Private Giants Flooding into Public Markets, Private Equity Can Support Valuations but Not Endless Cash Burn
Super private companies that have long remained in the private equity market are now concentrating on selling equity to the public market. In the first half of 2026, traditional IPO fundraising in the U.S. reached approximately $114.2 billion, over seven times the $14.8 billion in the same period of 2025; including follow-on offerings, by the end of June, U.S. IPOs and refinancings totaled about $251 billion, setting a record for the same period.
The volume is supported by a few names. SpaceX went public in June, pricing its valuation at approximately $1.75 trillion to $1.77 trillion, raising about $85.7 billion after exercising the over-allotment option, surpassing Saudi Aramco's 2019 record and exceeding the previous largest U.S. IPO, Facebook, by about 17 times in valuation terms. AI chip company Cerebras raised about $5.5 billion to $5.6 billion, with its opening price rising about 89% from the issue price. SK Hynix raised about $26.5 billion in its U.S. stock issuance. As of late July, there have been about 86 IPOs in the U.S. this year, raising about $251 billion, more than five times the $47.4 billion raised in all of 2025, with SpaceX accounting for about one-third of that.
The queue still consists of private giants. Anthropic secretly submitted its listing documents in June after a private valuation of about $965 billion, aiming for October; OpenAI submitted a confidential S-1 after a private valuation of about $852 billion, targeting a valuation close to $1 trillion, with a timeline fluctuating between the end of this year and 2027. Databricks has pushed its listing to 2027 while raising its valuation from about $134 billion to about $188 billion to $190 billion and continues private fundraising. Companies like Csquare, supported by Brookfield, Inspire Brands under Roark, and Jersey Mike’s supported by Blackstone are also in confidential filings. Anduril, Discord, Kraken, and Strava appear on banks' watch lists.
On the private equity side, there is another wave of "going public." Medline has become one of the largest private equity-backed IPOs in U.S. history, while Europe's Verisure has set a local record for private equity-backed listings. However, according to PitchBook, as of June 30, there are still 33,575 unsold companies in private equity portfolios, up from 32,451 at the end of last year and 15,923 a decade ago. Since 2022, only about 70 private equity-backed companies have gone public in the U.S. In London, the opposite is true: private equity is buying listed companies, with Bodycote at about £1.8 billion and Gamma at about £1.1 billion being subject to takeover bids, and the total amount of UK listed companies being acquired this year is about $182 billion.
The private equity fundraising window is thus yielding. In July, U.S. growth equity fundraising was $20.6 billion, the smallest monthly amount since 2026. The secondary market has previously replaced IPOs when the window closed: in the twelve months leading up to mid-2025, secondary venture capital transactions totaled $61.1 billion, exceeding the $58.8 billion raised by venture-backed IPOs in the same period. Now, large companies are returning pricing power to public book building, while smaller companies remain largely in private and extension funds.
Mechanically, buyers are public funds, indices, and active growth accounts capable of digesting trillion-dollar order books; sellers are venture capital, employee options, and some sovereign funds that have endured until exit years. The event-driven factors are AI and aerospace capital expenditures pushing private companies' cash needs beyond the limits of private equity rounds. Funds are flowing from "internal valuation transfers" to tradable stocks. Beneficiaries are investment banks that can price and allocate, as well as already listed giant growth stocks; those under pressure are ordinary private equity portfolio exits and medium-sized issuers that must compete for orders with SpaceX in the same window. After going public, Cerebras saw its stock price return to near the issue price within a month, indicating that the opening of the window does not mean that all private labels can stand firm in the secondary market.
Source: Public Information
ABAB AI Insight
Delaying IPOs for private companies is not a new strategy; it is a result of cheap funding in the 2010s combined with secondary transfers, employee stock ownership platforms, and extension funds, which have pushed the "IPO clock" from six or seven years to over ten. The combined private valuations of SpaceX, ByteDance, Anthropic, Databricks, and OpenAI once reached about $3 trillion, with clients repeatedly asking the same question: When will they IPO? The answer given for 2026 is not "full market normalization," but rather "only companies too large for private equity to absorb will go public." Harbourvest has framed this as a paradox: the headlines have never been stronger, yet the breadth remains narrow.
Capital pathways are diverging. AI labs need continuous financing for computing power and chips, with single rounds reaching tens of billions, and the depth of the public market exceeds the checks from sovereign funds and growth equity funds. Private equity, however, is stuck with inventory: the exit pace cannot keep up with portfolio aging, with assets over seven years still accounting for about 27%. Thus, two types of "going public" have emerged: super private companies selling stock to the public; and medium-sized listed companies in places like London being bought by private equity. Firms like Blackstone are concentrating realizations on already listed or near-listed assets while pushing new money into data centers and energy, rather than sending all 33,575 portfolio companies to exchanges.
Specific comparisons are necessary. The IPO boom of 2021 was quantity-driven, with many mid-sized tech stocks priced too high and subsequently falling; the situation in 2026 is weight-driven, with SpaceX's single IPO being heavier than all U.S. IPO fundraising in the past two years combined. Saudi Aramco represents a state selling production assets, while SpaceX represents a founder selling the narrative of aerospace, Starlink, and AI to global growth accounts. Medline and Verisure demonstrate that traditional cash cows supported by private equity can also succeed, but under Dealogic's metrics, only about 70 private equity-backed companies have gone public in the U.S. over the past four years, indicating that the pathway is only open to the top tier. The industry is in a phase where "private equity pricing power is returned to public markets, but only to the giants."
Structural judgments belong to capital concentration. The mechanism is: when a single company's financing needs exceed the risk limits and duration of private equity funds, the exchange regains pricing power; when the liquidity discount of medium-sized listed companies exceeds the control premium of private equity, the exchange loses pricing power. The result is not a convergence of public and private markets, but rather a simultaneous extremization at both ends— the largest private companies going public and mediocre public companies going private. Indices and active funds are forced to absorb a few super stocks, while ordinary private equity assets continue to circulate in extension funds and secondary discounts.
ABAB News · Law of Cognition
Private equity can support valuations but cannot sustain endless cash burn.
The window is only open to giants, while it remains closed to inventory companies.
The largest privatization and the largest publicization will occur simultaneously.