The window blows open

For most of the past decade the best companies stayed private for longer than anyone thought sensible. Venture capital, private equity and sovereign wealth funds supplied money in quantities that once required a stockmarket listing, and founders avoided the scrutiny that came with one. The number of companies listed in America fell by somewhere between 38% and half from its peak in 1996, depending on the count. In 2026 the tide turned, spectacularly.

Through the end of September, 83 traditional initial public offerings (IPOs) in America raised $120.2bn, according to PwC, an accounting firm, against $29.3bn in the same period of 2025. Counting cross-listings, Renaissance Capital, a research firm, tallied more than $146bn raised by mid-September, past the full-year record of $142.4bn set in 2021. The second quarter alone raised about $105bn, the biggest quarter on record.

202529.3

2026 excluding SpaceX~45.2

2026120.2

04080120

Source: PwC

One giant leap

The headline numbers flatter the market, because one deal dominates them. SpaceX’s listing in June raised roughly $75bn, the largest IPO ever, at a valuation reported at about $1.77trn. Excluding SpaceX, PwC counts about $45bn raised in 2026, which is still more than in the whole of 2025 but a long way from a record. Other big deals included Cerebras, a maker of AI chips, which raised $5.5bn in May, and a $26.5bn cross-listing by SK Hynix, a South Korean memory-chip maker.

More may follow. OpenAI said on June 8th that it had confidentially submitted a draft registration statement to the SEC, after a funding round in March reportedly valued it at about $852bn. Anthropic reportedly made a confidential filing on June 1st. Neither has set a date, and a confidential filing is no promise of a listing. Databricks, a data-software firm, has reportedly pushed its own plans into 2027.

Why now?

Three things changed. First, prices: with the S&P 500 at record highs and investors hungry for anything linked to artificial intelligence, valuations in public markets caught up with, and in some cases overtook, those in private ones. Second, size: the biggest private companies have grown so large that even sovereign funds struggle to keep writing cheques, and early investors and employees want to sell. Third, regulation is moving the same way: in May the SEC proposed letting companies report semi-annually rather than every quarter, a change that drew a record 200,000 comments and has not yet been adopted.

The catch

Investors who bought this year’s IPOs have not done especially well. PwC calculates that shares listed in 2026 had returned about 2% on average by the end of September, against roughly 12% for the S&P 500 over the same span. Big listings soak up money that would otherwise go to shares already trading, and a wave of selling by insiders when lock-up periods expire is a familiar drag on returns. Research on the dot-com bust found that the expiry of IPO lock-ups in 2000, which suddenly increased the supply of shares, helped time the collapse of internet stocks.

There is also a question about what the new listings do to the indices that most savers own. A handful of giant newcomers could, once included in benchmarks, push the market’s concentration in a few technology-related firms even higher. The ten biggest companies already make up about 39% of the S&P 500.

Still, the return of the IPO is good news for public markets. For years critics warned that ordinary investors were locked out of the fastest-growing companies, whose biggest gains happened before they ever listed. Some of those companies are now coming to market. Whether buyers pay too much for the privilege is, as ever, up to them.