Either the US stock market has entered a kind of financial heaven where earthly economic laws no longer apply, or the conditions are being created for a crash and a consequent financial crisis of major proportions.
The market boosters adhere to the former, basing themselves on the enormous changes being wrought by AI and its tremendous potential for lifting the productivity of labour. Others, however, are sounding increasingly loud warnings.
A recent editorial in the Wall Street Journal dismissed the latter with remarks typical of many.
Commenting on the initial public offering (IPO) of Elon Musk’s SpaceX and those to come shortly of OpenAI and Anthropic, it said that “amid the elegies [highlighting growing problems] about American capitalism, the latest mini-IPO boom is a welcome tribute to the dynamism of US markets that no other country can match.”
The IPO for SpaceX, founded in 2002 for space exploration but which has now extended into broadband, mobile satellite service and data centres for AI, was filed with the Securities and Exchange Commission last Wednesday. It is said to be the largest in history.
Very few of the company’s shares will be available to public investors; most will initially be in the hands of Musk. But under new rules recently introduced by the NASDAQ exchange, it will be included in indexes which are tracked by Exchange Traded Funds, meaning that billions of dollars will flood into the market to buy its shares.
According to estimates by JP Morgan, if 50 percent of the company’s shares are eventually floated, the market valuation will reach $2 trillion.
The IPO boom comes on top of a surge on Wall Street over the last two months. Since April and the announcement of a “ceasefire” in the war on Iran, it has powered ahead, with the S&P 500 index rising by 12 percent.
It has brushed aside rising energy prices, amid warnings that oil reserves are being run down at a rapid rate, the failure of the Trump administration to secure the opening of the Strait of Hormuz, indications of an inflation surge, flowing from, but going well beyond oil, a selloff in the US and global bond markets, which has seen yields (interest rates) reach levels not seen since before the global financial crisis of 2008, and the clear indication from the US Federal Reserve that cuts in its base interest rate expected for this year are now off the table.
The structure of the market, which was already dependent on high-tech stocks, has undergone a further concentration.
It is calculated that just five tech stocks—Alphabet (Google), Nvidia, Amazon, Broadcom and Apple—have accounted for more than 50 percent of the recent gains in the S&P index. At the start of the year the prevailing sentiment was that there would be a broadening of market gains.
According to Bloomberg, the AI chipmaker, Nvidia, which has been likened to the sun at the centre of an AI planetary system, has been responsible for nearly a fifth of the rise in the S&P since the start of the year and 15 percent of the $32 trillion rise in the market capitalisation of the index since 2023.
One of the key features in the rise of Nvidia and its role in powering the market is the degree of circularity involved. The company either invests in or lends money to companies which then buy its chips for AI development. Over the past 16 months Nvidia has committed some $90 billion in investments and partnerships with companies in such deals.
In an agreement with one company, which would use its chips, Nvidia agreed to buy some of its products and invest up to $2.1 billion to facilitate its expansion, meaning that it was at the same time a customer, supplier and potential shareholder.
As the Financial Times reported recently: “The spending spans more than 145 companies, from AI model developers and cloud providers to suppliers building the infrastructure underpinning the boom. ‘They are funding everyone,’ said a Silicon Valley banker.”
These examples point to a broader phenomenon—the narrowness of the Wall Street boom.
The FT reported that the rise since late March “has been driven by the smallest number of stocks on record.” According to analysis by UBS, an index of stocks called “effective constituents” of the S&P 500 went to a record low of 42 at the beginning of this month, compared to the level of 100 which has been characteristic of the market’s rise in recent decades.
And much of this boom is not based on profits made by the AI firms today but the expectation that the investments, amounting to hundreds of billions even trillions of dollars, will bring massive returns in the future. The three major firms at the centre of the new round of frenzied activity are all making losses.
Anthropic has said it expects to turn a profit in the second quarter of this year. OpenAI has said it expects to burn through $600 billion cash before becoming profitable in 2030. SpaceX, whose operations are “something of a financial mystery” in the words of the New York Times, boosted its revenue by 33 percent in 2025 to $18.7 billion. But it lost $4.9 billion in 2025 and in the first quarter of this year recorded a $4.3 billion loss.
The contrast between the Wall Street surge and developments in the economy and the global financial system—oil prices set to rise further, yields on bonds rising to their highest levels in nearly two decades and the prospect that central banks are moving to raise rates—has prompted warnings that some sort of “correction” is bound to come, with the question being how far it might go.
As one capital markets strategist commented to the FT there was “an incompatibility between having equities at all-time highs … and at the same time, interest rates and energy [markets] pricing in a lasting impact on the economy.”
Former Goldman Sachs executive and treasury secretary Robert Rubin pointed to historical experience in an op-ed piece in the FT under the title “Americans beware: markets can be out of sync with reality.”
He began by noting that markets and the economy seem to be inured to the chaos resulting from the policies of the Trump administration and that while they were not yet reflected in short-term data, “enormous damage” had been done to the economy “with effects likely to play out over time.”
“Markets can be out of sync with reality for an extended period, and then react rapidly and harshly. In the 18 months leading up to October 1987, the stock market soared, despite concerns about looming risks. Then, on October 18, the Dow fell by 22 percent in a single day.”
Rubin did not say it, but a fall of that magnitude, or even significantly smaller today, would have far greater consequences than the 1987 crash because of the enormous growth of debt and the complex, arcane, debt-based financial mechanisms developed in the four decades since then.
One of the key issues, in the final analysis, is what will be the longer-term impact of the development of AI.
In a recent comment piece in the FT, Joachim Klement, managing director of the UK investment bank Panmure Liberum, began by noting that over the past four quarters the US economy had been largely growing because of the tech boom, with 93 percent of its expansion explained by tech investments and if this drops, even by a relatively small amount, “the US economy will enter recession very quickly.”
As far as the stock market is concerned, he estimated that in the longer run the return on the massive investments by the hyperscalers would be “highly negative” for all of them with the exception of Amazon.
According to Klement, his calculations showed that “if the hyperscalers continue on the current trajectory, the AI boom will become one of the largest destructions of shareholder value in history.”
The issues at hand go even deeper because there is an inherent contradiction at the very centre of AI development under capitalism. This is because the “success” of AI is dependent on its capacity to return profits for the firms that use it, bringing a massive increase in the wealth of the oligarchs which own them. And that is determined by the number of jobs that can be eliminated. Those that do it are rewarded by Wall Street with a boost in their share price while those that do not face being eliminated.
Amid warnings by investors that 80 percent of all jobs could be done by AI within years, the initial numbers are in. In the first four months of this year American employers announced more than 300,000 job cuts with technology and AI cited as the main reason.
The argument is sometimes advanced that, as in the past, while technological developments destroy jobs, they will also create new ones, and that this will be the case with AI. What this argument ignores is that those “new” jobs will likely be able to be done by AI itself.
There is no question that AI is a massive development in the productivity of labour and lays the basis for a tremendous advancement for humankind. But for that to take place it must be freed from the destructive grip of the capitalist profit system through the taking of power by the working class and the reorganisation of the economy on new, socialist, foundations.
The frenzied boom on Wall Street and the growing indications of its direction is a warning that the political struggle by the working class to achieve this task is not something for the distant future but has become an urgent necessity.