Is the U.S. stock market facing a "PetroChina moment" with OpenAI's listing?
Author | Ding Mao
Editor | Zhang Fan
Since the beginning of this year, as the valuations of US tech giants in the stock market have repeatedly hit new highs, the market's concerns about the overheating signs of AI have been increasing. Recently, the rumor of OpenAI's massive IPO has pushed these concerns to the peak.
According to Reuters, OpenAI may submit its listing application as early as around 2026, with a target to raise more than $60 billion, and its valuation could reach up to $1 trillion. If this comes true, OpenAI will become one of the largest IPOs in the history of the US stock market.
While the market is cheering for this "AI gold rush", an undeniable law is ringing the alarm: "Gigantic IPOs" in history have often been accompanied by the "curse" of the market reaching its peak. This forces investors to calmly examine the current valuations of US stocks and be vigilant against the reappearance of the ghost of the 2000 Internet bubble.
So, do "gigantic IPOs" in history really indicate a bubble market? Does an AI bubble exist in the current US stock market?
The "Curse" of "Gigantic IPOs"
From a historical review, the emergence of "gigantic IPOs" does have a certain risk warning effect on the market. The logic behind this is mainly reflected in two aspects:
On the one hand, the most direct impact is that "gigantic IPOs" will have a strong siphon effect, quickly draining the blood from other sectors. When the market's liquid funds flow towards the gigantic companies, it is easy to cause a structural "blood loss" in the market, thus intensifying the overall market volatility.
On the other hand, the deeper reason is the "emotional over - draw" at the end of a bull market. "Gigantic IPOs" usually occur during periods of extremely optimistic market sentiment. During such periods, investors have a high risk appetite and a good tolerance for the company's fundamentals, and are willing to pay in advance for the company's growth expectations in the next few years or even more than a decade. Periods with these characteristics are usually at the end of a bull market, and the emergence of a massive IPO may exacerbate the depletion of market funds and trigger a reversal of market sentiment.
The most typical example is the famous "PetroChina curse" in the A - share market in 2007.
On November 5, 2007, PetroChina went public, raising 66.8 billion yuan, accounting for 0.23% of the total market value of the A - share market at that time. Its market value on the first day reached 8.05 trillion yuan, accounting for more than 25% of the total market value of the A - share market at that time.
The funds frozen for its subscription at the end of October were as high as 3.3 trillion yuan, while the residents' bank deposits that year were less than 17 trillion yuan, which was enough to show the almost crazy pursuit of the market for it.
In fact, before this, two giants, China Construction Bank and Shenhua Group, had successively listed on the A - share market, raising 58 billion and 66 billion yuan respectively. The total fundraising scale of the three giants reached 191.4 billion yuan, accounting for 43% of the total IPO amount that year.
The concentrated listing of these giants coincided with a critical turning point in the macro - environment. In 2008, the overseas sub - prime mortgage crisis broke out, and the domestic monetary policy tightened. Market sentiment began to reverse, and giants like PetroChina absorbed a huge amount of funds in the short term, directly accelerating the withdrawal of market funds.
Against this background, the Shanghai Composite Index plummeted from its historical high of 6124 points to 1664 points, a decline of more than 70%. The stock prices of giants like PetroChina tumbled, causing heavy losses to investors who bought at the high level.
Chart: The A - share market crashed due to the concentrated listing of gigantic IPOs in 2007. Data source: Wind, compiled by 36Kr
The IPO of Guotai Junan in 2015 once again confirmed this curse.
On June 26, Guotai Junan went public, raising 30.058 billion yuan. Although its scale was smaller than that of PetroChina, its listing time coincided with the peak of a new round of bull market in the A - share market. Under the extremely inflated market sentiment, its huge fundraising still caused a liquidity drain on the fragile market, accelerating the decline of the market index.
In the month of its listing, the market index reached its peak of 5178 points and then fell all the way. By early 2016, the decline was nearly 50%. At the same time, after Guotai Junan's stock price reached a peak of 37.30 yuan on July 1, it never exceeded that high point again.
Chart: The A - share market crashed after the listing of Guotai Junan in 2015. Data source: Wind, compiled by 36Kr
Compared with the A - share market, the "gigantic curse" in the US stock market seems less significant. This is mainly because the US stock market has a larger market value scale, more active trading, is dominated by institutional investors, and has a higher degree of market - oriented pricing. Therefore, the market can digest large - scale IPOs and reduce their impact on the market index.
However, from a historical review, massive IPOs still have a structural warning effect. The listing of Rivian Automotive in 2021 is a good example.
As a new star in the electric vehicle industry, Rivian raised more than $13.7 billion. After its opening, its stock price soared, and its market value once exceeded $100 billion, surpassing traditional giants such as General Motors. In contrast, Tesla only raised $226 million during its IPO, and its valuation was less than $2 billion, which undoubtedly reflects the unprecedented pursuit of electric vehicles by capital at that time.
Looking at the background, after the pandemic in 2020, during the Fed's super - interest - rate - cut cycle, the market was flooded with a large amount of liquidity, which boosted the preference for technology growth stocks. Against this background, growth stocks like Rivian, which had not yet made a profit but had broad prospects, were given the expectation of being "the next Tesla".
