Selling Pressure on Korean Stocks Eases as Multiple Institutions Maintain Overweight
Alina Collins
KOSPI fell roughly 20% in July — its worst month since 2008 — but JPMorgan, BofA, and TS Lombard say the sell-off is nearing exhaustion, maintaining overweight and framing Korea as a lead indicator for US tech.
Was this a fundamental breakdown or a leverage blowup?
JPMorgan classified the sell-off as a liquidity shock, not a deterioration in fundamentals. This means → companies are still earning; the crash was driven by leveraged funds forced to liquidate simultaneously.
Since the June 22 peak, KOSPI has pulled back roughly 26%, driven by leveraged-ETF deleveraging, hedge-fund unwinds, and passive foreign outflows.
In plain terms = the plumbing broke, not the economy.
How extreme did the leveraged-ETF bubble get?
At the late-June peak, leveraged ETFs benchmarked to Korean assets hit $50 billion in AUM — roughly 4× the US-equivalent ratio relative to market size.
These products created a textbook positive-feedback loop: buying into rallies, selling into declines. Korea's fear gauge spiked to nearly 5× the US VIX, versus a normal ratio of about 1×. This means → the leverage structure amplified panic far beyond what fundamentals warranted.
Forced liquidations have since shrunk leveraged-ETF AUM from $50 billion to roughly $26 billion. JPMorgan estimates the deleveraging is about 75% complete.
What has the regulator done to hit the brakes?
Korean regulators are tightening in stages: from August 5, minimum account size rises from ₩10 million to ₩30 million; from August 19, initial margin must be in cash only.
From November, the minimum trade size for single-stock leveraged products jumps from 1 share to 20 shares, and new single-stock leveraged ETF listings are suspended.
In plain terms = regulators are raising the bar so retail traders can no longer use pocket change to run outsized leverage — squeezing the bubble at its source.
$110 billion in foreign outflows — will they continue?
Korean equities have seen more than $110 billion in foreign outflows year-to-date, an all-time record for a single Asian market.
JPMorgan stresses that roughly 90% of the outflows were concentrated in two memory-chip giants, driven by passive rebalancing under MSCI EM index weight caps — not active selling. This means → foreigners were not turning bearish on Korea; index rules forced them to sell once weights exceeded limits.
As prices corrected, the two companies' MSCI EM weights fell from 9.5% and 8.3% to roughly 7.5% and 5.7%, materially reducing the passive selling pressure.
Why do multiple houses say the worst is nearly over?
TS Lombard framed the decline as semiconductor-stock deleveraging, not a broader market crash, noting that current prices already reflect roughly 45% normalization in memory-chip earnings next year.
BofA flagged a technical signal: the broad market's decline has, for the first time since March, exceeded the decline in its large-cap leaders — and the March instance marked a cyclical bottom. This reflects a shift from indiscriminate liquidation to selective adjustment, a classic sign of easing pressure.
JPMorgan maintains its overweight on Korean equities with a 12-month KOSPI target of 12,500.
Why is Korea treated as a lead indicator for US tech?
Korea's market is heavily exposed to memory chips — widely seen as a core beneficiary of AI infrastructure buildout. Put simply = the global AI boom needs data centers, data centers need massive amounts of memory, and Korea is the largest supplier.
JPMorgan sees hyperscaler data-center ROI still running high, with no visible slowdown in capex plans and no signs of actual memory-demand decline across the supply chain.
Despite July's sharp pullback, KOSPI is still up roughly 60% year-to-date in 2026. Whether deleveraging runs its course on schedule and whether the AI capex cycle holds will be the key tests for the 12,500 target.
Content is for reference only, not financial advice.