When the US 10-year Treasury yield rises, Korean equities get hit three times over: growth-stock valuations compress as the global discount rate resets, the won weakens as the US–Korea rate gap widens, and foreign investors sell to escape a double loss on price and currency. And the speed of the move matters more than the level — traders often treat a 10-basis-point-plus overnight jump in the 10-year as far more likely to unsettle the KOSPI open than a slow grind to the same level.
The question is topical for a reason. In mid-September 2026, foreign investors logged a fourth consecutive session of net selling on the KRX on September 14, with the won closing around 1,345 per dollar that day. This guide explains the machinery behind weeks like that one — machinery that works the same way in any year.
Channel 1: The Discount Rate and Growth-Stock Duration
The US 10-year Treasury yield is the closest thing global markets have to a universal risk-free rate. Every equity valuation model — from a simple discounted cash flow to an analyst’s target multiple — discounts future profits back to the present using a rate built on top of it. When that rate rises, the present value of distant cash flows falls, and the further out the cash flows sit, the harder they fall.
This is why Korean technology and growth names are the first casualties of a yield spike. Samsung Electronics and SK Hynix are priced not on this quarter’s earnings but on expectations for multi-year memory cycles and AI-related capital spending. The KOSDAQ — Korea’s growth-heavy secondary board, populated by biotech, battery-materials, and software names whose profits often lie years in the future — is even more rate-sensitive than the large-cap KOSPI.
The actionable distinction is why yields are rising. If the 10-year climbs because US growth expectations are improving, cyclical exporters can partially absorb the valuation hit with better earnings forecasts. If it climbs because of inflation fears or term premium — investors demanding more compensation simply for holding long bonds — there is no earnings offset, and the selling tends to be indiscriminate. Before reacting to a yield move, check whether inflation expectations or real yields drove it.
Channel 2: The Dollar Channel and the Won
Interest-rate differentials steer currencies. When US rates sit well above Korean rates, holding won-denominated assets carries an implicit penalty: capital can earn more, at lower risk, in dollars. Rising US yields therefore tend to push USD/KRW higher — for reference, the pair stood at 1,344.64 at the close on 2026-09-14.
For a foreign investor benchmarked in dollars, a rising USD/KRW is a direct loss on every Korean holding, independent of what the stock does. A falling KOSPI combined with a weakening won is a double loss — and this can encourage some foreign investors to reduce exposure, which itself adds to won-selling pressure. This is the reflexive loop at the heart of most sharp Korean drawdowns: yields up, won down, foreigners sell, won down further.
How to use this: read the daily foreign-flow print and the USD/KRW move together. When foreigners are net selling and the won is weakening on the same days, the combination can indicate elevated downside and currency risk, though it does not reliably determine subsequent returns. When flows turn negative but the won holds steady, the selling is more likely idiosyncratic — index rebalancing, a single block trade — and less likely to persist.
Channel 3: Reading the Foreign Flow Print — A Worked Example
Korean flow data is quoted in eok won (KRW 100 million, roughly $70,000–80,000 at typical exchange rates). The daily print for foreign net buying across KOSPI and KOSDAQ is the single most-watched flow number in the Korean market. Here is an actual stretch of completed sessions from September 2026:
| Session (2026) | Foreign net flow (eok won) |
|---|---|
| Sep 2 | -19,173 |
| Sep 3 | +414 |
| Sep 4 | +4,796 |
| Sep 7 | +25,532 |
| Sep 8 | +6,316 |
| Sep 9 | -4,348 |
| Sep 10 | -26,217 |
| Sep 11 | -22,984 |
| Sep 14 | -33,363 |
Take the last row and read it step by step:
- The raw figure: foreigners net sold 33,363 eok won of Korean equities on 2026-09-14.
- Convert the unit: multiply by 100 million — that is roughly KRW 3.34 trillion in a single session.
- Translate to dollars: at the 2026-09-14 USD/KRW close of 1,344.64, that is approximately $2.5 billion of net selling in one day — a large daily print.
- Check persistence and acceleration: the table shows a swing from strong buying (+25,532 on Sep 7) to four consecutive selling sessions from Sep 9, totaling 86,912 eok won — roughly KRW 8.7 trillion. The selling was not only sustained; it was accelerating, with the largest print coming last.
- Interpret: one negative day is noise — arbitrage unwinds, dividend-related flows, and index rebalancing all distort single prints. A multi-session streak that grows in size is worth taking seriously on its own; and when such a streak coincides with rising US yields and a weakening won, that combination is the signature of the dollar-channel loop described above.