However, at the end of 2021, the Fed accelerated its tapering and hinted at interest - rate hikes. Market sentiment reversed, and high - valuation and high - risk technology growth stocks were the first to be sold. Less than a week after its listing, Rivian's stock price tumbled continuously, and its market value evaporated by tens of billions of dollars.
Rivian's sharp decline was also regarded by the market as a signal of the bursting of the technology growth bubble, which accelerated the structural adjustment of funds and led to a correction in the overall Nasdaq.
Chart: The adjustment of the Nasdaq after Rivian's listing. Data source: Wind, compiled by 36Kr
The Essence of the Curse is Emotional Over - draw
In summary, there is indeed a "gigantic curse" to a certain extent in both the A - share and US stock markets. However, in fact, a massive IPO warns not about the company itself, but about the extreme market behavior and unreasonable valuation structure behind it.
Chart: The market performance after the listing of gigantic IPOs in the US stock market. Data source: Wind, compiled by 36Kr
Chart: The market performance after the listing of gigantic IPOs in the A - share market. Data source: Wind, compiled by 36Kr
Whether it was the madness of PetroChina or the blind optimism about Rivian, they are all signs that the irrational market sentiment has reached its peak in the later stage of a bull market. During such periods, after the market has experienced a long - term rise, investors are blindly confident, believing that the stock market is a myth that only goes up and never goes down. Their risk appetite is significantly increased, so their tolerance for fundamentals and high valuations is also improved.
For companies and issuing institutions, in the later stage of a bull market, the market has abundant liquidity and high valuation multiples, so they are more willing to strive for a good cashing - out price. This is the fundamental reason why gigantic IPOs are more likely to occur at the end of a bull market.
From this perspective, rather than saying that the listing of "gigantic" companies leads to the end of a bull market, it is more accurate to say that the irrational prosperity of the market at the end of a bull market gives rise to the emergence of gigantic companies. And the siphon effect of these massive IPOs indirectly intensifies the structural imbalance of funds, leading to an overall market adjustment, accelerating the market's de - leveraging process, and ultimately triggering a chain reaction, resulting in the end of the bull market.
Is the AI Bubble Coming?
After several years of AI fever and continuous new highs, investors are worried about a bubble in the US stock market. This concern is not unfounded but stems from a historical comparison of a series of structural indicators.
(1) In terms of the market's crowding degree, the current crowding degree of the AI concept in the US stock market has increased significantly, which is highly similar to the period of the Internet bubble.
Data shows that in the past 15 years, the US stock market has continuously outperformed the global market, and its market value now accounts for more than 60% of the total global stock market scale. Within the US stock market, the total market value of the top five technology companies has exceeded the sum of the Euro Stoxx 50 Index, the UK, India, Japan, and Canada, accounting for about 16% of the global public stock market.
Chart: The US stock market value accounts for a major part of the global scale. Data source: Goldman Sachs, compiled by 36Kr
(2) Valuations and the Buffett Indicator also point to the possibility of a bubble. Currently, the average price - earnings ratio of the top five giants in the US stock market is about 37 times, significantly higher than the historical average of the S&P 500 Index.
At the same time, looking at the Buffett Indicator (the ratio of the total US stock market value to the US GDP), the current ratio is about 225%, higher than the speculative critical line of 200%, and even higher than the 180% level at the peak of the 2000 Internet bubble.
Therefore, both the sharply rising valuation level and key references such as the Buffett Indicator indicate that there are indeed signs of an AI bubble in the US stock market at present.
However, compared with the 2000 Internet bubble period, this round of AI market is mainly dominated by technology giants. In the past few years, their valuation expansion has been accompanied by performance growth, and the support for future performance realization is higher. Therefore, in comparison, although the absolute valuation level of current AI technology stocks has risen, the relative level is still significantly lower than that during the bubble period. This is also a key reference for some sellers to believe that the AI bubble has not formed.
Chart: Historical comparison of the expected P/E of the S&P 500. Data source: Goldman Sachs, compiled by 36Kr
(3) From the perspective of the market's securitization degree and capital structure, the significant increase in AI debt financing also brings some risk warnings. In particular, the huge supplier financing around OpenAI is very similar to the mutual equipment purchases among telecommunications suppliers during the Internet bubble period.
In 2000, the market's expectation of the demand for network bandwidth increased, giving rise to a large number of start - up telecommunications operators. At that time, Lucent Technologies, the industry leader, provided huge loans or guarantees to its customers in order to seize market share and maintain its "myth" of high - speed growth, and these customers had to use this money to buy Lucent's telecommunications equipment. At its peak, the scale of this supplier financing once exceeded $15 billion.
With the bursting of the Internet bubble, global telecommunications companies significantly cut their capital expenditures. Lucent's performance then tumbled, the huge loans became non - performing assets, the company fell into losses, and its stock price also plummeted by more than 90%, becoming a representative of the bursting of the Internet bubble.
From historical experience, this large - scale supplier financing highly depends on the subsequent performance realization of OpenAI. If its commercial profitability fails to meet expectations, its procurement commitments to suppliers will not be fulfilled. At that time, the huge orders of suppliers may become non - performing assets, having a negative impact on performance, triggering a wide - ranging chain reaction, and ultimately leading to a systemic risk.
Chart: The internal cycle of supplier