"Microsoft, Alphabet and Amazon Report in 48 Hours: Three Earnings That Will Test the AI Trade"

The next 48 hours will decide what the AI trade is worth, and the schedule is unusually compressed. On Wednesday, October 28, the Federal Reserve announces its rate decision at 2 p.m. Eastern; a few hours later, Microsoft and Alphabet report quarterly results after the close. On Thursday morning, US third-quarter GDP and the September PCE inflation reading arrive before the open; after that day's close, Amazon reports. In two days, the market gets the rate decision, the cloud results, and the inflation data — the three things the whole AI trade is currently priced on.

The backdrop is unforgiving. Last week brought an AI-linked selloff, the Nasdaq's worst two-week decline since April 2025, a flight into Treasuries that pushed the 10-year yield back below 5%, and high-yield credit spreads widening to 340 basis points. Gartner warned this week that AI data center owners could face up to $4 trillion in asset write-down risks by 2032 if hyperscaler revenue fails to cover massive capital expenditure commitments. Meta has already reported: revenue beat expectations, but a plan to issue non-voting stock to fund its spending, plus a $25 billion privacy penalty in New Mexico, dented the optimism. Tesla, reporting the same evening, cut its spending forecast. The lesson from this season so far is that a beat alone may not be enough — beats have to be big to move prices, and guidance matters more than the headline number, because the market has already paid for a good quarter in advance.

What to watch is therefore specific. At Microsoft, Azure's growth rate and any framing of fiscal 2027 capital spending. At Alphabet, whether Google Cloud's growth and AI monetization in Search show revenue catching up with the company's AI investment. At Amazon, AWS growth against the roughly $220 billion in 2026 capital spending management guided on its July call, and whether fourth-quarter retail guidance holds. The thread connecting all three is the same question behind the credit anxiety: whether AI spending is producing revenue fast enough to justify itself. The cloud results are where that question gets answered in numbers rather than narratives.

My judgment is that all three will beat revenue consensus — not because beats are easy, but because the demand signals point the same way: Meta's revenue beat, enterprise AI spending still expanding, and Amazon's own statement that it is adding capacity for demand it cannot yet serve. Earnings beats are also correlated rather than independent: a strong quarter for enterprise technology spending tends to lift all three clouds together. The honest uncertainty is that the bar is already high — consensus expects strong quarters, so "beat" means clearing a bar the market has already raised, and as Meta showed, a revenue beat can still be met with a selloff if guidance disappoints. But the prediction is about revenue against consensus, not about the share price reaction, and on that narrower question the evidence leans positive. The answers arrive Wednesday and Thursday evening; there is no waiting until next quarter to find out.