-analysis piece instead of a product story.

Microsoft generated $80.8 billion of operating cash flow in the first six months of fiscal 2026. That is enormous. But the same statement shows additions to property and equipment of $49.3 billion, up from $30.7 billion in the comparable prior-year period. That is about a 60% increase.

This is the number that makes the title honest. Azure grew 39%. Property and equipment additions grew even faster.

That does not automatically make the spending reckless. Microsoft can afford it. The same filing also shows the company still returned capital through $13.1 billion of share repurchases and $12.9 billion of dividends during the first half. The business remains very cash generative.

Still, the quarter makes one thing clear: Microsoft is funding AI infrastructure at a scale where future returns have to stay visible. Investors are no longer being asked to underwrite only near-term product momentum. They are underwriting a multi-year datacenter and model-capacity buildout.

What Matters for MSFT From Here

The cleanest way to read the stock now is to track three relationships instead of one.

First, watch whether Azure growth stays strong enough to justify the pace of infrastructure additions. A cloud business can support heavy capex when growth is still this high. The risk shows up if growth normalizes faster than spend does.

Second, watch commercial remaining performance obligation and Microsoft Cloud gross margin together. The backlog line tells you whether demand is still forming in front of revenue. The margin line tells you how expensive that demand is to fulfill.

Third, keep segment profit in the picture. Intelligent Cloud operating income is what stops the story from collapsing into "great demand, terrible economics." Right now, those economics are still solid. The point is not that Microsoft's spending is out of control. The point is that the stock increasingly depends on continued proof that the spending is productive.

That is why this quarter is useful. It does not support a lazy bear case, because demand and profits are both real. It also does not support a lazy bull case, because the infrastructure bill is no longer small enough to ignore.

The most defensible conclusion is that Microsoft still looks like a high-quality compounder with real AI demand behind it. But the stock's next leg will depend less on whether AI is exciting and more on whether Azure growth, backlog conversion, and cloud profitability stay comfortably ahead of the capex curve.