AI Bubble Bursting or Catching Its Breath? US Stocks Hit Liquidity Limits as Nasdaq Slides 5%
Summary
- The report said the Nasdaq plunged more than 5%, while the Nasdaq 100 Index and S&P 500 Index moved toward a bear market and a correction, respectively.
- It said concerns are growing that US stocks are hitting liquidity limits as new share issuance reached $125 billion including Anthropic's listing, while the AI risk premium widened.
- Experts said the focus will shift from AI infrastructure companies to robotics, pharmaceuticals, and traditional industrial companies as cracks emerge in the techno-dollar structure and hawkish rate-hike risks build.
Forecast Trend Report by Period



The Nasdaq plunged more than 5% on Nov. 20, prompting fresh questions over whether the artificial-intelligence frenzy that had underpinned US equities is fading fast. With global technology shares falling in tandem from New York to Seoul, the Nasdaq 100 is nearing bear-market territory. On Wall Street, strategists also argue the S&P 500 has entered a correction after its longest losing streak since the 2020 pandemic shock.
Reports from the Financial Times and Bloomberg said the market mood has swung from indiscriminate buying to panic-driven selling. The shift goes beyond routine profit-taking. Investors are now grappling with an onslaught of stock and bond supply that is straining market liquidity, cracks in the AI-funded "techno-dollar" cycle, and a collision of rate-hike fears with geopolitical risks.
A $125 Billion Supply Shock: "The Market Coughed It Up"
The most immediate trigger for the selloff was, ironically, the listing of AI startup Anthropic, which had been expected to be one of the biggest market debuts on record. Anthropic entered the market with a $1.5 trillion valuation, but its shares fell sharply soon after listing, erasing most of the gains from its offering price. In general, overheated blockbuster initial public offerings are often read as a sign of a short-term market top.
Market professionals say the sheer scale of recent fundraising is testing the market's limits for absorbing new supply. Including Anthropic's $100 billion listing this week, US equity markets alone were hit with $125 billion of new share issuance. That marks the largest weekly issuance total in more than a decade.
Even established Big Tech companies including Alphabet and Meta have recently turned to follow-on share offerings. SpaceX also sold $32 billion of AI chip-related bonds this week at interest rates above 8%. That is unusually high for debt rarely associated with investment-grade borrowing costs, underscoring how much the market's AI risk premium has widened. For the past 20 years, US corporations had been net buyers of stocks through share repurchases and other programs. Goldman Sachs projects that, as large AI companies including OpenAI continue to list next year, net equity issuance in the US stock market will turn positive for the first time since 2000, while corporate bond issuance climbs to $2.8 trillion, the highest in at least a decade.
Adam Rich, deputy chief investment officer at Vaughan Nelson, said the market "literally coughed up supply" as soon as Anthropic began trading. His biggest concern is whether enough liquidity will remain to absorb the wave of fundraising still in the pipeline for next year. Bryce Doty, a bond fund manager at Sit Investment Associates, said that when a single company pulls in massive sums, someone else has to sell existing stocks and bonds to provide the cash. That, he said, raises the risk of crowding out across the market.
Unease Over the "Techno-Dollar" Cycle
At a broader level, doubts are growing over the "techno-dollar" structure that has helped sustain the surprising resilience of US equities.
In the past, Middle Eastern oil exporters reinvested dollars earned from crude sales — the petrodollar — into US Treasuries, the world's largest bond market. More recently, Asian semiconductor exporters such as South Korea and Taiwan have been pouring dollars earned from US Big Tech into American equities, where US stocks account for 64% of the MSCI All Country World Index. That "techno-dollar" flow has helped prop up the market. Deutsche Bank said South Korea's semiconductor exports surged 200% this year, lifting its semiconductor trade surplus to nearly $250 billion. For the first time, global semiconductor trade volumes also surpassed oil and gas trade.
The problem is the structure's circularity. AI infrastructure spending by US technology giants feeds semiconductor exports from Asia, and that money then flows back into US stocks, reinforcing yet another round of AI investment. Once demand weakens, the cycle can spin into reverse and trigger sharp volatility in global capital flows. Jan Hatzius, Goldman Sachs' chief economist, said he remains optimistic about AI's economic impact. But he added that once a large amount of capital equipment has been built, investment growth could eventually stall and turn negative.
Hawkish Fed Signals and Geopolitical Risks Add Fuel to Market Fatigue

Fatigue around the AI trade had already been building after a summer slump in Asian memory-chip shares. Bond investors had grown tired of absorbing debt from tech companies facing mounting questions. US communities were also pushing back against the construction of massive data centers. At the same time, institutional investors had become increasingly frustrated with what the article described as fragile corporate governance at Nvidia, whose market capitalization reached $5.7 trillion in October, more than double the total value of the UK's FTSE 100.
With investor sentiment already extremely fragile, outside shocks hit all at once. Inflation remained sticky, and Federal Reserve Governor Christopher Waller made hawkish remarks by not ruling out a 0.5 percentage-point rate increase. That deepened concerns over higher funding costs. Repeated reports of major cyberattacks and technology failures darkened sentiment further. Short sellers also published a report raising links between AMD and China, putting export-control risks tied to China back in focus.
Outlook: From "Building AI" to "Using AI" as Stock Picking Intensifies
The S&P 500 is up only about 5% this year. Without a year-end rebound, it would deliver one of its weakest annual performances of the past 20 years. Market professionals note that every transformative technology tends to pass through the same stages: innovation, explosive expansion, excess, and then a harsh separation of winners from losers. By that reading, the AI market has entered a new phase.
Vincent Mortier, chief investment officer at Amundi, said investors have so far been fixated on core AI infrastructure companies. By 2027, that focus will shift completely, he said. With monetization still difficult and elevated valuations becoming harder to justify, investors should pay closer attention to sectors that directly use AI, including robotics, pharmaceuticals and traditional industrial companies.
As Scott Bauer, chief executive officer of Prosper Trading Academy at the Chicago Board Options Exchange, put it, this week's market action was marked by "face-ripping volatility." That suggests the turbulence will be hard to avoid for now. Still, money is moving quickly into safe-haven US Treasuries, while major asset managers including Loomis Sayles are drawing a line between a default crisis and a valuation correction. That leaves the market enduring the pain of cooling excess while searching for a new balance.
Kim Dong-yoon, Hankyung.com reporter
This content was produced based on Epic AI, an artificial-intelligence investment information service developed by Hankyung AI. It does not constitute investment advice, and investors are responsible for their own decisions.
Kim Dong-yoon, Hankyung.com reporter
oasis93@hankyung.com
Korea Economic Daily
hankyung@bloomingbit.ioThe Korea Economic Daily Global is a digital media where latest news on Korean companies, industries, and financial markets.