Tonight is the night. After the closing bell, four of the largest companies on earth report earnings in a span of about 80 seconds. Microsoft, Meta, Alphabet, and Amazon all drop their results within minutes of each other, and options traders are pricing in more than $800 billion in market cap movement as a result. That’s the stock equivalent of a controlled demolition and a building boom happening at the same time.
I’ve been writing about the AI spending question all week. Monday we covered the Microsoft-OpenAI breakup. Tuesday we watched the market wobble on a Wall Street Journal report about OpenAI’s internal numbers. Now, tonight, we get something closer to actual answers. The companies doing the spending will have to justify it.
Here’s what I’m watching, for each one.
Microsoft (MSFT)
The setup: Analysts expect roughly $81.4 billion in revenue, up about 16% from a year ago. Earnings per share around $4.06 to $4.07. But the headline number almost doesn’t matter. What matters is Azure.
Microsoft’s Azure cloud has been the cleanest AI growth story in tech. Last quarter, Azure grew 31%. Analysts are looking for that to accelerate. If Azure comes in at 33% or higher, the stock goes up. If it misses or if guidance is cautious, all the goodwill from the OpenAI deal restructuring last Monday gets unwound fast.
Two things I’ll specifically listen for: How does management frame the revenue impact of the OpenAI partnership change? And do they mention any acceleration from customers running OpenAI workloads through Azure now that OpenAI has more cloud freedom? That second question didn’t exist two weeks ago. Now it matters.
Alphabet (GOOGL)
Alphabet heads into tonight with its core search business under genuine pressure for the first time in years. AI competitors are reshaping how people find information. The question isn’t whether Google is losing market share yet. The question is whether investors can see it coming.
The thing I want to hear about is Google Cloud. Last quarter it grew 48% and hit a run rate above $70 billion per year. That’s a real business now, not a talking point. If cloud keeps accelerating, Alphabet has a credible story that it wins even if search slowly loses some traffic to AI assistants. If cloud slows, that story falls apart.
Alphabet guided $175 to $185 billion in capital spending for all of 2026. Tonight they need to show the first signs that spending at that scale is generating returns. Any sign of margin pressure without corresponding revenue acceleration will hurt the stock.
Amazon (AMZN)
Amazon is the most complex of the four because it’s playing multiple games simultaneously. The consumer business is exposed to tariffs and inflation. AWS is the gold standard in cloud computing. The advertising business has become huge and mostly gets ignored. And the company just restructured its relationship with OpenAI on the backend.
AWS is what I’m watching. If AWS growth reaccelerates tonight, that will be the first concrete sign that the OpenAI partnership opening is already generating real enterprise demand. It’s been less than three days since the deal closed, but management may hint at early signals in their guidance commentary.
Amazon reports North American operating margin too, which has been a happy surprise for the last several quarters. If that holds above 6%, the consumer side looks fine regardless of macro noise.
Meta Platforms (META)
Meta is the one I’m watching most closely. Not because it’s the most important company tonight, but because it’s the clearest test of whether AI spending pays off for a company that actually has to sell ads to justify it.
Meta is spending north of $60 billion on AI infrastructure this year. Unlike Microsoft or Alphabet, Meta doesn’t have a cloud business where customers pay for compute. Meta has to monetize AI through better ad targeting, through Llama, through creator tools. The question is whether that spending is showing up in revenue per user or in engagement metrics.
Last quarter, Meta’s revenue-per-user numbers surprised everyone to the upside. If that continues tonight, it validates the idea that AI can directly lift ad revenue. If it slows, the $60 billion capex number looks like a problem, not an investment.
One More Thing — The Fed
As of this afternoon, the Federal Reserve held rates steady at 3.5% to 3.75%. Jerome Powell held what is likely his final press conference as Fed Chair. A divisive meeting, per the New York Times. His successor, Kevin Warsh, is expected to take over soon.
None of that directly moves tonight’s earnings. But it sets the backdrop. If Powell said anything that changed rate expectations, that will layer onto whatever big tech reports. Worth knowing going in.
What I’ll Do With What I Hear
I’m not planning to make any trades based on tonight’s results. I’ve said before that trading earnings is a coin flip most of the time, and the short-term noise after a report is usually just that — noise.
What I’m listening for is signal about the next six months. Are these companies seeing actual revenue from AI infrastructure? Are their customers getting value, or just spending? Is the buildout getting faster or starting to slow? Those answers will shape how I think about the whole technology sector for the rest of 2026.
I’ll have thoughts on what we learned tomorrow. Tonight, I’ll be watching.
— Tom
