The AI trade that powered South Korea's stock market to a $5.1 trillion peak has reversed with brutal speed. According to a report in The Hindu BusinessLine by Nishanth Gopalakrishnan, the benchmark KOSPI Composite Index lost 22 per cent in July — its third-worst month on record, after the 27 per cent crash in October 1997 and the 23 per cent plunge in October 2008. The report notes that $1.2 trillion of the country's market capitalisation has been erased in a matter of weeks, a brief demonstration of the possible fallout if the AI trade were to go south across the globe.
South Korea's concentrated AI bet
The rout is concentrated in the two semiconductor giants that dominate the Korean bourse. According to the report, SK hynix and Samsung Electronics account for roughly 50 per cent of KOSPI companies' market capitalisation. Both slumped to intra-week lows in July — SK hynix down 20 per cent and Samsung Electronics down 24 per cent versus the previous week's close. The report says these were the very stocks that had taken the Korean market to record highs.
The damage was intensified by single-stock leveraged ETFs, high-risk funds that use derivatives to multiply daily returns. The report highlights the CSOP SK Hynix Daily (2x) Leveraged Product ETF, which gains or loses 2 per cent for every 1 per cent move in SK hynix. From its 52-week high on June 25, the ETF has lost over 78 per cent. Such products have wiped out the portfolios of thousands of Korean investors, the report says, and the situation became so dire that the country's finance minister apologised and admitted the products were introduced "without careful consideration."
| Market event | Decline |
|---|---|
| KOSPI, July 2026 | 22% |
| KOSPI, October 1997 (Asian financial crisis) | 27% |
| KOSPI, October 2008 (global financial crisis) | 23% |
| S&P 500, 2000–2002 dotcom crash | 50% |
| Nasdaq Composite, 2000–2002 dotcom crash | 78% |
AI front-runners lose $6 trillion
The report tracked 16 front-running AI stocks — spanning hyperscalers, chip design (Nvidia, Broadcom), semiconductor manufacturing, neoclouds (CoreWeave, Nebius) and AI investor SoftBank. From their 52-week highs, these stocks are down 30 per cent on average, erasing about $6 trillion of investor wealth, according to The Hindu BusinessLine. Oracle is one of the starkest examples: after announcing Q1 FY26 results in September, the stock zoomed about 43 per cent to a 52-week high; it has since corrected 62 per cent.
S&P 500 bubble math
The report says the AI unwind is no longer a Korea-only story. Many of the 16 AI constituents are part of the S&P 500, accounting for about 30 per cent of the index's total market-cap and earnings. Consensus estimates show the index's total net income moving from $2.1 trillion in 2025 to $2.9 trillion in 2026, and the AI constituents account for one in every three dollars of the incremental $840 billion in earnings, the report states.
The valuation picture is stretched: the S&P 500's price-to-earnings multiple based on CY25 net income is 33x, a level the report calls "bubble territory." Based on CY26 earnings estimates, the P/E cools to 23x. This expected earnings growth is the "thin line" dividing the bulls and the bears.
If these companies fail to meet earnings expectations, the index being in bubble territory brings back memories of the dotcom crash in which the S&P 500 corrected 50 per cent and the Nasdaq Composite 78 per cent.
Earnings beats, but capex worries
If earnings are strong, what is troubling the market? The report notes these companies have delivered earnings beats "almost all the time in the last four quarters." The problem, it says, appears to be capex of "astronomical proportions." Alphabet is cited as a case in point, according to the report.
The dotcom deja vu is hard to ignore, and the report suggests the outcome depends on whether the AI cohort can convert heavy capital expenditure into earnings growth that justifies current valuations. For investors, the South Korean experience with leveraged ETFs is a cautionary example of what concentrated, leveraged exposure to AI stocks can deliver in a correction.
Sources:
