Microsoft passed the free-cash-flow test. Alphabet failed it by $9 billion — and Amazon is next

Mitrade Insights — The test was set on October 26: with hyperscaler AI capex running near $800 billion a year, does cloud growth and contracted backlog pay for the buildout? Microsoft and Alphabet reported after the close on October 28, and the scorecard is unusually clean. Microsoft's Azure grew 48% and free cash flow was positive $18.6 billion. Alphabet's cloud grew 103% — faster — but free cash flow was negative $9.07 billion, and the stock fell about 9.3% after hours, from a regular close of 321.89 to 291.85. Amazon reports tonight, and it is the third referee.

* Chart source: TradingView (Alphabet Class C, Nasdaq: GOOG).

Microsoft spent $45.2 billion on capex in the quarter, below its own roughly $50 billion forecast because some third-party capacity arrived late. The market did not punish the shortfall; it rewarded the output. Azure revenue was $33.17 billion, up 48%, and management guided next-quarter Azure growth to 49%–50%. Commercial remaining performance obligations — the backlog — remain the cushion.

The $18.6 billion cushion

Free cash flow was positive $18.6 billion, well above the roughly $8.54 billion consensus. That is the number that matters. Microsoft is still spending heavily — FY2027 capex is projected above $190 billion — but it is generating cash while doing so. In a market with the 10-year Treasury at 5.045%, a company that can fund its own capex is worth a different multiple from one that cannot.

Alphabet's $9.07 billion hole

Alphabet's cloud revenue was about $30.7 billion, up 103%, more than double a year earlier. That is faster than Azure. But capex was $51.4 billion, 2026 capex is expected above $200 billion, and free cash flow was negative $9.07 billion — its second consecutive negative quarter. Revenue of $127.58 billion and EPS of $2.12 missed consensus. The market looked at the growth, then looked at the cash, and sold the stock.

Oracle added 10.6% after lifting its FY2030 revenue target to $225 billion and capping additional fundraising at $20 billion; Micron rose 5.9%. The market is not abandoning AI infrastructure. It is paying for the parts of the chain that produce cash today — and punishing the parts that cannot yet.

Amazon reports after the close tonight. AWS growth, retail margins and capex guidance decide whether the third hyperscaler confirms the spending story or joins the de-rating that started with Meta. Amazon closed at 254.60 on October 28; it is above its 200-day average and testing its 20- and 50-day averages near 254–256. The bar is no longer "did capex go up?" The bar is "did free cash flow survive it?"

What would make the second leg real

Three things would change the story from "AI capex is a cost" to "AI capex is an investment". First, Amazon's AWS growth has to beat the slowdown implied by its capex guide. Second, Alphabet's free cash flow has to turn positive in the next quarter; a third negative quarter would confirm the de-rating. Third, Microsoft's Azure guide of 49%–50% has to hold; a miss there would remove the last cushion.

The de-rating case is wrong if all three happen at once: Amazon's AWS accelerates, Alphabet's free cash flow turns positive, and Microsoft's Azure holds 49%–50%. Until then, the market will keep paying for cash, not growth — and the 10-year at 5.045% is the discount rate that makes the difference.