Top IPO Scholar on Unprecedented IPO Wave & Why IPOs Underperform the Market | Jay Ritter

Jay Ritter describes the current IPO period as historic even after adjusting for inflation. The SpaceX IPO is approximately twice the size of Japan's 1987 NTT privatization in inflation-adjusted terms, which previously held the record for the largest inflation-adjusted IPO ever. Without SpaceX, Anthropic and OpenAI would likely rank first and second all time; instead they will likely rank second and third. If all three go public in 2026, their combined proceeds as a percentage of US market cap would be in the same ballpark as 1999 to 2000 and 2021, making it the largest IPO wave in history by that measure. The absolute number of companies going public is nowhere near records, with software and biotech seeing few IPOs, partly because AI is perceived as threatening software business models. Ritter draws a parallel to 1999 and 2000, when enormous enthusiasm for internet companies coexisted with a depressed IPO market in industries labeled old economy.

Ritter argues that big IPOs signal market tops with only about 51 percent accuracy, making them unreliable predictors. He notes that Bob Shiller's irrational exuberance comment came in late 1996, more than three years before the US market peaked at a much higher level, as a caution against assuming current enthusiasm is necessarily misplaced.

On average IPOs underperform the market over the three years following a first-day jump, but Ritter stresses this is an equally weighted average. Companies going public with annual revenue of at least $100 million in inflation-adjusted terms do not on average underperform after their IPO, because institutional investors dominate large deals and on average get valuations right by pricing relative to comparable companies. However, companies with significant revenue but very high price-to-sales ratios have on average underperformed after going public. SpaceX went public at a price-to-sales ratio of over 90. To justify a roughly $2 trillion valuation at a price-to-earnings ratio of 20, SpaceX would need approximately $100 billion in annual after-tax profits, a level achieved by only a handful of companies globally. Ritter says SpaceX would need to demonstrate the kind of durable competitive moat seen at Meta, Apple, Alphabet, Microsoft, and NVIDIA, where network effects or expensive proprietary technology allowed ownership of a vertical without competition eroding profits.

SpaceX's prospectus cites a total addressable market of $29 trillion, much of it from AI infrastructure and data centers in space. Ritter is skeptical because data centers in space face substantial technological hurdles and are more speculative than Starlink or Starship. He also notes SpaceX is spending significant money on data centers on earth, a more competitive business with lower profit margins. He characterizes SpaceX as more of a conglomerate than other large tech companies because its AI and Starlink businesses are distinct from each other, whereas OpenAI, Anthropic, and NVIDIA are focused on much narrower businesses. SpaceX acquired Cursor for approximately $60 billion around the time of the IPO, and whether that deal makes SpaceX a serious AI player remains uncertain. Ritter compares it to Facebook's acquisition of Instagram just before its 2012 IPO, which was controversial at the time but proved to be a great investment, while noting the Cursor deal puts SpaceX into a very different business than the launch and Starlink verticals that originally excited investors.

Unlike the internet bubble, companies going public now are those where it is much clearer who the winners are, because companies stay private longer with more venture capital money available. Ritter argues that so much money has flowed into private equity and venture capital that the illiquidity premium is probably close to zero, and venture capital returns are clustered largely in the top decile, so investors who cannot access top funds may be better off in public markets. He identifies volatility washing as a significant problem in private markets, where stale pricing makes volatile assets appear less volatile, and says some investors are willing to overpay for what he calls fake lower volatility. Institutional investors who sue general partners risk being shut out of future investment opportunities, creating a disincentive to litigate, while the introduction of retail capital into private markets may force more rigorous mark-to-market practices.

Ritter says the history of technological change shows the main beneficiaries are consumers and workers rather than owners of capital, citing airline investors as an example of wealth destruction despite air travel being transformative. Competition forces companies incorporating AI to lower product prices, meaning benefits flow to consumers. He acknowledges legitimate concern that AI is disrupting white collar jobs at a pace faster than prior technological transitions but argues that limiting technological change would forgo large productivity gains. He says there are very good reasons to be excited about AI but that determining the right price is much harder, and whether current valuations in US tech have already priced in future growth is the key unanswered question.