Azure Just Crossed $100 Billion. The Market Spent 18 Months Getting This Wrong.

Imagine a toll road company that spent two years laying asphalt while critics said there was no traffic coming.

Then one evening, $678 billion worth of signed traffic contracts appeared on the books.

That is what happened to Microsoft $Microsoft (MSFT.US)$ last night.

What Microsoft Reported

Revenue: $90.01B (+18%)

EPS: $4.74 vs $4.24 estimate

Azure growth: +43% YoY

Azure annual revenue: crossed $100B for the first time ever

Cloud backlog: $678B (+84% YoY)

Copilot paid seats: 30 million enterprise

FY2027 capex guidance: $255-260B (up from $190B)

Stock reaction: +8% after hours

On a night when the $Dow Jones Industrial Average (.DJI.US)$ fell 840 points and the Fed held with three dissenters voting to hike.

Azure Crossed $100 Billion. Here Is What Microsoft's $678B Backlog Means for Alphabet, Amazon, Meta and Apple

The Market Was Wrong for 18 Months

For 18 months the question was the same: you are spending $190 billion on AI infrastructure, where is the revenue?

The stock fell 18-20% from its highs. The multiple compressed from 30x to 20x forward earnings. The narrative was: capex is accelerating, returns are unproven.

Last night Microsoft showed a reservation book with $678 billion in committed enterprise spending. Up 84% in one year. That is not a forecast. That is signed paper.

The market did not have a capex problem. It had a disclosure problem. The business was compounding the whole time. Azure grew 40% in Q3. 43% in Q4. Copilot went from zero to 30 million paid seats in under 24 months. The backlog was growing the whole time the stock was falling.

What It Means for the Other Four

Alphabet ( $Alphabet-C (GOOG.US)$): Microsoft's $678B backlog validates Alphabet's $514B backlog. Combined that is $1.19 trillion in signed enterprise AI cloud commitments. The market sold Alphabet on capex fear after its earnings. Microsoft answered that argument last night. The narrative shifts from "is the capex justified" to "is execution fast enough."

Amazon ( $Amazon (AMZN.US)$): Reports tonight after close. The last hyperscaler. If Amazon discloses a comparable backlog, three companies hold $1.7 trillion in committed enterprise AI cloud spending. Watch AWS growth rate vs Azure's 43% and Google Cloud's 82%. Azure at 43% vs AWS at 28% last quarter is a gap that needs an answer.

Meta ( $Meta Platforms (META.US)$): The contrast case. Revenue beat (+28%). EPS missed badly ($6.18 vs $7.17). Expenses up 55%. Operating margin from 43% to 31%. FCF $784M on $60B revenue. Stock -9.64%. The difference: Microsoft has $678B in signed contracts justifying the capex. Meta has advertising optionality. The market is not anti-capex. It is anti-capex-without-proof.

Apple ( $Apple (AAPL.US)$): Reports tonight after close. Different game entirely. Apple monetises AI through hardware upgrades and Services attach rate across 1.5 billion active devices. Watch Services revenue above $27B and any commentary on Apple Intelligence driving upgrade demand.

The market spent 18 months asking the wrong question.

The question was never whether AI infrastructure was expensive. It was always whether anyone had actually signed up to pay for it. Last night Microsoft showed $678 billion in signed contracts. A week ago Alphabet showed $514 billion. Tonight Amazon closes the picture.

That is $1.19 trillion in committed enterprise spending before Amazon reports. Not projections. Not analyst estimates. Signed paper.

The cost phase was always going to be visible. Big numbers are always visible. What the market missed is that the return phase does not announce itself with a press release. It announces itself with a backlog. And the backlog is already here.

The businesses that own the next decade are not the ones spending the most. They are the ones with the most customers already committed to paying. Microsoft proved that distinction last night. The rest of the proof arrives this week.

231K Views