Analysis
The US IPO market has quietly logged 238 listings in 2026 as of September 7 -- a modest but real 2.15% ahead of the 233 IPOs recorded by the same point in 2025. It's not the number getting attention. That belongs to the handful of AI megadeals -- OpenAI, Anthropic, SpaceX, Canva -- that Wall Street is bracing for, under pressure to prove business models durable enough to scale into a public listing. But a broad-based, modestly-up count of 238 is arguably the more useful data point for reading how open the IPO window actually is beneath the AI headlines.
The gap between the two stories is the point: a 2.15% year-over-year increase in total IPO count says normal-sized companies are getting out the door at roughly the same pace they were a year ago, while the market's attention is almost entirely absorbed by a handful of listings that haven't happened yet. A New York Times report from June indicated OpenAI is now leaning toward 2027 rather than 2026 specifically over valuation concerns -- meaning the single most-watched IPO of the cycle may not even land within this year's count at all.
Oura's S-1, filed September 3 with a five-bank underwriting syndicate targeting a raise of up to $3 billion, is itself a useful signal here: there is real banking capacity available for a well-understood, profitable consumer story right now, even while the industry's collective attention stays fixed on whether the AI labs will list at all. A steady 238-deal count with modest year-over-year growth is what a functioning IPO market looks like when it isn't being driven by a handful of outlier names -- worth remembering the next time a single delayed AI listing gets treated as evidence the whole market is closed.