The KTB Follow-Through and the Bank of Korea’s Bind
Korean government bonds — gukgochae, or Korea Treasury Bonds (KTBs) — do not price independently of Treasuries. Global bond investors arbitrage across markets, so when US long yields rise, KTB yields are dragged up with them even if nothing changed in the Korean economy. That raises borrowing costs for Korean firms and households regardless of what the Bank of Korea does with its policy rate.
It also constrains the central bank directly. If the Bank of Korea cuts rates while US rates stay high, the rate differential widens further, adding pressure on the won — and a weaker won imports inflation through energy and food, undermining the point of the cut. Rising US yields therefore shrink the Bank of Korea’s room to support the domestic economy. Practically, watch the direction of the 10-year KTB–Treasury spread: a Korean yield rising purely in sympathy with Treasuries, with no domestic catalyst, is imported tightening, and it weighs on rate-sensitive domestic sectors such as construction and consumer finance.
Speed vs. Level: What Actually Moves the KOSPI Open
The Korean cash market opens at 09:00 Seoul time, hours after New York closes, so the KOSPI open embeds the entire overnight Treasury move at once. Two heuristics matter far more than the absolute yield level:
- The size of the overnight change. A move of roughly 10 basis points or more in the US 10-year overnight is the kind of jump traders commonly watch as a risk-off warning for the KOSPI open — an informal rule of thumb rather than a hard empirical threshold — with semiconductors and the KOSDAQ typically in the lead. The same yield level reached gradually over weeks tends to be digested with far less drama, because portfolios adjust incrementally.
- Round-number breaks. Psychological levels — historically thresholds like 4.5% or 5% on the 10-year — matter disproportionately, not because the math changes at 5.00%, but because breaks of round numbers generate headlines, trigger systematic de-risking, and force every investment committee to revisit assumptions on the same morning.
The common mistakes are the mirror image: trading the yield level instead of the change (a market that has lived at 4.8% for a month has already priced it); ignoring the composition of the move (real-yield-driven vs. inflation-driven); and treating a single violent open as the start of a trend when the flow data has not yet confirmed persistent foreign selling.
The Domestic Buffer: Deposits and Margin Debt
Foreign selling does not translate one-for-one into index losses, because domestic investors sit on the other side. Two KOFIA statistics tell you how much cushion — and how much hidden fragility — the domestic side carries:
- Investor deposits (tuja-ja yetakgeum) — idle cash parked in brokerage accounts, the standard gauge of retail dip-buying ammunition. This stood at roughly 107.1 trillion won as of 2026-09-11. High deposits mean foreign selling can be absorbed; shrinking deposits during a drawdown mean the buffer is being spent.
- Margin lending balance (sinyong gongyeo janggo) — stock bought with borrowed money, roughly 32.3 trillion won as of 2026-09-11. This is the amplifier: if a yield-driven decline persists, leveraged positions face forced liquidation, turning orderly foreign selling into a domestic cascade. A high margin balance heading into a Treasury spike is a fragility flag, not a bullish sign of enthusiasm.
Read the two together: large deposits with modest margin debt is a resilient setup; the reverse is the configuration in which US yield shocks do the most damage to the KOSPI.
FAQ
Does a rising US 10-year yield always hurt Korean stocks?
No. When yields rise because global growth expectations are improving, Korean exporters — autos, shipbuilders, machinery — often benefit enough to offset the valuation drag, and the index can grind higher. The damaging pattern is a yield rise driven by inflation fears or term premium, especially a fast one, because it compresses multiples with no earnings offset.
Why do Korean bond yields follow US yields if the Bank of Korea sets its own policy rate?
Short-term rates follow the Bank of Korea, but long-term KTB yields are set in a global market where investors compare returns across countries. When Treasury yields rise, foreign and domestic institutions demand comparable compensation on KTBs, dragging Korean long yields up regardless of domestic policy — effectively importing US financial conditions.
Where can I check the daily foreign flow number myself?
Naver Finance publishes the daily investor-type breakdown (foreign, institutional, retail) for KOSPI and KOSDAQ, and the KRX data portal provides the official downloadable series. KOFIA’s statistics service carries the investor-deposit and margin-lending balances. All are free and updated daily; the flow figures are quoted in eok won.
Is there a single yield level that automatically breaks the KOSPI?
No fixed number works across cycles, because what matters is the change relative to what is priced in. Round-number breaks and overnight moves of 10 basis points or more are common rules of thumb for short-term stress rather than proven triggers, and a level that shocked the market on first contact is routinely absorbed once earnings expectations and positioning adjust to it.
Sources
- KRX data portal — data.krx.co.kr
- Naver Finance — finance.naver.com
- KOFIA statistics — freesis.kofia.or.kr
- Bank of Korea ECOS — ecos.bok.or.kr
Disclaimer: This is an information and analysis publication, not investment advice. See our Methodology for data sources, standards, and our corrections policy.