
- KOSPI Circuit Breaker: How the Historic July 2026 Sell-Off Unfolded
- Why Did the KOSPI Crash? Five Drivers Behind the Sell-Off
- Why SK Hynix Stock Fell Despite Record Q2 2026 Earnings
- Why South Korea Apologised for Single-Stock Leveraged ETFs
- How Retail Selling and Margin Calls Deepened the KOSPI Crash
- How Leveraged ETFs Magnified SK Hynix and Samsung Sell-Off
- Is the KOSPI Crash a Normal Correction or a Structural Breakdown?
- What Analysts Say About the KOSPI, SK Hynix and Samsung Stocks
- Is the KOSPI Cheap After the Crash? Valuation Explained
- Five Risks to Watch After the KOSPI Crash
- What the KOSPI Crash Means for Investors
South Korea's finance minister stood in front of parliament on Wednesday and apologised. Not for a scandal, not for a policy failure in the usual sense, but for a stock market product his own government approved nine weeks earlier.
A few hours later, SK Hynix, the company at the centre of the entire Kospi story, reported a 557% jump in quarterly operating profit, the best quarter in its history, and its stock still closed down 9.6%. The Kospi itself triggered a circuit breaker for the second day running, something that had never happened before in the index's history, down roughly 16% in two trading days and about 40% from its June peak.
Let's break down why a record profit and a government apology landed on the same day, what a little-known corner of Korea's ETF market has to do with both, and whether the maths says this is a scary but ordinary correction in KOSPI or something that has actually broken.
KOSPI Circuit Breaker: How the Historic July 2026 Sell-Off Unfolded
Wednesday started deceptively. The Kospi opened 1.09 percent higher at 6,089.11, a small relief rally after Tuesday's 10.84 percent rout. It did not last. A sell-side sidecar, a mechanism that briefly pauses program sell orders, kicked in around 10:55 a.m. local time, and by early afternoon the index was down as much as 12.63% intraday to 5,262.77, its lowest level in almost six months. A circuit breaker halted all trading for 20 minutes at 12:32 p.m., and the Kosdaq, Korea's smaller-cap index, tripped its own breaker minutes later after falling more than 8%.
According to the Korea Exchange, this was the first time in the market's history that circuit breakers had fired on two consecutive trading days, and the first time the Kospi and Kosdaq had both halted on the same day.
| What happened | Detail |
| Kospi close | 5,663.24, down 5.98% |
| Kospi intraday low | 5,262.77, down 12.63% |
| Circuit breaker | Triggered 12:32 p.m. KST, second straight day, a first |
| Kosdaq close | 662.68, down 6.12% |
| Two day index loss | Roughly 864.5 trillion won in market value |
| Retail net selling | Around 2 trillion won ($1.38 billion) |
| Foreign net selling | Around 1.2 trillion won |
Samsung Electronics and SK Hynix, which together make up somewhere between half and 60% of the Kospi's weight depending on the day you check, both fell by double digits at one point before paring losses. Samsung closed down 5.23%, and SK Hynix closed down 9.61%, after falling as much as 17% intraday, largest single-day drop on record before buyers stepped back in during the afternoon.
The decline from the index's June peak, hit barely five weeks earlier, is more than 40%. That is the headline everyone already has. The more useful question is what actually caused a company's best quarter ever to be treated by the market like bad news, and why the government felt it had to apologise on the same afternoon.
Why Did the KOSPI Crash? Five Drivers Behind the Sell-Off
Five forces interacted together which led to the massive sell-off in Korean Equities:
- A fundamental trigger
- Index concentration
- Leverage
- A buyer strike
- A policy-confidence shock
The first layer explains the repricing. The next three explain its violence. The fifth explains the government response.
Why SK Hynix Stock Fell Despite Record Q2 2026 Earnings
At 9 a.m. Korean time, hours before the worst of the selling, SK Hynix released preliminary second quarter results that were, by any normal standard, extraordinary.
| Metric | Q2 2026 | Change |
| Revenue | 79.32 trillion won | Up 257% year-on-year |
| Operating profit | 60.54 trillion won | Up 557% year-on-year, 76% margin |
| Net profit | 93.92 trillion won | Up over 1,200% year-on-year |
| First half revenue | Crossed 100 trillion won | First time in company history |
One number in that table needs a flag before anyone gets excited about it: net profit of 93.92 trillion won is actually larger than operating profit, an unusual pattern that typically points to a large one-off gain sitting below the operating line. SK Hynix's release does not spell out exactly what that gain was, so treat the headline net profit figure as inflated by something outside the core memory-chip business rather than as pure operating strength.
