The old MicrosoftMSFT-- story dominated 2026: a capex machine running faster than it could ever justify, burning through cash while AI returned nothing but depreciation. The stock fell 29% from its 52-week high. Investors treated every data-center order as evidence the spending had no ceiling.
The fiscal fourth-quarter earnings report on July 29 tore through that narrative. Azure revenue grew 43% year over year, well ahead of the roughly 39% consensus. Microsoft cloud revenue reached $59.3 billion for the quarter, up 27%. Non-GAAP EPS of $4.74 smashed through the $4.24 consensus - an 11.8% beat. The stock rallied 15.5% the next day and has gained roughly 25% over the past five sessions, from the $389 area to around $489 where it sits today.
The headline numbers explain the reaction, but the number that matters for the next 12 months is the commercial remaining performance obligation.
Quick Backtesting Tool
RPO - remaining performance obligation - is contracted future revenue that Microsoft has already won but not yet recognized. It grew 84% year over year to $678 billion. That is not a soft pipeline signal or a management aspiration. That is revenue customers have signed for but won't show up on the income statement for months. An 84% increase at this scale tells you something the capex fear missed: Microsoft isn't building data centers into a vacuum. It's building them against contracted demand that's accelerating faster than the infrastructure itself.
The capex question
Here's the piece that still keeps the bears awake. Capital expenditures hit $35.8 billion in the fourth quarter, up 109.6% year over year. Full-year FY26 capex reached $115.9 billion. Analysts expect FY27 capex growth in the 20% to 30% range, implying roughly $139 billion to $151 billion.
The arithmetic that makes this tolerable is the one that the selloff ignored. Azure's revenue growth rate - 43% last quarter, guided at 45% for the first quarter of fiscal 2027 - is still running ahead of the spending growth. When the engine grows faster than the fuel burn, the buildout is funding itself. The market punished Microsoft for the same capex anxiety it had applied to other hyperscalers earlier this year, but the growth math is different here. Azure just crossed $100 billion in annual revenue. Copilot surpassed 30 million paid seats. AI Foundry now serves 100,000 clients, with revenue more than doubled year over year. These are not pipeline promises; they are consumption numbers from the quarter that just closed.
What the valuation is actually saying
Microsoft trades at roughly 27 times trailing earnings and 36 times forward earnings. That forward multiple might feel stretched for a company approaching $4 trillion in market capitalization - except the PEG ratio (price-to-earnings divided by earnings growth rate) sits at 0.87, meaning the stock's earnings power is growing fast enough to eat through that multiple. Revenue is growing at 18% year over year. Operating margin is 46.8%. Return on invested capital is 28.3%. These are not the margins of a company that's losing its way through a building spree.
Free cash flow for the trailing twelve months is $67 billion - down 6.5% year over year, dragged by the capex surge. That's the number that keeps bears up at night, and it's fair to examine. But free cash flow fell because capital spending rose, not because revenue or margins collapsed. Operating cash flow is $183 billion. The balance sheet carries $20.9 billion in cash and a net debt position of negative $36.6 billion - meaning Microsoft has more liquid assets than total debt. A dividend yield of 0.73% and a 20.7% payout ratio mean there's plenty of runway if the capex cycle softens.
The question isn't whether Microsoft can afford the spending. It's whether the spending keeps generating the kind of contracted revenue that the $678 billion RPO already proves it does.
The next 12 months
Management guided Azure growth of 45% for the first quarter of fiscal 2027. If that holds and revenue scales proportionally across the remaining quarters, full-year FY27 revenue runs near $390 billion. Applying a forward multiple in the 33x to 36x range - not a stretch for this growth and margin profile - puts the stock in the $600 to $650 range over the next 12 to 18 months.
That's not a DCF exercise or a mechanical multiple calculation. It's a bridge between the contracted revenue backlog that grew 84%, the Azure growth rate that's accelerating, and a valuation that was punished by a spending cycle it's already monetizing. AInvest's aggregate analyst consensus labels Microsoft a Buy, which doesn't tell you much by itself, but it does confirm the broader sell-side hasn't turned negative even after the capex headlines.
What would break the setup
A Q1 FY27 Azure print below 35% would be the first real warning sign. That growth rate is the hinge - if Azure decelerates that sharply while capex keeps climbing, the arithmetic flips and the spending genuinely does start outrunning the engine. A material slowdown in RPO growth from the 84% pace would also raise questions about whether the backlog is durable or front-loaded.
But until Azure growth softens below that 35% threshold, the old capex fear looks like a misread. The stock's 25% rally in a week closed the discount that existed before earnings, but it hasn't yet priced the $678 billion of contracted revenue that Microsoft just proved it can grow at nearly double-digit multiples of the previous year's pace.
Discipline over ego. If Azure stays above 35% and the RPO keeps expanding, the run toward $600-plus has legs. If it doesn't, the tripwire is clear.
