Meta Platforms fell 7.5 percent on Thursday to $644.30, erasing the post-earnings pop that might have followed a quarter that topped estimates on every headline metric. Revenue of $64.6 billion beat the $63.37 billion consensus, driven by a 10 percent rise in ad impressions and a 14 percent increase in average ad price. Earnings per share came in at $3.83, though the company did not clarify whether that figure is directly comparable to the $5.75 analysts had modeled. The family of apps hit 3.64 billion daily users, and Threads now claims more monthly active users than X.
The capex number that spooked the room
The slide coincided with a revised capital expenditure outlook. Meta lifted the floor of its full-year capex range to $135 billion from $130 billion, keeping the ceiling at $145 billion. That incremental $5 billion at the low end represents the minimum additional outlay the company is now committing to, on top of the hundreds of billions already deployed since the AI pivot began roughly eighteen months ago. Investors who tolerated the spending ramp when it was framed as infrastructure build-out are now being asked to fund a consumer-facing rollout of Muse, the AI agent Meta hopes will turn coding automation traction into a mass-market revenue stream.
A new currency for the next phase
To underwrite that ambition, the board authorized a Class C non-voting share class. Existing holders will receive one Class C share for each share owned, effectively a two-for-one split, with trading expected to start in January. The company said the new stock could be used to raise capital, fund acquisitions, and cover employee compensation over a multi-year horizon, potentially unlocking hundreds of billions in financing capacity. The structure preserves voting control while giving Meta a liquid instrument it can issue without diluting governance, a mechanism that looks suspiciously like a permanent acquisition currency.
The bet is now on consumer AI
Zuckerberg framed the spending increase as a vote of confidence in Muse’s early progress, particularly its adoption among developers. The strategy has shifted from proving the models work to proving they can sell. That is a different, harder question. The market’s reaction suggests it is not yet convinced the consumer pivot will generate returns commensurate with the capital intensity required. The quarter was strong. The guidance was expensive. The new shares are a blank check. Whether Muse becomes a profit engine or a very expensive science project is the only variable that now matters.