The number that should stop you is not the 2.70%. It is the shape of the buying.
On Monday, September 28, South Korea's KOSPI closed at 6,889.75, down 2.70% as markets reopened after the Chuseok holiday. Foreign investors sold 3.23 trillion won — roughly $2.4 billion. Institutions, led by financial investment, investment trust and pension funds, added another 1.02 trillion won, about $750 million, to the exit. Retail investors bought.
Early in the session, their net purchases sat near 145 billion won, about $106 million. By the close, that figure had reached 2.21 trillion won, roughly $1.6 billion. The KOSDAQ, Korea's secondary board, closed slightly higher.
Read the progression twice. Retail did not make a decision at the open and hold it. They escalated into the close. Whatever floor was built on Monday was constructed in the final hour, not the first, and that tells you more about the market's structure than any single price level.
Context: who the ants are, and why they matter
Korean individual investors have a nickname for this behavior: the ants. The label dates to 2020, when they absorbed a wave of foreign selling during the pandemic crash and were largely rewarded by the rebound that followed. It stuck because it described something real — a retail base that treats falling prices as an invitation rather than a warning, and that shows up in quantities large enough to move a national index.
South Korea is not a casual retail market. It is the country that gave the world the Kimchi premium, the persistent gap between local and offshore crypto prices that became a case study in how reflexive retail demand can distort a market's relationship to its underlying asset. The behavior on display Monday is not new. It migrated. A population trained on local crypto exchanges, on five-minute candles, on the emotional rhythm of a twenty-four-hour market, carried those reflexes into equities — an inheritance that the ETF era mostly erased from the crypto side of the ledger and quietly deposited somewhere else.
That is the context you need before reading the trade.
Core: the volatility inversion
Here is the part I keep returning to, and the part the headlines skipped. The KOSPI has been more volatile than Bitcoin this year.
Sit with that. I have spent most of my career explaining to non-technical investors that crypto is the volatile asset — the one you size down, the one you treat with respect, the one that will humble you if you mistake conviction for a hedge. During the 2022 drawdown I watched that framing save ordinary people from decisions they could not recover from. The point was never that crypto is bad. The point was that risk must be anchored to reality rather than to narrative.
The anchor has moved. Post-ETF Bitcoin has compressed its realized volatility. The approval cycle pulled it into an institutional regime, and institutional holders dampen moves rather than amplify them; the sheer weight of steady, unemotional allocation has changed the shape of the order book. Meanwhile, the two semiconductor names at the heart of the KOSPI — Samsung Electronics and SK hynix — both fell more than 5% on Monday, extending swings that have made the index itself carry more day-to-day variance than the asset the world still calls risky.
That is not a rhetorical flourish. It is a change in where risk actually lives, and most portfolios have not been re-underwritten for it. The volatility did not leave crypto and reappear in equities. It rotated, driven by the same force in both places: the arrival of large, structurally different holders who rewrote the mechanics of price discovery.
The SK hynix leveraged exchange-traded fund that lost 45% of its value earlier this year is the clearest exhibit in the file. Retail investors held it as a long-term position rather than a short-term trade. That is a category error, and it is one I have watched destroy portfolios in crypto for a decade. A leveraged product is a decaying instrument. It is a bet on a path, not on a destination. Holding it through a drawdown and calling that conviction is not patience; it is mistaking the shape of a trade for the shape of an investment. The people who held it were not reckless. They were told, implicitly, that a leveraged wrapper is a better version of the underlying. It is not. It is a different animal with a clock in its chest.
The intraday structure deserves a closer examination, because it is the kind of signal I would flag to students in a sideways market. Net buying that starts small and compounds into the close is rarely a single thesis. It is a sequence of small surrenders by sellers meeting a sequence of small buys by people averaging down, each participant a little more certain than the last. That is how a floor gets built — not by a decision, but by attrition. It is also how a trap gets built. The difference between the two is visible only afterward, which is precisely why the ants' conviction cannot be scored in real time. For anyone navigating a consolidation regime, the useful question is not whether retail wins. It is where the marginal seller lives, and whether it is the same wallet that shows up again tomorrow.
Which brings me to the framing both sides of Monday's trade will reach for. Foreign investors blamed a stronger US dollar and short-term doubt about Korea's semiconductor cycle. Institutions agreed and joined the selling. Retail read the same headlines and reached the opposite conclusion. But "retail versus institutions" is a comfortable story that hides the mechanism. The passive and index products wrapped around these same names are increasingly retail-funded, and the pension flows that sold on Monday are the same flows that carried prior rallies. The ants were not betting against a crowd. In aggregate, in a market this retail-dense, the ants are a large part of the crowd.
Contrarian: the precedent is being misapplied
Here is the blind spot. The 2020 template — ants absorb the selling, market rebounds, ants are vindicated — is being applied to present conditions as if the two situations were equivalents. They are not.
In 2020, the selling was a liquidity shock against a recovering macro backdrop. The dip-buy worked because the thing being sold was temporarily mispriced, not structurally repriced. Today, the same semiconductor names are being sold by the holders who tend to be right about cycle timing. The 2020 playbook was written in a market where institutions and retail wanted the same asset at different moments. This time, the exit looks less like a pause than a rotation out of a narrative, and the buyer on the other side is the participant least equipped to price a cycle it did not create.
Solidarity over speculation is easy to say when the tape is climbing. It means something harder when the community consensus is itself the crowded trade. The ants are not wrong to buy. They are wrong to believe that buying together is the same thing as buying correctly. Those are two different sentences, and only one of them is a strategy.
Takeaway
Watch Samsung and SK hynix first. If the ants picked a floor, both names stabilize within days and the KOSPI retests 7,000 quickly. If they did not, Monday's foreign selling may only be the opening move, and every optimistic target gets harder to reach. Code is law, but ethics is conscience — and the honest position is that nobody, least of all the enthusiastic buyer, knows yet how this resolves. Culture on-chain, heart on-screen.