Monday, in this very newsletter, we said Wall Street was walking into Broadcom’s print wanting one thing: roughly 140% AI growth. The company delivered 143%. The stock fell anyway.
Welcome to the strange math of a market priced for perfection.
The quarter itself was a monster. Broadcom posted record Q2 revenue of $22.2 billion, up 48% year over year. AI semiconductor revenue hit $10.8 billion, growing 143%. Management guided Q3 AI revenue to $16 billion — a 200% year-over-year leap. Adjusted EBITDA came in at a record $15.2 billion, 69% of revenue.
So why did shares drop as much as 13% in after-hours trading? Because total revenue came in a hair light against the loftiest estimates — $22.19 billion versus roughly $22.27 billion expected. On a name that had run to a $2 trillion-plus valuation on relentless AI enthusiasm, a “narrow miss” on the headline line was enough to trigger profit-taking, even as the part of the business everyone actually cares about — AI — blew the doors off.
This is what happens when expectations get stretched to the ceiling. The bar wasn’t “grow fast.” The bar was “grow fast and beat every line.” Miss one and the algorithms sell first, ask questions later.
The Play: When a stock sells off on a 143%-growth quarter, the move is a verdict on expectations, not on the business. The AI-backlog story Broadcom told didn’t get worse on Wednesday night — it got better. The thing to watch is whether the knee-jerk reverses once cooler heads read past the headline revenue line to the $16 billion AI guide. Reactions to “priced-for-perfection” prints often look very different a week later. (Research, not advice.)