Seoul's 18% Rebound: Why Korea's Leveraged Washout Is Not a Bottom
South Korea's record rebound shows forced selling has eased, not that the market has found a durable floor. The next trade is about structure, not speed.
South Korea's benchmark index has just delivered the kind of rebound that tempts investors to declare the crisis over. The KOSPI surged 17.9% on Friday, its biggest daily gain on record, after losing more than 40% from its June peak, then gave back nearly 5% on Monday. Foreign investors bought 7.2 trillion won, or about $5 billion, on Friday, more than double the previous one-day record. The arresting conclusion is that the market has found its floor. The less comfortable one is that forced selling has eased while price discovery is still being run by leverage, concentration and the availability of a next buyer.
That distinction matters because the Korean selloff was never a clean referendum on the country's semiconductor earnings. It was a market-structure event that temporarily overwhelmed the fundamentals. Single-stock leveraged exchange-traded funds launched in May, drawing concentrated bets into Samsung Electronics and SK Hynix just as the two companies were carrying more than half of the KOSPI's weighting. When the trade reversed, the feedback loop worked in the other direction. The index became a high-beta expression of positioning before it became a signal about the economy.
Steve Lawrence, chief investment officer of Balfour Capital Group, told Reuters on Aug. 3: “This was a leverage event, not an earnings event.” That is the cleanest summary of the past two weeks. It does not mean the shares are cheap, or that the next move must be lower. It means the usual shortcut from a sharp price decline to a macro conclusion is unreliable here.
The first buyer is not the same as a new bull market
Friday's foreign buying is important, but it should not be mistaken for a broad restoration of conviction. Foreigners had been net sellers of Korean stocks all year before that record purchase. The timing suggests that some investors were covering shorts, re-establishing exposure after a forced unwind or buying a market whose index had become mechanically oversold. Average short interest, weighted by position value, was about 4.3%, down from a recent peak of 5.3%, according to S3 Partners data cited by Reuters.
The Monday reversal reinforces the point. A market that can rise nearly 18% and then fall nearly 5% in the next session has not returned to normal volatility. It has moved from a one-way liquidation into a two-sided trading regime. That is a better environment for specialists than for investors using the rebound as permission to chase the same crowded exposure that caused the damage.
The mechanics are unusually visible. Reuters reported that the Hong Kong-listed two-times leveraged ETF tracking SK Hynix had fallen 83% in a month from its late-June peak, while still holding HK$31.9 billion, or roughly $4 billion, in assets. Samsung and SK Hynix together accounted for more than 80% of KOSPI trading volume on some days this year, according to Reuters calculations. The KOSPI volatility index remained above 80 for six weeks and hit a record 97.99 on June 19, a level that makes conventional valuation work look precise only by accident.
Michael Green, chief strategist and portfolio manager at Simplify Asset Management, described the dynamic to Reuters on July 29: “The combination is creating an incredible feedback loop that's driving volatility in the semiconductor space.” The important word is combination. The chip companies are not irrelevant; their earnings determine the long-run opportunity. But the ETF flows, daily rebalancing and index concentration determine how far prices can travel before those earnings are allowed back into the conversation.
South Korea's regulators now have to manage the after-effects without pretending that product rules can erase a global risk appetite cycle. The government proposed limiting an individual's investment in single-stock leveraged ETFs to 20% of total investment assets, raising trading costs and requiring a minimum cash deposit of 30 million won, or $20,646. New listings were halted and advertising restricted. Those steps may reduce the next wave of retail leverage. They cannot force existing holders to sell, prevent similar products listed in Hong Kong or New York from transmitting flows into Korean shares, or restore confidence among investors who have just watched a large part of a market's value disappear and reappear in days.
Kim Jin-wook, an economist at Citi Korea, told Reuters on July 30 that a liquidity put such as a market-stabilisation fund would have a greater effect than the announced measures. The comment points to the policy problem beneath the ETF headlines. Seoul is trying to curb leverage after the fact, but the market's deeper vulnerability is the absence of a credible liquidity architecture for a concentrated index. A cap can slow the next acceleration; it does not automatically create a buyer when volatility is highest.
Strong chips, weaker price signals
None of this cancels the underlying export and earnings story. South Korean exports rose 62.8% year over year in July to $98.89 billion, according to preliminary data reported by Reuters on Aug. 1. The gain beat the 59.0% median forecast in a Reuters poll and was powered by global demand for chips and computers. That is a genuine macro signal, and it is precisely why the market's price action is so difficult to read: the economy can be improving while the index is dislocating.
Quarterly results tell the same story with a warning attached. SK Hynix's April-to-June operating profit rose 557% to a record 60.5 trillion won, yet missed the 64 trillion won LSEG SmartEstimate. Revenue rose 257% to 79.3 trillion won, below an 84 trillion won estimate. In a market priced for near-perfect execution, an enormous profit increase can still be treated as a disappointment. Gary Tan, portfolio manager at Allspring Global Investments in Singapore, told Reuters on July 29: “SK Hynix delivered strong results, but in today's AI market, strong is no longer enough.”
“Investors were looking for additional catalysts, particularly around long-term agreements and shareholder returns, to support a memory sector that has become the epicentre of the AI trade. Without those signals, we expect volatility in AI-linked equities across Asia to persist as leveraged positions unwind and the market resets expectations.” — Gary Tan, portfolio manager at Allspring Global Investments, quoted by Reuters, July 29, 2026.
That is the contrarian point for equity investors. The rebound can coexist with a lower valuation multiple and stronger exports, but a better fundamental backdrop is not the same thing as a stable market. When two companies dominate an index, the market can be directionally right about earnings and still wrong about the path. The first question is no longer whether Samsung and SK Hynix participate in a memory upswing. It is whether the ownership base can hold those positions without another round of forced deleveraging.
There are already signs that the market is trying to separate those questions. Morgan Stanley upgraded South Korean equities to overweight after the washout and put a 9,000 target on the KOSPI, implying 36% upside, according to Bloomberg on Aug. 2. That view may prove right if the leverage has genuinely been cleared and the export cycle remains intact. But the target is a twelve-month statement about earnings and valuation, not a guarantee that the next week will be calm.
The better way to read the foreign buying is as an attempt to own the reset, not a verdict that the reset is complete. Investors who bought on Friday may be expressing confidence in semiconductor cash flows, but they may also be trading the technical consequences of a market where short interest has fallen, leveraged holdings have shrunk and the most liquid names have become the easiest way to re-enter Asia risk. Those motives can produce the same green candle and very different outcomes afterward.
Our view is that the KOSPI should be treated as a post-leverage regime change, not a conventional dip-buying opportunity. Selective exposure to companies with visible cash generation, durable customer commitments and credible shareholder-return policies can make sense. A broad rush back into the index does not. The data say the Korean economy is still strong; the tape says the market's plumbing is still fragile. Until those two messages converge, the risk is not missing the first day of the rebound. It is mistaking the first buyer for the last buyer.
This note is for informational purposes only and does not constitute investment advice. Figures and quotations are attributed to the Reuters and Bloomberg reports linked in the text, dated July 29 to August 3, 2026.
