Driven by the artificial intelligence wave, the US stock market has repeatedly hit record highs. However, a growing number of market observers are sounding the alarm, warning that this strong rally fueled by AI may have entered a phase of rapidly accumulating risks. Prominent economic research institutions point out that the current US stock market is exhibiting typical late-stage bubble characteristics. Although market sentiment remains high in the short term, with related indices still having room to push higher by the end of this year, the true period of risk realization may arrive in the next year or two, at which point the market could face a significant and deep correction.
This cautious stance is not driven by anomalies in a single data point, but rather by multiple historical peak indicators flashing red simultaneously. The cyclically adjusted price-to-earnings ratio of the US stock market is currently approaching the extreme levels seen during the previous internet bubble, and the relative valuation of the S&P 500 against US Treasuries is also at historical highs. Meanwhile, market expectations for corporate earnings growth over the next year have reached historical extremes. If subsequent corporate earnings fail to meet these optimistic expectations, the lofty valuations will only be corrected through falling stock prices.
At the industry level, the massive capital expenditures in the AI sector have become a Sword of Damocles hanging over the heads of tech giants. Over the past two years, leading tech companies have poured unprecedented amounts of capital into data centers, chips, and cloud computing infrastructure in their race to dominate the AI high ground. However, this intense investment has not yielded commensurate cash returns. Analysts estimate that the combined free cash flow of top AI cloud service providers could turn negative in the near future. This signifies that the AI industry is transitioning from merely painting a blueprint for growth to the ruthless stage where commercial returns must be validated with hard cash.
The structural vulnerabilities of the US stock market are also becoming increasingly apparent. The continuous rise of the indices relies heavily on the performance support of a few mega-cap tech companies, and the high concentration of funds masks internal imbalances within the market. At the same time, IPOs and follow-on offerings in the primary market have significantly heated up, with a massive supply of stocks poised to enter the market. Historical experience shows that a clustering of equity financing by companies is often a hallmark of overheated market sentiment, typically followed by a rapid reversal in market trends.
If the aforementioned factors are hidden risks accumulating within the market, then the rise in US Treasury yields could be the external blade that bursts the bubble. Recently, the yield on the 10-year US Treasury briefly approached the critical psychological threshold of 5%. For global asset pricing, the trajectory of this benchmark interest rate is crucial. Looking back at major asset bubbles over the past few centuries, their bursts have often been accompanied by a sharp surge in the financing costs of core assets.
The high-interest-rate environment is particularly devastating for the AI industry. AI infrastructure construction is a typical capital-intensive business that requires continuous capital injection. If financing costs remain persistently high, tech giants will inevitably re-evaluate and scale back their capital expenditure plans. Meanwhile, the rise in the risk-free rate will directly compress the reasonable range for stock valuations, subjecting the AI sector to a triple squeeze of increased financing costs, constrained capital expenditures, and valuation downgrades. A more profound impact is that the US government debt has exceeded 100% of GDP; high interest rates will lead to a surge in interest expenses, forcing the government to issue more bonds, thereby creating a vicious cycle that pushes yields even higher.
Faced with these complex risks, even the most steadfast optimists on Wall Street are beginning to adjust their expectations. Some senior analysts have downgraded the probability of long-term prosperity for US stocks and correspondingly raised their forecasts for entering a bear market. While they remain optimistic about short-term index levels, they universally emphasize that the logic for future gains must shift from valuation expansion to the actual realization of earnings.
The US stock market is currently at a subtle crossroads. On the one hand, the industrial transformation brought about by AI technology, the resilience of corporate earnings, and the stress tolerance of the macroeconomy continue to provide upward momentum for the market. On the other hand, lofty valuations, massive capital expenditures, rising US Treasury yields, and extreme market concentration are continuously pushing the bubble burst threshold higher. In the short term, as long as US Treasury yields do not break through key thresholds and AI investments can be smoothly converted into cash flow, high valuations might be sustained for a while. However, if interest rates continue to rise while earnings growth falls short of expectations, the market narrative will rapidly shift from chasing the technological revolution to re-evaluating asset prices, forcing investors to face the reality of whether current stock prices have already overdrawn the future.
