Will AI be the giant waiting in line to go public, marking the "last hurrah" of the U.S. stock market?

Article by: Dong Jing

Source: Wall Street Insights

A feast comparable to the peak of the internet bubble IPO frenzy is taking shape. OpenAI, Anthropic, and SpaceX, three AI giants, are racing to go public, each aiming for a valuation of over one trillion dollars, with a combined scale enough to reshape the U.S. stock market landscape. This unprecedented wave of IPOs is not only the ultimate stress test for AI investment logic but also the biggest variable influencing the trend of risk assets this year.

On May 22, according to Wall Street Insights, OpenAI has prepared to secretly submit an IPO application to regulators, with the earliest possible listing in September this year, targeting a valuation exceeding $1 trillion, aiming to raise about $60 billion, surpassing Saudi Aramco’s 2019 IPO record of $25.6 billion by more than double.

Meanwhile, competitor Anthropic is also advancing its own listing plans and disclosed that its second-quarter revenue is expected to double sequentially to $10.9 billion, potentially achieving its first quarterly operational profit. Deutsche Bank’s research report pointed out that the way these two IPOs are executed "could very likely become a major swing factor in risk asset trends this year," making them key macro themes to watch closely.

However, beneath these glamorous valuations, the financial fundamentals of the two companies are starkly different. OpenAI’s first-quarter revenue was $5.7 billion, but its adjusted operating profit margin was -122%, meaning it loses $1.22 for every dollar of revenue generated. It is expected to achieve positive cash flow only around 2029 to 2030. Anthropic’s revenue in the same period was $4.8 billion, with second-quarter revenue expected to jump to $10.9 billion, and it is projected to achieve an operating profit of approximately $559 million, having already crossed the profitability threshold.

Analysis indicates that although these two companies are competing on the same stage, they exhibit fundamentally different business logics, presenting a rare dilemma for public market investors.

The Largest IPO in History: How Shocking Are the Numbers?

Deutsche Bank’s research report states that whether it’s OpenAI or Anthropic, each IPO will raise more than twice the amount of Saudi Aramco’s 2019 IPO, even after adjusting for inflation, making them easily the largest IPOs in history.

In another report, Deutsche Bank noted that if OpenAI achieves its target valuation of over $1 trillion, it will become the 14th largest company globally by market capitalization, ranking just behind Berkshire Hathaway and surpassing Eli Lilly.

In comparison, Berkshire Hathaway last year had revenue of over $370 billion and net profit of $67 billion; Eli Lilly’s sales exceeded $65 billion with a profit of $21 billion. OpenAI, on the other hand, is not yet profitable, with annualized revenue around $30 billion and only a few thousand employees.

From the market capacity perspective, Deutsche Bank believes that the current total market value of U.S. stocks is about $70 trillion, five times the peak of the internet bubble, indicating a market absorption capacity far stronger than in the late 1990s.

At that time, nearly 500 companies IPO’d annually, whereas in this decade, the average is only about 120, and today’s listed companies are generally more mature.

Moreover, a single IPO of $60 billion is only slightly below the total U.S. IPO fundraising in 1999 and 2000 (each around $65 billion), which is about half of the record $119 billion raised in 2021.

The “Siphon Effect” of Giants and the Massive Shift of Passive Funds

As these giants go public, their draining effect on liquidity in the U.S. stock market has raised high alert on Wall Street.

The clustering of SpaceX, OpenAI, and Anthropic IPOs, combined with the Nasdaq’s newly introduced “Fast Track” inclusion mechanism, is brewing an unprecedented massive shift of passive funds, creating an “AI giant siphon effect.”

According to Wall Street Insights, JPMorgan estimates that if SpaceX’s target valuation reaches $2 trillion and 50% of its shares are floated, passive funds will have to sell about $95 billion of their holdings in the eight major tech stocks on Wall Street (Nvidia, Apple, Microsoft, Amazon, Google, Broadcom, Meta, Tesla) to make room for the new shares.

Strategas’s chief ETF strategist Todd Sohn pointed out that since IPO initial float is usually only about 5%, and ETFs track assets worth trillions of dollars, this extreme supply-demand imbalance will make the index inclusion process “somewhat crazy,” forcing passive investors to buy at high prices.

