Nasdaq drops 1.25% after the FT puts OpenAI's revenue run rate $20 billion short

Set side by side, the two newspaper figures describe a delay. Earlier expectations had put OpenAI's September run rate at $70 billion ^[19], so the FT's number is a shortfall of roughly 29% against what had been expected ^[20]. Bloomberg's year-end figure is that same $70 billion ^[10]. If both accounts are accurate, the expected number arrives about one quarter late ^[18], and getting there means the run rate compounding at roughly 11.9% a month through December ^[16].

Dow Jones Newswires, in a story carried by Morningstar, linked the fall in AI-linked stocks to the same FT report ^[3]. The selling was concentrated. The Dow closed up 0.10% and the S&P 500 fell 0.47% ^[9], leaving the Composite 1.35 percentage points behind the Dow for the session ^[17].

Crypto Briefing described sentiment swinging from worry to relief within roughly 24 hours on the strength of two press reports ^[11]. OpenAI has no public shares ^[14], so figures like these reach public holders secondhand. One round trip establishes that a reported $20 billion gap in a private company's run rate can move the Composite by more than a percent ^[3][8]. Whether a smaller or vaguer number would do the same is not something a single session can show.

The session can be read three ways. In the first, Bloomberg's number holds, and the market sold a one-quarter delay as if it were a permanent $20 billion shortfall. In the second, the run rate stalls near $50 billion, and the drop was the first repricing of chipmakers, cloud providers and AI funds. Crypto Briefing noted that all three trade partly on the assumption that AI spending keeps growing ^[14]. In the third, oil and interest rates did part of the selling. Oil pulled back on October 9 from highs set during the selloff session, by Dow Jones's dating ^[7]. Front-month Brent fell 1% to $103.20 after President Trump said the U.S. wouldn't attack Iran before the midterms, citing "productive" talks with Tehran ^[6]. Ten-year Treasury yields were at 5.244% that morning, having hit multidecade highs earlier in October ^[13].

I'd put most weight on the first reading, with some of the third mixed in. A Dow that finished higher on the same day ^[9] fits a sale of one sector better than a market-wide oil or rates shock. The counter-case is that the compounding Bloomberg's figure implies is steep, three straight months near 12% ^[16]. Crypto Briefing also pointed out that the recovery depended on a forecast that has not yet been delivered ^[11]. If AI names turn out to have tracked oil and yields through that session more closely than they tracked the FT story, the attribution weakens, and this view with it.