Put this week's IPO-adjacent news side by side and a pattern emerges that's easy to miss story by story: 2026's IPO class is really two separate markets wearing the same label. On one side, mega-cap AI-infrastructure listings are trading like high-beta risk assets. SpaceX has round-tripped from a post-IPO peak near $211 back to roughly $123, landing near its $135 offer price just weeks after going public. Csquare, the Brookfield-backed data-center operator, priced its own IPO two dollars below range this week and slipped further on its NYSE debut. Both are getting punished for the same underlying anxiety currently hammering chip stocks: that AI-infrastructure growth assumptions built into 2026 valuations were too aggressive.
On the other side, a steady cluster of biotech S-1 filings kept arriving throughout the same stretch, apparently untouched by any of it. NuvOx Therapeutics and Valion Bio both filed S-1s on July 17, joining earlier July filings from Attovia Therapeutics and Braveheart Bio, each disclosing plans to raise roughly $100 million in their public debuts. None of that pipeline slowed down or showed any sign of repricing around the chip-stock selloff happening in the same news cycle.
“On the other side, a steady cluster of biotech S-1 filings kept arriving throughout the same stretch, apparently untouched by any of it.”
The explanation is straightforward once you look at what actually drives each category's investor base. AI-infrastructure names trade largely on comparable-company multiples and growth-rate assumptions that move in lockstep with sector sentiment -- when Nvidia-adjacent stocks wobble, SpaceX and Csquare wobble too, regardless of their own fundamentals. Biotech IPOs trade on clinical-trial data, regulatory timelines and pipeline maturity, variables that have nothing to do with semiconductor demand forecasts, so the investor base pricing them simply isn't reacting to the same headlines.
For bankers and founders planning 2026-2027 listings, the practical takeaway is to benchmark against the right tape. An AI-infrastructure company should expect its pricing to move with sector sentiment on the week it lists, the way Csquare's just did -- that's now the base case, not a tail risk. A biotech company should expect its pricing to depend almost entirely on its own data and calendar, largely insulated from whatever the SOX or Nasdaq did that week.
What to watch next: whether the next mega-cap AI-infrastructure IPO in the pipeline delays its listing to wait out sector volatility, and whether the biotech S-1 cluster from this month converts into a wave of actual pricings in August, testing whether that tape's apparent insulation holds once several offerings compete for the same investor demand simultaneously.