High valuation of U.S. stocks superimposed with multiple risks

I. Introduction: A Landmark Event

On June 12, 2026, SpaceX officially began trading on the Nasdaq under the ticker SPCX. Its IPO price was $135, its closing price on the first day was $160.95, and its total market capitalization exceeded $2.1 trillion. This event itself may not be entirely surprising—after all, the market had long been eager for the listing of this space exploration company founded by Elon Musk. What is truly worth examining is that, in the quarter before the IPO, the company recorded a net loss of $42.76 billion; since its establishment in 2002, its cumulative losses have been about $41.3 billion; and it achieved profitability only once in 2024.

Estimated using an annualized revenue of $20 billion, SpaceX’s price-to-sales ratio is close to 90x, higher than any company in the S&P 500. Some analysts point out that its price-to-sales ratio has exceeded 112x—far above Tesla’s 15x and Nvidia’s nearly 20x. A company with persistent massive losses and less than 5% of its total shares available in circulation receives a market valuation of over $2.1 trillion—so is this pricing a grand vision for the stars, or just another footnote to irrational exuberance in the market?

II. Overall Valuation: Multiple Indicators Point to Extreme Levels

SpaceX is not an isolated case. By mid-2026, the overall valuation of the U.S. stock market has reached historically extreme levels.

The Shiller P/E ratio (CAPE) is one of the most persuasive long-term valuation indicators. As of June 2026, the S&P 500’s Shiller P/E ratio has risen to roughly the range of 39.5 to 41.7x, the highest level in the past 25 years. It is only below the extreme peak of about 44x during the 1999 internet bubble. Data show that this indicator is about 42x, while the 155-year historical median is about 16x. This is only the third time since records began in 1871 that the valuation has broken through the 40x threshold. The market’s implied future return, as reflected by the Shiller P/E ratio, has fallen to about 1.3%.

The Buffett Indicator (the ratio of total U.S. stock market capitalization to U.S. GDP) also sends a strong warning signal. From May to June 2026, the indicator has surged to the 227% to 238% range, setting a century-high record and far exceeding the dot-com bubble peak of about 175% in 2000 and the high of about 190% in 2021. Based on current valuation levels, some estimates suggest that the annualized return on U.S. stocks over the next 8 years will be only about 1.1% to 1.2%.

As for the forward P/E ratio, the S&P 500 is around 21 to 22x, while the 25-year historical average is about 16 to 17x—implying a premium of roughly 25% to 30%. Since 2026, the S&P 500 index has already set 23 new all-time highs.

These three independent indicators point to the same conclusion: current U.S. stock valuations have severely deviated from historical averages. Those supporting the “new paradigm” argument need to explain why, this time, the market’s performance after every previous instance of extreme valuation would differ—especially when the historical record is not on their side.

III. AI Bubble: A Misalignment Between Investment Frenzy and Return Reality

The core narrative behind the current market’s valuation expansion is artificial intelligence. The AI theme has driven continued gains for technology giants related to semiconductors, cloud computing, and AI infrastructure. As a result, a small number of large tech stocks account for the vast majority of the S&P 500’s overall increase.

However, the reality of AI investment is facing severe tests.

The mismatch between capital expenditure and returns is becoming increasingly prominent. Renowned short-seller Jim Chanos has pointed out that there is a huge “financial mismatch” in the current AI industry chain. He revealed that even in the current environment of GPU shortages and data center capacity falling short of demand, the pre-tax return on invested capital (ROIC) for related transactions is mostly only 5% to 8%. If during periods of supply shortages the return can only reach single digits, then once supply-and-demand conditions reverse, the business model will become even harder to sustain the high valuations.

Data center construction is encountering real-world headwinds. In the first quarter of 2026, across the United States, the total value of data center projects that were blocked or delayed was as high as approximately $130 billion—roughly equal to the total for all of 2025. A JPMorgan report noted that among data center projects planned to be completed in 2027, more than 60% have not yet started construction. The main reasons relate to power supply bottlenecks—data centers currently account for 5% of U.S. electricity demand, which could grow by 3x by 2035—as well as a surge in opposition from community groups.

The nine major global cloud service providers have raised their 2026 capital expenditure to $830 billion, but whether such a massive level of investment can generate corresponding returns remains an open question. The combination of rapid depreciation in AI data centers, ongoing energy consumption, and uncertainty in commercialization pathways makes the foundation of this investment frenzy far more fragile than it appears on the surface.

IV. Geopolitical Shock: Market “Immunity” or a Delayed Reaction?

On February 28, 2026, the U.S. and Israel jointly carried out airstrikes on Iran. Iran’s Islamic Revolutionary Guard Corps then announced the closure of the Strait of Hormuz. This “energy chokepoint,” which carries about one-fifth of the world’s oil and liquefied natural gas supplies, was blocked for a full 110 days.

During the blockade, crude oil export volumes from ports in the Arabian Gulf plunged by 95%, and LNG carrier traffic fell by 99%. International oil prices at one point went into a frenzy: Brent crude rose from $70 to a high of $119, an increase of more than 70%. An International Energy Agency report showed that the closure of the strait caused Gulf states’ oil production to be 14.4 million barrels per day lower than pre-war levels, and that over the past 100 days the cumulative global oil supply was reduced by more than 1.2 billion barrels.

