From SpaceX’s breaking of its IPO price to OpenAI delaying its listing: has a turning point arrived for the tech IPO market?

On July 24, 2026 (Beijing time), SpaceX (SPCX) closed at $118.24, up 2.59% on the day. This seemingly mild rebound—almost unimaginable just a month and a half ago—followed June 12, when SpaceX made its Nasdaq debut with an offering price of $135. The fund-raising size (including the over-allotment option) reached a record of about $86 billion. On the third trading day after listing, the stock surged to an all-time high of $225.64.

Then came a steady pullback. On July 15, SpaceX first fell below its offering price during trading hours. Even though it rose 2.59% on July 24, the stock is still down about 47.6% from its all-time high and down about 12.4% from its offering price.

SpaceX is not an isolated case. The “hot IPO then fading” pattern among mega-IPOs is reshaping the market’s overall assessment of the prospects for tech unicorns going public.

IPO market sentiment shift: from “chasing trends” to “doing the math”

In recent years, the capital markets’ pricing logic for tech companies has revolved around one core keyword: narrative. The explosion of AI large models, the commercial rollout of commercial space, the replacement logic behind new energy, and domestic substitution in deep tech—these sweeping narratives have supported a sustained expansion of primary-market valuations. A “2026 Global Unicorn List” released by Hurun Research Institute in June 2026 shows that the number of unicorn companies worldwide reached 1,603. Total value rose 43% year over year to $8 trillion. Among them, AI unicorns reached 215—adding 87 net new companies within a year—and accounted for 36% of the total value.

But in the second half of 2026, the market’s attention is shifting elsewhere. Data compiled by Zhitong Finance and Economics shows that, as of July 15, the weighted average return for 2026 U.S. IPO companies had fallen to 6%, lagging the S&P 500’s 11% gain over the same period. Weighted average returns for biotech and pharmaceutical company IPOs reached 55%, while after excluding SPACs and other financial instruments, the overall weighted average return for the U.S. IPO market was a loss of 4.4%. The market is not cooling across the board; instead, there is a clear structural divergence—segments with real profit support are still being chased, while sectors priced based on future expectations face repricing.

Behind this shift is investors refocusing on three core metrics: profitability, cash flow, and debt levels.

SpaceX’s valuation dilemma: Starlink keeps the lights on, Starship burns cash

SpaceX’s stock performance exposes the fragility of “narrative-driven” valuations.

From fundamentals, SpaceX shows a typical “two-polarization” pattern. Starlink is the company’s only profitable segment. In 2025, revenue was $11.387 billion, and operating profit was $4.423 billion. Its global satellite communications monopoly advantage is difficult to shake in the near term. But Starship R&D and xAI’s artificial intelligence business continue to burn massive amounts of cash. In 2025, SpaceX posted a total net loss of $4.937 billion, and losses expanded further in 2026 Q1.

In the early days after listing, the market gave SpaceX an extremely high valuation for its “space vision” and the “Musk premium.” At its peak, the company’s market cap briefly topped $2 trillion. But after the bubble burst and the hype faded, fundamentals began to take back pricing power. Even after a large pullback, SpaceX’s forward P/S ratio is still about 46x, significantly higher than other tech stocks.

A more worrying signal comes from the bond market. SpaceX’s 30-year bond yield rose from 6.7% at issuance to 7.4%, edging toward “high-yield junk bond” levels; meanwhile, credit default swap (CDS) spreads widened to 158 basis points, up sharply from 110 basis points at the end of June. Bond investors typically perceive risk earlier than stock investors—when the bond market starts using “junk bond” pricing logic to measure a company, the pressure on its equity valuation is self-evident.

Short-selling forces are also gathering speed. According to S3 Partners, SpaceX’s short position is about 185 million to 196 million shares, representing 29% to 31% of the float—an enormous increase from about 40 million shares roughly three weeks earlier. As of July 24, short-sellers’ paper gains had reached about $15.5 billion.

The August lockup-unwind wave will be another major test. The first insiders and employees’ shareholdings that meet the release conditions (together up to 911.5 million shares) will enter the market—equivalent to more than 1.4 times the entire IPO float. The combination of a sharp expansion in float and persistent short-selling pressure means any valuation recovery for SpaceX will be a long process.

Ripple effects: OpenAI delayed, AI unicorns under pressure

SpaceX’s trajectory is generating broad ripple effects.

Most directly affected are other AI heavyweights that are also in the IPO preparation stage. OpenAI originally planned to list in the second half of 2026, but after SpaceX traded below its offering price, the company leaned toward postponing its IPO to 2027. Management continues to insist on a $1 trillion valuation target and is unwilling to lower expectations in exchange for a faster listing. Wall Street investment banks generally judge that model-driven enterprise IPOs originally expected to start in the second half of 2026 may be pushed back to the first half of 2027 as market risk appetite declines.