Valérie Noël, trading director at Syz Group, said that the market has already begun betting on a downward pressure on existing large-cap stocks.

According to disclosures on March 28 this year, OpenAI’s going public will be a substantial referendum on the entire AI investment logic. The data shows that OpenAI’s revenue in 2025 will reach $13.1 billion, but it is expected to incur a net loss of $14 billion in 2026.

Meanwhile, OpenAI has committed to investing about $1.4 trillion in infrastructure before 2033. If S&P Global, FTSE Russell, and Nasdaq adopt fast inclusion rules, passive funds may be forced to buy between $24 billion and $48 billion immediately after listing.

Faced with such a massive capital restructuring, ordinary investors’ portfolios will be passively reshaped regardless of whether they are active or passive.

Deutsche Bank’s research states that the way these IPOs are executed will be a major swing factor for risk assets this year. PitchBook’s analysis is more straightforward:

The private equity market is experiencing a “systemic quality inversion”—the companies with the highest valuations score the lowest on actual business quality metrics when priced in the public market.

For ordinary index fund or ETF investors, this game is hard to avoid: whether active or passive, their portfolios will be passively reshaped as index rules change.

For active investors, when S-1 filings are public and all financial secrets are laid bare, the market will face a clear choice: trust a company that has already found a profitable model, or a giant requesting more years and billions to explore profitability?

The answer will determine whether this carnival is the start of a new cycle or the last dance before the feast ends.

Ice and Fire: Anthropic Profitable, OpenAI Massive Losses

Despite the valuations soaring, the financial situations of these two leading AI companies are worlds apart. Anthropic has already started turning a profit, challenging the traditional notion that huge AI expenditures hinder short-term profitability.

According to Wall Street Insights, on Wednesday, The Wall Street Journal reported that Anthropic’s second-quarter revenue is expected to grow more than double to $10.9 billion, with an operating profit of about $559 million.

Anthropic’s gross margin has jumped from 38% to over 70%. Its CEO Dario Amodei jokingly said that revenue growth has become “too hard to handle.”

The company’s success is mainly due to explosive demand from enterprise clients for its programming tools, with about 85% of revenue coming from corporate and developer customers. This model has clear willingness to pay and lower service costs.

In contrast, OpenAI is still losing money.

Wall Street Insights mentioned that data shows OpenAI’s first-quarter revenue was $5.7 billion, but its adjusted operating profit margin was -122%, meaning it loses $1.22 for every dollar earned.

About 85% of OpenAI’s revenue is related to ChatGPT consumer subscriptions. Despite having 55 million paying users, it has over 900 million weekly active users, with a huge free user base creating a significant inference cost black hole.

OpenAI expects to achieve positive cash flow only around 2029 or 2030. Its CEO Sam Altman and head of application business Fidji Simo are trying to shift focus to commercially profitable clients.

In IPO narratives, the two tell very different stories. Anthropic, with verified quarterly profits, can be compared to Salesforce or ServiceNow—a logic of an enterprise software company.

OpenAI, on the other hand, needs to convince the market that AI agents, image generation, and even advertising will eventually turn massive consumer traffic into profits.

In Sam Altman’s plan, by 2030, ChatGPT’s advertising business could generate about $102 billion in revenue, but this will take time, and time is the most scarce resource for OpenAI when using losses to fuel growth.

Are AI Giants’ IPOs Essentially “Passing the Hot Potato” to Retail Investors?

Wall Street Insights notes that this wave of AI giant IPOs, in Joachim Klement’s view, is essentially a “risk transfer,” a large-scale cash-out by early investors passing the risk to retail investors, pension funds, and other institutions.

He believes that OpenAI, Anthropic, and others are choosing to accelerate their IPOs amid high investor sentiment to cash out at high valuations before the hype subsides. Early institutional investors can fully exit in the public market, while retail investors and pension funds will face the risk of financial logic ultimately returning to reality.

He directly characterizes this process as “massively shifting investment risk from current holders to those willing to buy into the story.”

Klement cites Greenspan’s 1996 warning of “irrational exuberance”—three years before the bubble burst. He judges that AI hype could continue into 2026, with large cloud computing firms unlikely to cut investments; but “impossible math” will eventually return to reality, perhaps not in 2026, but possibly in 2027 or 2028.