However, U.S. stocks did not experience a sharp decline during this period; instead, they repeatedly hit record highs. This “market immunity” phenomenon has prompted two interpretations: first, the market believes the impact of the geopolitical conflict is temporary and controllable; second, the market may be experiencing an upswing driven by some non-fundamental factors.

Some analyses suggest that the rise in U.S. stocks during the Iran war may have been partly due to a “gamma squeeze” effect—investors bought large quantities of call options, forcing market makers to buy spot shares to hedge risk, creating a self-reinforcing cycle of gains. After SpaceX’s listing, its options market has also been accused of potentially triggering a gamma squeeze. If such mechanical trading behavior dominates market pricing, then once options demand fades, the market could face a rapid reversal.

On June 18, the U.S. and Iran signed a memorandum of understanding, and the Strait of Hormuz officially reopened. Brent crude fell back to around $80. But the 110-day blockade has already produced irreversible economic consequences—inflation spread across Japan, South Korea, and countries in Southeast Asia, and even after oil prices declined, local prices did not retreat. The follow-on effects of this geopolitical shock are far from fully reflected in asset prices.

V. Gold and Silver: Why Did Safe-Haven Assets Fail?

In theory, geopolitical conflicts, inflation concerns, and expectations of currency depreciation should push up the prices of safe-haven assets such as gold and silver. However, reality presents a different picture.

On June 23, 2026, spot gold fell below $4,100/oz, setting a two-week low and dropping 2.35% during the day; spot silver fell below $62, down 4.80%. Several stocks in China’s A-share precious metals sector hit their daily trading limit down.

Multiple international investment banks have recently turned bearish on gold’s short-term outlook. Goldman Sachs cut its gold target price for end-2026 sharply to $4,900 per ounce, a reduction of $500, citing that the Fed is unlikely to cut rates within the year. Citibank expects the gold price to reach $4,300/oz in the next 0 to 3 months. Morgan Stanley reduced its gold target price for the second half of 2026 to $5,200/oz.

In China, many domestic banks in June intensively increased margins for precious-metals trading; for some contracts, margins were raised from 100% to 120% to 140%. Industry insiders said that recent volatility in the international precious metals market has been severe, and daily swings of over $100 have become the norm.

Gold’s inability to deliver its traditional safe-haven role in the short term may be due to factors including: hawkish signals from the Fed that have pushed up real interest rates; some countries selling gold to obtain U.S. dollar liquidity; and widespread selling caused by market liquidity tightening during wartime. But from a longer-term perspective, the global central bank de-dollarization trend, the trust crisis triggered by the U.S. freezing of Russian assets, and ongoing geopolitical uncertainty still provide fundamental support for gold’s long-term upward trajectory.

VI. Government Direct Entry Into the Market: From Regulator to Shareholder

A noteworthy institutional change is that the U.S. government is shifting from being a market regulator and purchaser to becoming a direct shareholder in multiple private companies.

Since January 2025, federal agencies have put nearly $21 billion into 17 companies, buying stocks and equity. The funds have mainly flowed into two industries: semiconductors account for 52.4%, and critical minerals account for 42.5%.

The largest investment is in Intel. In August 2025, the U.S. government agreed to buy 433.3 million shares at $20.47 per share, taking a 9.9% stake and becoming one of Intel’s largest shareholders. As of mid-June 2026, the government’s paper profit exceeds $43 billion, and Intel’s stock price has risen a cumulative 440% since the investment. A Morningstar senior tech analyst said that the government’s involvement and shareholding helped the stock take off in 2026.

In addition, senior U.S. government officials have held preliminary discussions with major U.S. AI companies about the possibility of the government taking equity stakes in these firms.

This trend has raised profound questions about institutional nature. When the government simultaneously plays three roles—market regulator, industrial policy maker, and corporate shareholder—conflicts of interest are almost unavoidable. Companies in which the government holds shares effectively receive de facto “government endorsement,” and their stock price gains partly reflect this political premium rather than purely commercial value. Even more alarming is that if the government can support specific stocks and even the entire market through direct entry, the price discovery function of the market pricing mechanism will be severely distorted.

VII. Conclusion: The Market Amid Multiple Imbalances

The U.S. stock market currently faces an overlap of multiple imbalances: valuations are at historical extremes (Shiller P/E at 41x, the Buffett Indicator above 230%); the AI investment frenzy faces a dual bind of low capital returns (5%-8%) and obstacles to infrastructure construction (with $130 billion in projects stalled); the long-term consequences of geopolitical conflicts have not yet been fully priced in; and direct government entry into the market is blurring the boundary between the market and politics.

Historical experience shows that the force of mean reversion in markets is like gravity—it can be temporarily offset but cannot be permanently eliminated. In 155 years of history, a Shiller P/E above 40x has occurred only three times; the current level of the Buffett Indicator is far above the peaks of 2000 and 2021. These extreme readings themselves do not point to the precise timing of a crash, but they clearly lead to one conclusion: long-term expected returns have been severely compressed.

Market irrationality can last for a long time—CAPE was already high in 2024, and the market has risen by more than 30% since then. But “long” does not mean “forever.” When four engines are smoking at the same time and the flight altitude keeps climbing, no matter how confident the pilot is, gravity will eventually take effect. The question has never been “whether the market will revert to the mean,” but “when and in what way it will revert”—and the longer the wait, the more violent the reversion often is.