Generated by Codex with GPT-5

This summary covers The Economist’s May 31st, 2026 Finance & economics Buttonwood column listed in the contents as Giga-IPOs and published under the headline The incredible shrinking stockmarket.

A strong market with fewer public companies

The article argues that America’s stockmarket can look triumphant and diminished at the same time. Major indices are at record highs, but the public market itself is thinner than it used to be. The number of initial public offerings has fallen sharply from the pace of the 1990s, and the count of listed American companies has dropped from about 8,000 in 1996 to roughly 3,900 last year.

That matters because public markets are supposed to let ordinary savers share in the growth of ambitious companies. If fast-growing firms stay private for longer, or avoid listing altogether, much of their most explosive value creation accrues to founders, employees, venture funds and private-market investors before index funds or retail investors can buy in.

The coming flotations of SpaceX, OpenAI and Anthropic might seem to solve this problem. They are among the rare private companies large and important enough to excite the whole market. SpaceX has filed to raise perhaps \$75bn, while OpenAI and Anthropic are preparing their own enormous listings. Once public, all three would probably enter major indices and therefore become part of many ordinary investors’ portfolios.

The article’s central point is that this is not a cure. It is evidence of the disease.

Why the giga-IPOs are not enough

The three companies are exceptional even by technology standards. SpaceX, OpenAI and Anthropic can plausibly claim to sit near the center of the artificial-intelligence and space-economy revolutions. Their expected combined market value is close to \$4trn. By contrast, the next three largest American startups cited by the article, Databricks, Stripe and Anduril, are together worth less than \$400bn.

That scale creates a problem for public investors. Broad stockmarket returns are not driven by the average newly listed company. Most IPOs underperform the wider market over the following years. The reason public-equity investors have historically done well is that a few extraordinary winners compensate for a long tail of disappointments.

Late listing weakens that mechanism. When companies go public while still young and relatively modestly valued, public investors can capture a large share of the upside if those companies become giants. Amazon and Nvidia are the classic examples. Their post-IPO returns have been spectacular because public shareholders were allowed in before the companies became dominant.

SpaceX, OpenAI and Anthropic are arriving at the public-market gate after much of that repricing has already happened. If SpaceX lists at a valuation around \$1.75trn, the article notes, it would need to exceed \$470trn in market value to match Amazon’s post-IPO return. Matching Nvidia’s return would require a number so vast as to be absurd. The point is not that SpaceX is a bad company. It is that even a brilliant company can list too late and too large to offer public investors the old kind of windfall.

Private markets have become too comfortable

The article is skeptical that regulation alone can reverse this shift. American officials want to revive IPOs by easing some disclosure burdens and changing reporting requirements. A friendlier market may encourage smaller firms to list in the wake of the big technology flotations.

But the deeper incentive has changed. Venture capital is abundant, and private secondary markets increasingly let early investors and employees sell shares without forcing a company onto the stock exchange. That reduces one of the old reasons to go public: liquidity. The article cites PitchBook data showing that secondary deals for venture-backed firms nearly doubled in the first quarter of 2026 to \$112bn and, for the first time, exceeded money raised through IPOs.

For founders, staying private can mean less scrutiny, fewer disclosure obligations and more control. For venture investors, it preserves access to the richest phase of value creation. For employees, secondary sales can provide cash before an IPO. Each group has a rational reason to accept a longer private life. The cost is borne by public-market investors, who receive access later and at higher valuations.

The AI boom sharpens the issue. If artificial intelligence produces a handful of companies worth many trillions, public investors will still benefit after those firms list. But they will benefit less than they would have if the companies had gone public earlier. A market in which ordinary investors are offered only mature giants, not young compounding machines, becomes less democratic and less dynamic.

The takeaway is deliberately uncomfortable: giga-IPOs may be exciting events, but they do not prove that America’s public markets are healthy. They show that the most important companies can now grow to staggering size before the public is invited in. Unless that changes, the stockmarket will remain visible, liquid and powerful, but less central to the creation of new wealth than its record highs suggest.