Read Full SK Hynix Earnings and Stock Fall Analysis Here
Even so, the operating numbers alone would normally be celebrated. TradingKey and Reuters both reported that analysts had penciled in roughly 84 trillion won in revenue and 64 trillion won in operating profit, so the actual print, spectacular as it was in isolation, still missed what the market had already priced in.
On a same-day call, SK Hynix executives said they saw no signs of a slowdown and that keeping up with customer demand, not finding it, was now the company's biggest problem. The stock fell anyway. This is the same pattern that played out with Samsung Electronics on July 7, when its own 19 fold jump in preliminary operating profit to about 89.4 trillion won still triggered a 6.9% share price decline because revenue landed just under consensus.
Why South Korea Apologised for Single-Stock Leveraged ETFs
One part of Wednesday’s KOSPI crash deserves more attention. As the market continued falling, Finance Minister Koo Yun-cheol apologised for regulators’ handling of single-stock leveraged ETFs linked to Samsung and SK Hynix. These products aim to magnify a stock’s daily movement, often by two times, making both gains and losses much larger.
“The country has turned into a casino,” lawmaker Lee Jongwook told Koo, calling the rollout a policy failure. Koo said existing measures could be strengthened further if needed to stabilise the market. The speed of the policy reversal explains the backlash:
| Date | Development |
| April 28 | Single-stock ETFs were permitted |
| May 27 | First leveraged products began trading |
| July 16 | New listings and advertising were halted |
| July 31 | Minimum cash requirement rises from ₩10 million to ₩30 million |
| July 29 | Regulators apologised and held an emergency meeting |
The FSC originally argued that domestic products would offer better protection than overseas alternatives, widen investor choice and reduce money flowing abroad. However, introducing them into a market dominated by a few large stocks and increasingly leveraged retail investors made price swings more severe.
Regulators have now:
- Halted new leveraged, inverse and covered-call single-stock products.
- Banned their advertising.
- Increased mandatory investor education and added tougher tests and loss warnings.
- Raised the cash requirement to ₩30 million, with stocks and bonds no longer accepted as substitutes.
- Proposed increasing the minimum trade from one share to 20.
Some restrictions are already active, while others are still being implemented. Individual investment limits also remain under consideration. On July 29, Korea’s finance minister, central bank governor and top financial regulators held their second emergency meeting in two weeks. Authorities are now reportedly considering restricting these products to professional investors entirely.
How Retail Selling and Margin Calls Deepened the KOSPI Crash
Foreign selling is common during risk-off periods. The bigger warning on July 29 was retail investors, who had cushioned the earlier decline, also turning into sellers.
After buying a net ₩11.47 trillion of KOSPI shares over three sessions, individuals sold ₩1.97 trillion on Wednesday. Foreign investors sold another ₩1.23 trillion that day and ₩11.94 trillion over four sessions. With both groups exiting, the market lost a crucial source of demand.
Retail investors were also heavily indebted. Margin loans, or money borrowed directly to buy shares, reached a record ₩38.63 trillion on June 24 and remained at ₩34.37 trillion on July 15. Total investor debt exceeded ₩60 trillion by May-end.
Goldman Sachs estimated that more than 1.2 million leveraged retail accounts had received margin calls by July 13, with 320,000–360,000 already liquidated. These estimates predated the latest crash.
The key concern was therefore not just foreign outflows. Korea’s indebted dip-buyers were losing the capacity to keep supporting the market. Their reversal could mark final capitulation or leave more leveraged positions vulnerable. One rebound cannot determine which.
How Leveraged ETFs Magnified SK Hynix and Samsung Sell-Off
A separate risk came from single-stock leveraged ETFs. Unlike margin loans, where investors borrow directly, these products build leverage into the fund itself.
