Analysis
2026 has been the biggest IPO year in a decade by volume -- Q2 alone produced $104.8 billion in US listings, the largest quarter on record, and SpaceX's June debut at a $1.77 trillion valuation was the largest IPO ever completed, according to Dealroom's 2026 IPO tracker. What that volume doesn't tell you is how those newly public companies are actually performing once the debut-day headlines fade, and the early scorecard is a lot messier than the raise numbers suggest.
## The split between debut and delivery Cerebras is the clearest cautionary data point. The AI chipmaker fell roughly 14% after its second earnings report as a public company -- even while raising full-year guidance to $880-890 million in core revenue. Core revenue actually grew 103% year-over-year, a genuinely strong number, but GAAP revenue of $180 million missed the $194 million analysts expected, and the market punished the miss harder than it rewarded the raised guidance. That's a company executing well on the metric that matters long-term (core revenue growth) while still getting marked down on the metric Wall Street reacts to in the moment (GAAP revenue versus consensus).
Reddit sits at the other end of the spectrum. The company is set to join the S&P 500 on August 18, becoming only the second pureplay social media company in the index after Meta -- a milestone that reflects two full years of sustained execution since its 2024 IPO, not a debut-week pop. Index inclusion requires sustained profitability and market-cap thresholds that a hot opening trade alone can never satisfy; it's the clearest available signal of which IPO-class companies have actually built durable public-market businesses versus which ones just had a good first week.
“## The split between debut and delivery Cerebras is the clearest cautionary data point.”
SanDisk offers a third pattern entirely: not a 2026 IPO at all, but an existing public company whose stock is up more than 450% year-to-date on an investor-day growth plan and the same AI-driven memory-chip demand tightening supply across the sector. It's a reminder that 2026's biggest stock moves aren't confined to the newly public cohort -- the AI infrastructure demand story is repricing established chip names just as aggressively as it's pricing fresh listings.
The pattern across all three: initial IPO pricing and post-listing stock performance are only loosely correlated once a few quarters of real trading data exist. A company that primed a strong debut on narrative and demand can still get marked down hard on its first disappointing earnings print, while a company executing well on the metrics that matter can take years to get full market credit through mechanisms like index inclusion.
What the record IPO volume headline misses: a busy IPO market means more companies are getting priced by public investors, but it says nothing about how many of those companies will still be trading above their debut price a year from now -- and 2026's dataset so far is genuinely mixed, not uniformly bullish.
The next real test is whether Anthropic and OpenAI's expected IPOs, both targeting valuations in the trillions, follow SpaceX's post-listing trajectory or Cerebras's -- and given how differently this year's smaller-cap debuts have traded, that outcome is far from decided by the size of the offering alone.