Korea's most prominent bearish economist says the AI-driven market rally carries the same hallmarks as the three worst financial crises of the past century.
Quay lại
Kim Young-ik warns AI bubble mirrors 2000 crash as KOSPI plunges 25%
Korea's most prominent bearish economist says the AI-driven market rally carries the same hallmarks as the three worst financial crises of the past century.
Kim Young-ik, the Hanyang University economist known as Dr. Doom, warned the AI boom displays the same pattern of excessive optimism, debt-fueled speculation and inflated asset prices that preceded the 1929, 2000 and 2008 crises.
"The issue is not whether the technology is real, but whether expectations for it are excessive," Kim, author of "The End of Debt Created by the AI Bubble," said in an interview. "Signs show that excessive optimism about AI and semiconductors is passing its peak."
Korea's benchmark KOSPI index has fallen more than 25% from its June high of 9,114.55, closing at 6,820.60 on July 16 after a single-day plunge of 6.4%. Kim pointed to Korea's Buffett Index — total market capitalization divided by nominal GDP — which reached 220% at the end of June, more than triple the 70% average since 2000. The ratio of market cap to M2 money supply hit 146%, up from a historical average of 58%.
Kim warned that if a crisis erupts, governments will have fewer policy tools than in 2008 or 2020 because debt levels across major economies have surged. "A crisis under such conditions — where governments and central banks lack weapons — could be severe and prolonged," he said.
Korea's semiconductor export growth rate peaked in June, Kim said, and while exports are still rising in absolute terms, the month-on-month growth rate has been declining. Samsung Electronics' stock price, which has tracked semiconductor export growth since the 1990s, fell sharply in early July despite the company posting record-breaking second-quarter earnings — a pattern Kim described as typical of markets where expectations have overshot reality.
The Bank for International Settlements has echoed the concern. In a recent report, BIS said the scale and speed of current AI investments exceed pre-crisis levels of past technological booms, and warned that big tech companies' practice of issuing corporate bonds or taking private loans to fund data center construction blurs the lines of collateral and debt ownership, creating a potential trigger for cascading defaults.
Private Credit Looms as $2 Trillion Risk
Kim identified the private loan market as the most dangerous structural vulnerability. The BIS estimates the market at a minimum of $2 trillion, though the true figure is unknown because the financing sits outside regulated banking channels. A recent National Bureau of Economic Research paper by University of Pennsylvania professors warned that big tech companies borrowed heavily for AI investments but are not generating sufficient returns, leaving private loan investors exposed.
"Credit is paramount in finance," Kim said. "If fear spreads that borrowed money cannot be repaid, a bank-run-like situation — where everyone rushes to withdraw funds first — could occur."
Kim pointed to SpaceX's June stock market listing as a warning signal. The Elon Musk-led company, which raised funds in both equity and bond markets on the premise of building AI data centers in space, now trades at $124, below its $135 offering price. Both stock and bond investors are underwater.
"If investors hesitate to provide funds, companies' financing becomes difficult, forcing U.S. AI firms to reduce investments," Kim said. "Next year may be the turning point."
Defensive Rotation Underway
Kim advised investors to lower return expectations and increase cash holdings, recommending short-term government bonds over U.S. equivalents given the still-elevated won-dollar exchange rate. He noted that Warren Buffett has increased Berkshire Hathaway's cash allocation to approximately 30% of assets.
"Stocks cannot always rise like that," Kim said. "It's time to significantly lower expected returns. If prices rise too much, they fall sharply afterward. This is my conclusion after analyzing the stock market on the front lines for 40 years."
The warning comes as analysts begin lowering price targets for semiconductor companies, a shift Kim described as a "narrative transition" from peak optimism toward a more sober assessment. Meta has reportedly explored leasing AI facilities to external customers after overinvesting, while Apple has lobbied the U.S. government to allow the use of Chinese-made memory chips to reduce costs — each report triggering sharp declines in semiconductor stocks.
This article is for informational purposes only and does not constitute investment advice.