A 2x ETF targets twice a stock’s daily return, not twice its long-term return. To reset that leverage each day, the fund or its hedging partners reduce exposure after the stock falls and increase it after it rises. This automatic rebalancing can add selling pressure during a downturn without investors actively choosing to sell. Goldman Sachs attributed an earlier phase of the decline to this type of program-driven, mechanical selling rather than weaker memory-chip fundamentals.
Daily resetting also creates volatility drag:
| Scenario | Day 1 | Day 2 | Net result |
| Stock falls 10%, then rises 10% | 100 → 90 | 90 → 99 | Down 1% |
| 2x leveraged ETF | 100 → 80 | 80 → 96 | Down 4% |
The stock ends almost unchanged, but the ETF loses four times as much because daily returns compound from a lower base.
Korea Exchange data show that some inverse ETFs designed to profit from declines in Samsung and SK Hynix lost up to 31% after launch, even though both stocks fell overall. Investors predicted the direction correctly but still lost because the sequence of daily moves mattered more than the final price.
By July 20, retail investors had bought a net ₩14 trillion of these products, compared with ₩2 trillion from foreign investors. The largest of the 16, the KODEX SK Hynix Single Stock Leverage ETF, fell 45%–47% despite SK Hynix reporting record profits.
These losses demonstrate the risks of daily-reset leverage, although they do not prove that ETFs alone caused the wider crash. They also explain why Korean regulators apologised and tightened the rules. India avoids this specific risk by not permitting single-stock leveraged ETFs for retail investors.
Is the KOSPI Crash a Normal Correction or a Structural Breakdown?
The honest answer, based on the evidence available, is that both a sharp ordinary correction and a real structural wobble are happening at once, and it is worth separating them rather than picking one story.
The case for "ordinary, if unusually fast, correction" rests on scale and history. The Kospi gained roughly 297% from its April 2025 low to its June 2026 peak, its strongest bull run since the 1980s, according to data compiled by Seeking Alpha. A correction of 33% to 34% from that kind of peak, which is roughly where the index sits now, is close to the average Kospi bear market decline of 33.5% going back decades.
What is unusual is not the depth so much as the speed: this one arrived in about five to six weeks, versus a historical average bear market length of 229 days. Judged purely on size, this looks like a normal, if compressed, bear market.
| Event | Move | Time to fall |
| 1997 Asian Financial Crisis, worst month | Down 27% | One month |
| March 4, 2026 single day, unrelated to chips | Down 12.1% | One session |
| July 2026 chip selloff, current | Down 32% to 33% | About five weeks |
| Average Kospi bear market | Down 33.5% | 229 days on average |
The case for something genuinely structural rests on three things that are not just sentiment.
- Concentration: Samsung and SK Hynix's combined share of Kospi market value has been reported anywhere from roughly half to roughly 60 percent depending on the source and the week, up sharply from about a quarter at the end of last year, so two earnings prints now move the entire country's index.
- The leveraged product mechanic described above, which manufactures extra selling regardless of what the underlying fundamentals are doing.
- A market microstructure problem regulators themselves are now admitting to in public: products introduced in May, at the peak of enthusiasm, that neither the industry nor the regulator appears to have stress tested for a scenario where both stocks fall together.
None of this proves the AI memory story itself is broken. HBM4 shipments began in the second quarter and SK Hynix's own management says demand still exceeds supply. What it does suggest is that a meaningful share of the last two days' violence was Korea's own market plumbing amplifying a real but more modest earnings disappointment into a historic rout.
What Analysts Say About the KOSPI, SK Hynix and Samsung Stocks
Wall Street's own record on calling the bottom of this specific selloff has been shaky, which is itself useful information.
Morgan Stanley, previously nicknamed the "grim reaper of semiconductors" in Korean market commentary for its bearish memory calls, flipped to calling the correction a buying opportunity around July 22, only for the index to fall further afterward.
