How to keep your micro coin gains

The lesson from last night is blunt: beating estimates no longer protects a hyperscaler when the market decides capex is the story. Alphabet shares slumped more than 10% after the search giant reported negative free cash flow during the third quarter. EPS came in at $2.12 against a $3.02 consensus, and revenue of $127.58 billion missed estimates of roughly $128.2 billion. Capital spending of $51.36 billion outpaced operating cash flow of $42.29 billion, leaving free cash flow at negative $9.07 billion. Microsoft showed strong cloud growth, yet the market treated the two companies very differently: shares popped after hours for Microsoft before pulling back, while Alphabet sank. Microsoft’s free cash flow fell 27% year over year to about $18.6 billion. The reason in both cases was the same: investors are selling down companies that spend far faster than their cash flow can absorb.

Amazon walks into that exact room tonight. Analysts forecast earnings per share of $1.96 and revenue of about $202.2 billion. Amazon expects to spend around $220 billion on capital expenditures in 2026. The surface-level risk is obvious. What experienced traders will look past that surface for is the quality argument Amazon has that Alphabet could not make as cleanly.

The core distinction is the backlog. AWS annualized revenue stands at roughly $169 billion, with performance obligations, primarily related to AWS, of about $496 billion as of June 30, 2026. In Q2, AWS grew about 37% and posted an operating margin around 39%. The consensus for Q3 AWS revenue sits near the mid-$40 billions, implying growth in the high 30s. Bank of America has argued for upside versus consensus on AWS growth, citing capacity and AI-related demand.

That distinction matters for how you read tonight. When Amazon raised its 2026 capex guidance to $220 billion after Q2, investors rewarded the AWS acceleration over the cash burn and sent the stock up 15.3% the next session, because the spend looked demand-driven rather than speculative. Alphabet did not get that pass last night. The critical question is whether Amazon can repeat that outcome by demonstrating, again, that its capex is contracted demand it cannot yet fill rather than a build-first bet that may not pay off.

Amazon’s capex argument rests on more than cloud contracts. AWS’s in-house chips division has crossed a run rate that Wall Street has not yet priced the way it prices dedicated semiconductor names.

Disciplined traders will not react to the EPS line alone. Watch AWS growth relative to that high-30s bar. A number that clearly undershoots expectations reopens the deceleration question on a bigger base and invites the same reaction Alphabet received. A number that clearly beats reopens the acceleration story. Watch operating income against the top of guidance at $26.5 billion. And watch Q4 guidance: analyst models cluster around roughly $239 to $246 billion in revenue, with operating profit expectations spanning the high $20s to mid $30s billions. Management’s tone on holiday demand will move the stock as much as the quarter itself.

One practical lesson from the last 24 hours: the market is not punishing AI spending indiscriminately. It is punishing spending without visible returns. Alphabet’s free cash flow turned negative for a second consecutive quarter at about minus $9.07 billion. Microsoft’s free cash flow declined 27% year over year to about $18.6 billion. If Amazon’s numbers show capex converting into accelerating, high-margin cloud revenue, the reaction could be the opposite of what traders saw last night. If AWS growth decelerates and management hedges on the Q4 outlook, expect the same selling that hit its peers, regardless of whether the headline EPS beats. Know which outcome you are watching for before the report drops.