Anthropic’s situation is more nuanced. After completing its Series H funding round in May 2026, the company’s valuation reached $965 billion, and it has already secretly submitted an IPO application, potentially as early as October. But SpaceX’s IPO performance has set an unideal reference point for Anthropic’s pricing. If a company raising $86 billion and owning a profitable segment like Starlink can still trade below its offering price, how should an unprofitable AI company be valued?

Secondary-market valuation signals are also not encouraging. A database maintained by Stanford University professor Ilya Strebetskaya shows that, as of May 2026, among 1,900 unicorn companies, 332 had raised funding at valuations no higher than their peaks—of which 212 valuations had already fallen below the $1 billion threshold. As many as 383 companies disclosed no new funding plans in the past 3 years. PitchBook estimates that as these companies reprice or go bankrupt, the total net reduction in valuations would add up to $500 billion to $1 trillion.

“Zombie unicorns” are emerging in large numbers—companies that were overvalued in the primary market, unable to list and unable to be acquired, and struggling to continue raising funding are becoming the new normal in Silicon Valley.

From “growth stories” to “fundamentals”: a paradigm shift in IPO pricing logic

SpaceX’s IPO path reveals a deeper trend: capital markets are shifting from “growth-story investing” to “fundamentals investing.”

Over the past decade, a low interest-rate environment and abundant liquidity encouraged investors to pay premiums for “potential market size” and “future cash flow.” AI, the space economy, new energy, and deep tech—these narratives were compelling enough to sustain the continued expansion of primary-market valuations. The 1,603 unicorns and $8 trillion total valuation in the Hurun list, to a large extent, are products of this loose environment.

But the environment is changing. Expectations of Fed rate hikes, a sharp surge in tech giants’ capital expenditures, and uncertainty around the pace of AI commercialization collectively drive a retreat in risk appetite. Investors are starting to ask a basic question: when will this company start making money?

This question is particularly sharp for tech unicorns still in the “burn cash to buy growth” stage. Even SpaceX—with a profitable segment like Starlink—has traded below its offering price. AI startups with smaller revenue scales and more severe losses will face even tougher valuation tests.

From a more macro perspective, the IPO market is undergoing a deep restructuring of its valuation framework. In recent years, there has been a notable valuation inversion between primary and secondary markets—valuations in private markets often exceed the pricing ability of public markets. SpaceX trading below its offering price, SK Hynix giving back gains, and OpenAI postponing its listing all point to the same direction: the secondary market is recalibrating the value of tech companies with stricter standards.

Conclusion

SpaceX’s roller-coaster ride from $225.64 to $118.24 is not just one company’s stock volatility. It is a snapshot of a shift in capital-market pricing logic—when the era of abundant liquidity and narrative dominance gradually fades, profitability, cash flow, and debt levels become the core reference coordinates of valuation.

For tech unicorns preparing to go public, this means a harsher valuation environment. OpenAI chooses to delay rather than cut its offering price, and Anthropic advances its IPO process on the basis of a $965 billion valuation—these decisions will be tested by the market in the coming months.

For investors, the IPO market’s paradigm shift means expected returns from IPO subscription strategies need to be recalibrated. The old IPO logic of “getting the shares means you profit” is being replaced by more complex fundamental analysis.

In 2026’s tech IPO market, a profound transformation from “stories” to “numbers” is underway. SpaceX is just the most eye-catching example of this shift.

FAQ

Q: Why did SpaceX’s share price fall sharply after its IPO?

The core reason is a disconnect between valuation and fundamentals. In the early days after listing, the market set an extremely high valuation based on its “space vision” and the “Musk premium.” However, besides Starlink, SpaceX’s Starship and xAI businesses have continued to post massive losses. Net loss in 2025 was $4.937 billion. As hype fades, the large-scale lockup unwind ahead in August approaches, and short-selling pressure builds up—so the stock price returns to valuation based on fundamentals.

Q: Why did OpenAI delay its IPO plans?

OpenAI originally planned to list in the second half of 2026, but after SpaceX traded below its offering price and market risk appetite declined, and the AI segment cooled, the company leaned toward postponing its IPO to 2027. Management continues to insist on a $1 trillion valuation target and is unwilling to lower the offering price to list.

Q: What is a “zombie unicorn”?

It refers to unicorn companies that secured a high valuation in the primary market but get stuck in difficult conditions because they cannot go public, cannot be acquired, and find it hard to continue raising funding. As of May 2026, about 212 unicorn valuations have already fallen below the $1 billion threshold.

Q: What does the shift in the IPO market from “growth-story investing” to “fundamentals investing” mean?

It means that market pricing logic shifts from “future potential market size” to “current profitability, cash flow, and debt levels.” Investors are no longer willing to pay high premiums for narratives that have not been realized, and instead require companies to demonstrate a real path to profitable operations.