That does not make every bullish call wrong this time, but it is a reason to read every fresh "buy the dip" note with a little more scepticism than usual until the price action actually confirms it.
| Firm or analyst | Stance around the current selloff | Stated reasoning |
| Morgan Stanley (Joseph Moore), index target 9,000 | Reversed an earlier bearish stance to call the pullback a buying opportunity, around July 22 | Cited intensifying memory shortages projected into 2027 to 2028 |
| JP Morgan, index target 12,500 | Maintained a 12 month index target well above the June record high, as of late July | Has not been revised down publicly to reflect this week's fall, so treat the gap between target and price with real caution |
| Kiwoom Securities (Han Ji-young) | More cautious | Said investors now want proof that AI capital spending keeps growing, not just a good quarter, and warned the market keeps finding new reasons to talk itself into selling |
| Fibonacci Asset Management (Jung In Yun) | Focused on who is bearing the losses | Said domestic retail investors have absorbed most of the damage from leveraged products bought as long-term holdings rather than short-term trades |
Two things stand out. Analysts broadly agree the underlying HBM and DRAM demand story has not collapsed. They disagree, sometimes within the same firm over a matter of weeks, on whether the price has now overshot the fundamentals to the downside or is still catching down to them.
J.P. Morgan believes the KOSPI may be nearing a bottom as leveraged-ETF unwinding is likely complete and hedge-fund deleveraging is roughly 90% done. It remains bullish on Korean stocks, citing attractive valuations, resilient earnings and opportunities beyond chipmakers.
A 12,500 index target sitting on the books at a firm while the index trades near 5,663 is not necessarily wrong, but it is a reminder that not every published target has caught up with the last ten trading days, and it is worth checking the date on any target before leaning on it.
Is the KOSPI Cheap After the Crash? Valuation Explained
As of July 1, before this week’s crash, the KOSPI traded at 22.95 times past earnings but only 7.82 times expected earnings. This wide gap showed that analysts expected profits to rise sharply. Even near June’s record highs, the index looked historically cheap based on forecasts, approaching 2008 crisis-era levels in some data series.
However, the same figure supports both the bull and bear cases. If Samsung and SK Hynix sustain their current HBM and DRAM profits, the KOSPI is attractively valued. If memory prices normalise, 7.82 times may be based on peak earnings and understate the market’s true valuation.
Korean stocks already trade at a long-standing discount because of governance concerns, concentrated ownership and average returns on equity of roughly 7% over the past decade. The low valuation therefore reflects both this structural “Korea discount” and doubts about whether record chip profits can continue. Samsung’s full results and the next few SK Hynix quarters should provide greater clarity.
Five Risks to Watch After the KOSPI Crash
| Risk | Why it matters |
| Leveraged ETF unwind is not finished | With 16 products still live and some Korean commentary suggesting only a partial liquidation of leveraged positions so far, another sharp move in either stock can still trigger fresh forced selling on top of it |
| Samsung's July 30 segment report | If the mobile and display businesses show the operating losses some analysts already expect, it complicates the "one clean AI story" narrative even if chips stay strong |
| A hawkish Fed outcome later on July 29 | The Fed's decision lands after Asian markets close and was not yet known as this was written, but a hawkish surprise would tighten conditions for exactly the kind of high-multiple trade that built this rally |
| China's memory ambitions | Chinese DRAM capacity has been expanding fast even if it is not yet matching HBM's technical bar, and it pressures the commodity end of Samsung and SK Hynix's business regardless of the AI story |
| Regulatory overcorrection | If Seoul restricts these leveraged products too abruptly, it could remove a source of retail liquidity precisely when the market can least afford another supply shock, a risk regulators themselves have flagged |
What the KOSPI Crash Means for Investors
The AI-led memory upcycle has not visibly broken. Two earnings disappointments triggered the sell-off, but Korea’s leveraged retail ecosystem turned it into a historic, self-reinforcing rout, a mechanism regulators now publicly acknowledge.
The bull case remains intact: Samsung and SK Hynix reported record, though below-consensus, profits; forward valuations were near multi-year lows even before the crash; management says demand still exceeds supply; and regulators are acting to curb forced selling.
The bear case is concentration and durability. Korea’s reliance on its two largest chipmakers has, by some measures, doubled in seven months. Goldman Sachs and Korea’s finance minister have linked leveraged products to mechanical selling, while the governance problems behind the long-standing “Korea discount” remain. Samsung’s July 30 results are the next major test.
The key question is whether Samsung’s segment results and SK Hynix’s next quarter confirm intact demand and a leverage-driven overreaction, or reveal the first turn in the memory cycle. EWY, FLKR and Nasdaq-listed SK Hynix shares do not carry daily-reset leverage. However, their investors still face the same concentration risk.