Here’s a question worth sitting with for a minute. When a company spends $41 billion in a single quarter on infrastructure, grows revenue 18% to $90 billion, beats earnings estimates by nearly 12%, and guides its next quarter’s cloud growth to 45%, what does the market do?

If you said “send the stock up 8% in after-hours and call it a day,” you’d be partially right. But here’s the part that matters more: before those results, Microsoft’s stock had already fallen roughly 22% from its all-time high. Shares were trading around $390. The P/E ratio had compressed to about 23x, against a 10-year historical average closer to 31x. That’s not a minor discount. That’s a company that looks measurably cheaper than it has been in most of the last decade.

So what went wrong? Nothing structural. What changed is the market’s patience.

The Concern That Created the Discount

Heading into this week’s earnings, Microsoft had spent aggressively on AI infrastructure. Capital expenditures surged more than 70% year over year to $41 billion in Q4 alone. Full-year capital expenditures reached roughly $116 billion. The company has also committed to an additional $329 billion in data center leases that haven’t even started yet. Investors watching those numbers alongside a slower-than-expected free cash flow figure, which came in at $19.6 billion for the quarter, started asking an uncomfortable question: is the return on all this spending real, or is Microsoft building a very expensive monument to a theme that might not pay off?

That fear, not any actual breakdown in the business, is what drove the stock down from its peak. Short interest heading into earnings had climbed to roughly 92 million shares, the highest level since 2015. The market was skeptical. And that skepticism is exactly what creates the kind of opportunity this publication exists to find.

What the Evidence Actually Shows

Azure grew 43% in Q4 of fiscal 2026. That’s the fastest pace in four years, and it cleared Wall Street’s estimate by more than 3 percentage points. For the full fiscal year, Azure crossed $100 billion in annual revenue for the first time. Microsoft 365 Copilot, the company’s AI work assistant, surpassed 30 million paid seats. The Intelligent Cloud segment, which includes Azure, posted $39.3 billion in quarterly revenue, up 32% year over year. Microsoft Cloud overall hit $59.3 billion, up 27%.

The company also returned $10.2 billion to shareholders in Q4 through dividends and buybacks. Total fiscal year revenue came in at $331.8 billion, up 18%. Operating income for the full year reached $155.2 billion, up 21%. Net income climbed 31% on a GAAP basis.

These are not the numbers of a company in distress. These are the numbers of a company whose spending is starting to deliver.

Slight tangent worth noting: the CFO Amy Hood acknowledged that new data center capacity is being monetized almost immediately after coming online. That’s a meaningful detail. It suggests the lag between spending and return is shorter than bears have been modeling.

The Valuation Disconnect

At roughly $390 per share before the post-earnings move, Microsoft was trading at a trailing P/E of around 23x. Its five-year average is closer to 32x. Its three-year average is nearly 33x. The price-to-free-cash-flow ratio had similarly compressed, sitting in the upper 30s versus a prior average in the mid-40s. Return on invested capital remains above 27%. Operating margins are running near 47%.

Compare that to a business growing revenue at 18%, guiding its largest cloud segment to 45% growth next quarter, sitting on $78 billion in cash, holding a contracted commercial cloud backlog of $678 billion (up 84% year over year), and you start to see what the market may be mispricing. The consensus analyst price target is around $557, suggesting roughly 40% upside from where shares were trading before last night’s results.

Is this cheap? In the traditional sense, no. Microsoft never trades at 10x earnings. It never has. The question is whether it’s cheap relative to its own history, its growth rate, and what a patient investor should reasonably expect from one of the highest-quality businesses on earth. And by that standard, the answer looks a lot more interesting.

The Risks Are Real. Don’t Dismiss Them.

Free cash flow declined 23% year over year as AI spending accelerated. Capital expenditures are expected to exceed $50 billion in Q1 of fiscal 2027. If Azure growth stalls, or if AI demand proves more cyclical than structural, the valuation case weakens quickly. The More Personal Computing segment continues to face a sluggish PC market. And competition from Amazon Web Services and Google Cloud is relentless.

There is also the accounting change to consider. Microsoft extended the useful life of its data centers from 15 to 25 years, which will reduce reported depreciation going forward. Management was clear that actual spending is unchanged. But it’s the kind of move that deserves scrutiny rather than a pass.

Where This Leaves Patient Investors

Microsoft is not broken. It is not even close to broken. What it has been is out of favor, which is different. The market spent much of the last year worrying that AI spending would outrun AI revenue. Last night’s results were the clearest evidence yet that the gap is closing. Azure at 43% growth, beating estimates by more than 3 points, guided to 45% next quarter, with demand still outpacing supply, is not the picture of a company that overbuilt.

The combination of a valuation sitting well below historical averages, a business compounding at double-digit rates, a $678 billion contracted backlog, and growing AI monetization is what disciplined investors look for. This is not a momentum call. It is a recognition that perception and fundamentals have diverged, and that the evidence is beginning to close that gap.

Whether the stock moves back toward its highs in one quarter or several is unknowable. What’s harder to dispute is the quality of what you’d be owning at a multiple this far below its own history.

The Cheap Investor