Alarm Bells Ringing: Global Tech Stock Plunge Reveals Reality Check on AI Valuation Bubble

Recently, the U.S. tech sector faced a sharp sell-off that rapidly spread to major global markets, as investor concerns over the sustainability of artificial intelligence valuations erupted.

Within a single week, the combined market capitalization of the U.S. tech “Big Seven” evaporated by nearly $800 billion, with Nvidia alone shedding $350 billion in value from its peak of $5 trillion. This plunge marks a market shift from grand AI narratives toward pragmatic scrutiny of profitability and valuation rationality.

Global Chain Reaction: U.S. Leads Decline as Asian and European Markets Follow Suit

The tech stock crash originated on Wall Street but rapidly escalated into a global financial event. All three major U.S. indices fell, with the tech-heavy Nasdaq dropping approximately 3% for the week — its worst weekly performance since the 2018 tariff war.

The fear gauge VIX surged to its highest level in over two weeks, climbing more than 8% in a single day, reflecting palpable market anxiety.

U.S. tech giants bore the brunt of the sell-off. Among the so-called “Big Seven,” Nvidia’s stock plunged nearly 3%, erasing $143.1 billion in market value overnight; Tesla, Oracle, and AMD all fell more than 3% at one point.

Palantir, a previously sought-after AI concept stock, also suffered, with its share price falling over 6%.

The ripple effects of this tech sell-off quickly spread to global markets:

Asian markets took a heavy hit, with Japan’s Nikkei 225 index dropping over 1.6%, marking its steepest weekly decline since April. South Korea’s Kospi index dropped 2.2%, facing its worst weekly performance since last November.

European markets also declined in tandem, with the pan-European Stoxx 600 index falling 0.4%, while Germany’s DAX 30 and Spain’s IBEX 35 both dropped over 1%.

Trigger: SoftBank’s Sell-Off and Earnings Warning Catalyze Plunge

Japanese tech investment giant SoftBank Group’s complete liquidation of its Nvidia holdings directly ignited this sell-off. In October 2025, SoftBank sold all its Nvidia shares — approximately 32.1 million shares valued at around $5.83 billion.

This marks SoftBank’s second complete exit from Nvidia — after its first liquidation in 2019, the firm missed Nvidia’s golden period of soaring from a $100 billion market cap to a trillion-dollar valuation.

Meanwhile, AI infrastructure company CoreWeave plunged over 16% after revising its full-year 2025 revenue forecast downward from a high of $5.35 billion to $5.05-$5.15 billion, citing slow data center construction.

This earnings warning intensified market selling pressure on AI-related stocks.

Notably, while liquidating its Nvidia holdings, SoftBank announced an additional $22.5 billion investment in OpenAI, with plans to complete the full investment through SoftBank Vision Fund 2 in December.

This move signals a capital reallocation within the AI sector, shifting focus from hardware companies to application-level ventures.

Valuation Bubble: Reality Check After AI Frenzy

Market concerns over AI stock valuations were laid bare during this sell-off. Palantir’s static price-to-earnings ratio once soared to 1,000 times. Even based on its 2025 and 2026 EPS forecasts, its forward P/E ratios remain exceptionally high at 246 times and 181 times respectively — far exceeding the typical 30–40 times range for established software companies.

A Goldman Sachs research report released in October noted that while current tech valuations are elevated, they fall short of the frenzy seen during the dot-com bubble. However, the report cautioned: “The biggest risk is that tech companies’ revenues begin to disappoint investors, raising doubts about whether current returns can be sustained.”

“Big Short” investor Michael Burry adopted a bearish stance, purchasing put options targeting AI companies like Nvidia and Palantir.

While some investment managers view this as hedging rather than outright shorting, the news has amplified market anxiety.

Wall Street targets for AI star stocks show significant divergence. Bank of America bucked the trend by raising Palantir’s target price from $215 to $255, calling it a “true AI practitioner.” Citigroup, however, maintained a relatively neutral stance with a $190 target.

Macro Environment: Government Shutdown and Lack of Economic Data Amplify Volatility

The U.S. government shutdown has entered its 38th day, becoming the longest in history, delaying the release of multiple key macroeconomic reports. The absence of official economic indicators has forced investors to rely on private-sector data, intensifying market volatility and the formation of speculative positions.

Amid this data vacuum, preliminary ADP payroll figures revealed U.S. private employers cut jobs in October, further dampening market confidence.

White House National Economic Council Director Kevin Hassett warned that the government shutdown’s economic impact is far more severe than anticipated, projecting a slowdown in fourth-quarter GDP growth.

Comments from Federal Reserve officials also heightened market uncertainty. New York Fed President John Williams stated that the Fed may soon need to expand its balance sheet through bond purchases to meet liquidity demands in the financial system.

He emphasized that any resumption of bond purchases would be purely a liquidity management operation, not signaling a return to “a new round of quantitative easing.”

Industrial Reality: The Gap Between Capital Expenditures and Actual Returns

The mismatch between massive capital expenditures in the AI sector and actual returns has become a core market concern. A Goldman Sachs report indicates that technology companies may have overinvested, increasing the risk that future returns on AI capital expenditures will fall short of current market expectations.

Data shows that since 2024, the credit market has seen over $20 billion in asset-backed securities (ABS) issued for data center investments. Tech companies included in Goldman Sachs’ AI ETF have issued $141 billion in corporate bonds since 2025, surpassing the $127 billion issued throughout all of 2024.

OpenAI’s financial situation is alarming. Disclosures reveal that while OpenAI generated nearly $13 billion in subscription revenue from millions of customers this year, it signed over $1 trillion in computing power procurement agreements for 2025, with the company’s valuation nearing $1 trillion.

Meanwhile, OpenAI recorded a net loss of $13.5 billion in the first half of 2025.

McKinsey research indicates that nearly 80% of companies deploying AI have not seen increased net profits as a result. A survey reveals that 95% of U.S. companies adopting generative AI have not profited from the technology, highlighting the severe challenges in commercializing AI applications.

Wall Street Divided: Bubble or Future?

Faced with the sharp correction in the AI sector, Wall Street institutions have shown starkly divergent views.

After conducting in-depth research on over 30 classic bubble cases spanning the past 400 years, renowned U.S. hedge fund Coatue released a report arguing that AI is currently in its early “displacement/adoption” phase and is far from reaching the peak of a bubble.

Coatue points out that the fundamental difference between today’s tech giants and those during the dot-com bubble lies in their robust profitability and cash flow. The average price-to-earnings ratio for the seven major U.S. tech companies now stands at approximately 28 times, significantly lower than the 67 times seen during the 1999 internet bubble.

However, bearish voices remain equally potent. Florian Ielpo, Head of Macro at Lombard Odier Investment Management, cautioned: “AI-related capital expenditures have reached staggering levels and increasingly rely on debt financing — strikingly reminiscent of the frenzied investments during the 2000 dot-com bubble.”

As year-end approaches, investor attention is focused on two key events: Nvidia’s upcoming earnings release and the resolution of the U.S. government shutdown. Louis Navellier, Chief Investment Officer at Navellier & Associates, notes: “We are two weeks away from Nvidia’s earnings release, which could serve as a catalyst to reaffirm the AI investment thesis.”

Capital markets never entirely reject innovation, but when valuations become too detached from reality, a “reality check” becomes inevitable.

Today’s tech giants possess cash flows and business models that the dot-com companies of 2000 lacked. Yet with AI-related stocks now accounting for over 44% of the S&P 500’s weighting, even minor fluctuations can trigger global market chain reactions.