AI sells-off triggers first Korean emergency fund since 2008
South Korea is preparing to reactivate its stock stabilization fund after AI-related selling crashed markets again. The move signals a structural shift — machine-driven selloffs are no longer anomalies but recurring threats to East Asian economies.
The market no longer thinks in hours anymore.
It thinks in microseconds. And when it decides to sell, human hands can only watch.
On November 20, Wall Street sold off hard — Dow down 1.93 percent, S&P 500 down 3.70 percent, Nasdaq down 5.28 percent — and within hours the shockwave reached Seoul. By the next morning, overnight KOSPI 200 futures had dropped nearly 7 percent. Foreign investors were gone. The question now gripping Korean policymakers is simple: when algorithms drive the selling, who do you call in?
The answer, apparently, is everyone the government already has on speed dial.
On November 22, Deputy Prime Minister and Finance Minister Lee Hyung-il convened an emergency meeting with the heads of the Financial Services Commission, the Financial Supervisory Service, and the Bank of Korea. Their mandate was straightforward: prepare to activate the stock stabilization fund before Tuesday’s regular session opens. If history means anything, that’s going to be a historic moment — the fund hasn’t been deployed since the 2008 financial crisis, when it received just 5 trillion won and spent it all fighting to keep Korean banks alive.
This time the money would go straight into stocks, not balance sheets.
The government is moving fast on parallel fronts. Starting November 24, listed companies will get a three-month window to raise their daily self-share purchase limits — essentially letting firms buy back their own stock aggressively when the market can’t find buyers. Illegal short selling will face concentrated surveillance. These are familiar tools in any central banker’s playbook. But the urgency behind them reveals something more unsettling: the AI market crash may no longer be treated as a temporary anomaly.
Why AI is different from every other bubble.
Goldman Sachs’ Panic Index climbed to 9 out of 10 after Monday’s rout. That’s not a mild jitter. It’s a market signaling it expects more pain ahead. Bloomberg noted that investors are increasingly questioning whether the enormous capital being poured into AI infrastructure — semiconductors, data centers, computing capacity — will ever generate profits large enough to justify today’s valuations. The doubt isn’t abstract. It landed squarely on Nvidia.
The chipmaker dropped 14.01 percent on November 20 alone. Against its October 6 high, Nvidia is down nearly 48 percent. Compounding the trouble is lingering uncertainty around CEO Jensen黄 — his scandals and rumored departure have injected another layer of doubt into the company most synonymous with the AI boom. The Wall Street Journal ran with the headline that the AI sector itself is shaking. In Seoul, that shakes harder than anywhere else.
The Samsung problem.
Here’s what English-language coverage of these events consistently misses: Korea’s stock market isn’t just affected by global AI sentiment. It is the single market where that sentiment translates most directly into pain because of its concentration. Samsung Electronics and SK Hynix together still account for roughly 49 percent of total KOSPI market capitalization — barely down from a peak of 59 percent. When the AI narrative sours globally, these two companies don’t drift lower modestly. They pull the entire index down with them.
The KOSPI 200 volatility index — the V-KOSPI, Korea’s version of the VIX — closed at 53.18 on November 20, up over 6 percent from the prior week. That number tells you exactly where fear lives in this market: in two companies. A broader, healthier equity market distributes risk. Korea’s does not.
Morgan Stanley Capital International’s ETF tracking MSCI Korea shares shed 8.42 percent overnight, a clean signal that international capital is exiting Korean exposure ahead of domestic trading even beginning.
The liquidity trap.
Kim Seong-noh at BNK Investment & Securities offered the sharpest diagnosis visible so far. He cited Nvidia’s collapse and Anthropic’s recent listing fallout as twin signals that U.S. market liquidity simply isn’t what it was. Less liquidity means smaller trades move bigger prices. It also means foreign sellers can exit larger positions without finding offsetting buyers — which is exactly what appears to be happening in Korea right now.
“Given this liquidity deterioration,” Kim said, “foreign selling in the Korean market will likely continue for some time.”
That’s a prognosis, not a wish. And it carries weight because Kim is describing a structural problem, not a temporary one. Algorithmic funds managing AI exposure don’t operate on national loyalties. When their models flag sector stress, they liquidate simultaneously across correlated markets — and Korea happens to sit directly in the crosshairs because of Samsung and SK Hynix’s outsized role in the semiconductor supply chain.
What happens next.
All eyes are now on Thursday’s U.S. personal consumption expenditures price index for October. The Federal Reserve treats it as its primary gauge when setting rates. If inflation comes in hot, the prospect of further rate hikes — and rising Treasury yields — will compress valuations further and make equity rebounds fragile at best. Two earnings reports add to the pressure: Hewlett Packard Enterprise on Wednesday and Dell Technologies on Thursday. Both are proxy tests for whether AI demand is real or merely reflected in stock prices.
South Korea’s government has laid out three defensive moves — the stabilization fund, relaxed buyback rules, and short-sale monitoring. Between them, they offer a partial shield. But none of them address the root asymmetry: an economy whose public markets depend overwhelmingly on two firms riding one technology wave, facing a global investment base that now sells faster than it ever has before.
The era of manual market correction may be over. The era of algorithmic shock transmission has arrived. Korea’s emergency response is a reasonable first step. Whether it’s enough is the question no stabilization fund can answer.