By Jonathan Levin The US may yet notch a record year in initial public offerings if Anthropic PBC goes public soon. But it would be a false dawn, the numbers flattered by the trading debuts of Space Exploration Technologies Corp. and the aforementioned maker of the Claude chatbot. Under the surface, America’s withered IPO market is crying out for revitalization following a decades-long structural shift toward private funding that’s curbed transparency and walled off access to the best investment opportunities.

When companies live out their fastest growth years in private, the profits accrue primarily to founders and country-club capitalists. Public investors don’t get their first opportunities to buy shares until the companies are somewhat mature and expectations are inflated, as evidenced by the underperformance of recent crops of large IPOs . That’s a key concern with the offerings still looming from Anthropic and OpenAI, which aim to join SpaceX in the over-$1 trillion public valuation club. To reverse the trend, the US should demand more disclosure from massive private firms to remove the perverse incentive they currently have to hide in the shadows for as long as possible.

When companies live out their fastest growth years in private, the profits accrue primarily to founders and country-club capitalists. Public investors don’t get their first opportunities to buy shares until the companies are somewhat mature and expectations are inflated, as evidenced by the underperformance of recent crops of large IPOs . That’s a key concern with the offerings still looming from Anthropic and OpenAI, which aim to join SpaceX in the over-$1 trillion public valuation club. To reverse the trend, the US should demand more disclosure from massive private firms to remove the perverse incentive they currently have to hide in the shadows for as long as possible.

Look beneath the headline number of $132 billion raised year-to-date and the US still has too few companies going public. Since the SpaceX listing in June, only about 22 IPOs have taken place in the third quarter, on track for one of the weakest quarters in recent years . Adding to the disappointment is a wave of IPO cancellations or delays this month, which reportedly include nuclear power services company Holtec Nuclear Corp., SoftBank-backed SB Energy Inc. and home insurance firm Bamboo Insurance Services Inc. Anthropic is expected to debut in the fourth quarter and OpenAI in 2027; yet unanswered AI safety concerns still have the potential to postpone those plans further. The big picture is unmistakable: America had significantly more IPOs in the 1990s than in the entire quarter century that’s followed.

The fate of IPOs matters because the public market is the best venue for matching great companies in need of funding with investors trying to save and expand their wealth. It has the strongest standard of disclosures, the best liquidity and no velvet rope to keep the general public out. When retail investors are robbed of the opportunity to buy stock in young, innovative companies, inequality widens, and society receives too little information about the firms dominating our economy.

Paul Atkins, President Donald Trump’s Securities and Exchange Commission chair, has his own diagnosis of the problem: He blames “regulatory adventurism” for making public life onerous for companies, and wants to cut the costs he associates with that burden, including reducing certain audit requirements and allowing companies to report semi-annually instead of quarterly.

He’s right that the recurring audits can be prohibitively expensive to smaller issuers, and that quarterly reporting is time-consuming. But Atkins’ proposed solutions amount to a poor tradeoff: The marginal increase in such IPOs wouldn’t compensate investors for lost disclosure and the dismantling of a layer of auditing protection that was instituted in response to the Enron and WorldCom frauds . As much as companies dislike audits, investors love them even more.

For the most exciting young companies in the innovation economy, the costs Atkins is targeting — which total in the hundreds of thousands or low millions of dollars per firm — aren’t really the main issue at all; those firms stay private to guard business secrets from competitors and the public.

Deregulationists point their fingers at that auditing reform — part of the broader Sarbanes-Oxley Act of 2002 — for perceived regulatory overreach. The more important development was a 1996 law that made it dramatically easier to raise private capital across the country, irrespective of differences in state blue sky laws, the overlapping securities rules which used to generate delays, paperwork demands and attorneys’ fees. Once founders and venture capitalists figured out how to raise billions while operating in the shadows, the game was forever changed. Management and early investors decided that the benefits of accessing public capital weren’t enough to offset the costs of having to disclose information about their businesses to the world. That reform, the National Securities Markets Improvement Act, created the conditions for companies to stay private for years, routinely waiting to IPO only once valuations had eclipsed $10 billion or more. Facebook (now Meta Platforms Inc.) would go public in 2012 at a market capitalization of $81 billion and SpaceX would do so in 2026 at $1.8 trillion.

Though Atkins recognizes the problem, his “Make IPOs Great Again” agenda gets the remedy exactly backward. Rather than reducing public market transparency, the right solution would bring more candor to private markets. One study by researchers Cyrus Aghamolla and Richard Thakor found that, after a 2007 legal reform that increased mandatory disclosures of clinical trials, biotech firms became more likely to go public. As my Bloomberg Opinion colleague Nir Kaissar and I have both written, we should obligate private companies above a certain size to start disclosing financial results.

Such a reform may need an act of Congress, which would be hampered by the powerful companies that enjoy notable advantages from the current setup. Privacy allows AI firms, for instance, to conceal their competitive positions. There may even be geopolitical concerns about mandatory disclosures, since AI prowess is being wielded as a form of national power.

One solution would require timely disclosures to relevant authorities for companies with valuations of $10 billion or higher, while making the results public with a one-year lag. That would give the public a taste of what the private behemoths are doing and curb the stay-private incentive, but also protect market share and other competitive secrets until they are effectively stale to competitors in, say, China. Such a delay would look similar to the 18-month lag on disclosure of patent applications, which exists for similar reasons.

The cost of failing to treat this problem correctly will be borne by the next generation of investors. The S&P 500 Index’s extraordinary returns over the past two decades were driven by the likes of Apple Inc., Microsoft Corp., Amazon.com Inc. and Nvidia Corp. — companies that went public from 1980 to 1999 at vastly smaller valuations even in 2026 dollars. Lawmakers and regulators must ensure that the next crop of innovation superstars grows up in the public eye, and that they enter public life early enough for regular investors to share in their success.

To borrow and adapt a term from Atkins, deregulatory “adventurism” won’t solve the problem at hand, but a concerted effort to increase oversight in shadowy private markets will. To contact the author of this story: at jlevin20@bloomberg.net