U.S. AI Market Faces Correction as Tech Stocks Decline

Original author: Dong Jing

Source: Wall Street Journal

Wall Street is experiencing a costly wake-up call. As large tech stocks face their most intense monthly sell-off since March 2025, skepticism over the return on investment in AI infrastructure is rapidly growing, and signs of a retreat in AI trading are becoming increasingly evident.

On the last trading day of October, U.S. market capitalization lost over $1.5 trillion for the month, with the Nasdaq Composite Index plunging 5.3%—its worst monthly performance since March 2025—and the S&P 500 posted its second consecutive monthly decline.

Senior analysts Bert Dohmen and Dion Dohmen of Dohmen Capital Research, a market research and investment analysis firm, recently wrote that the sell-off has spread beyond the AI and semiconductor sectors, triggering a chain reaction of widespread leveraged liquidations and margin calls, leaving liquid assets such as gold and silver also unaffected. Meanwhile, questions about whether AI capital expenditures can generate substantial returns have evolved from fringe concerns into the dominant cloud overshadowing the entire sector.

Prominent institutions such as Elliott Investment Management have already issued warnings. According to a Wall Street Insider article, citing The Wall Street Journal, Paul Singer’s Elliott explicitly stated in its October letter to investors: “Right now, nearly every alarm bell signaling a major bear market is ringing,” and compared the current market environment to the “Roaring Twenties” preceding the Great Depression.

This judgment strongly resonates with the analysis in the latest article by Bert Dohmen and Dion Dohmen—both point to the same core question: Who will pay for AI?

The market downturn in October was shocking. Bert Dohmen and Dion Dohmen note in their article that the $1.5 trillion in market value erased from U.S. stocks in a single month was three times the $500 billion lost in a single day during the 1987 "Black Monday" crash.

The fuse was lit one after another. After Meta released its Q3 earnings report after hours on October 21, market sentiment shifted sharply; on October 28, Microsoft and Google released their earnings, with Google’s parent company Alphabet plunging 13.5% in a single day, dragging down the entire large-cap tech sector.

The MAGS ETF, representing the "Mag 7," has seen a sharp loss of momentum since its peak on October 13 and closed below the key support level of 70 for two consecutive days at the end of October, confirming that the earlier upward breakout was a false breakout.

It is worth noting that the actual peak for tech stocks and the semiconductor sector formed as early as June this year, when it did not attract widespread market attention. Dohmen and his colleague emphasize in their article that true market tops often emerge months before the dramatic crashes that capture public attention; by the time most investors recognize the risk, they are often already trapped.

The fundamental driver behind this wave of AI trading retreat is the market beginning to confront a long-ignored question: Can massive investments in AI infrastructure be translated into sustainable commercial returns?

Dohmen wrote in the article that concerns surrounding AI funding continue to grow, "casting a massive cloud over the market."

In his letter to investors, Elliott outlined three specific concerns regarding the AI sector:

Circular investments between AI model companies and AI infrastructure manufacturers, debt-funded capital expenditure expansion, and the emerging anti-AI social movement.

Elliott warned that once AI product pricing comes under pressure, or market expectations for AI development contract, the impact on the market "could be historic, at least as severe as any bear market in the past several decades."

This logical feedback loop is now manifesting in the market: mega-cap tech stocks have enjoyed a significant premium based on the AI narrative, but when the commercialization of this narrative comes into question, the valuation framework faces pressure to be restructured. Once leveraged capital heavily invested in these stocks begins to withdraw, it triggers margin calls, forcing investors to sell other assets to raise cash, creating a chain reaction.

The underlying structure of market sentiment may be undergoing a qualitative change.

Elliott believes that the current bull market is supported by multiple factors, which together constitute systemic risk—should any two of the pillars—expanding U.S. budget deficits, persistent inflation, private credit vulnerabilities, or the AI bubble—begin to falter, market sentiment could abruptly shift from "fear of missing out" (FOMO) to "fear of getting obliterated" (FOGO), triggering a self-reinforcing cascade of collapse.

Elliott also characterized the growing popularity of zero-dated options trading in today’s market, along with meme stock short squeezes, as behaviors "closer to gambling," arguing that these speculative activities are distorting the market’s normal pricing mechanisms.

Elliott managing partner Jesse Cohn recently wrote on X:

The stock market feels different now because it truly is different. Winners keep winning, while losers are getting hammered, and market reactions during earnings season are becoming increasingly unpredictable and volatile.

In response to the aforementioned risk assessment, some leading institutions have begun proactively reducing their risk exposure and shifting into a defensive stance.

Elliott has shifted its investment focus to capital preservation. Singer and co-CEO Jonathan Pollock wrote in the letter:

Throughout Elliott’s fifty-year history, periods of silence have indeed occurred, but historically, these phases have often coincided with market turbulence, followed by rare opportunities. Our goal is to be thoroughly prepared—protected against downside risk while holding capital to buy cheap assets at market lows—waiting for the tide to turn again.

Dohmen also clearly advises in the article against buying the dip in AI and related stocks. They use a poker table maxim as an analogy: “If you can’t tell who the patsy is, you’re the patsy.”

Whether Dohmen or Elliott, both point to a counterintuitive market pattern: the true market top often forms before the public even notices.

Dohmen noted in the article that the top of the AI and semiconductor sectors quietly formed in June this year, but at the time, it did not trigger sufficient caution. Investors who added positions after the peak now face substantial unrealized losses and a psychological struggle against selling at a loss—this is the classic path through which trapped positions accumulate during bear markets, and the root cause of heavy selling pressure that will accompany every future rally.

Elliott compares the current situation to the "Roaring Twenties" before the Great Depression, when risk appetite driven by debt coincided with technological fervor. History does not repeat itself exactly, but it often rhymes.