This article is for informational purposes only and is not financial advice. TheGatBull may earn a commission from some links at no cost to you — see our disclosure and full disclaimer.
This is not financial advice. This is structure, not a call — and nothing here declares a top or a bottom.
Part 2 of 2. Part 1 (July 13) asked why Korea fell 9% in a single day, and found the answer in index concentration and a leveraged position coming apart — not in AI demand. This piece asks the question that comes next: is the unwind over?
On July 9 and 10, Korean media reported that the KOSPI had “exited bear market territory,” and traders started asking whether 7,200 had been the bottom. The very next session — Monday, July 13 — the index fell 8.95% to 6,806.93, lost the 7,000 line for the first time in two months, and tripped the seventh circuit breaker of 2026. So the short answer is no: 7,200 was not the bottom. The longer answer is the one that matters, and it is the thing foreign investors almost never see on a Korean chart: in this market the bottom is not a price level; it is a process — ₩38.6 trillion (~$25.7B, the June 24 record) of bittu (빚투, “investing with borrowed money”) working its way out of retail accounts. On the very day the headlines asked “was that the bottom,” the number climbing fastest was the one nobody put in a headline: the forced-liquidation ratio.
The claim: “Korea exited its bear market”
The KOSPI peaked at a record close of 8,801.49 on June 2. Through early July it sold off hard — the July 8 close of 7,246.79 was 17.7% below that high, and intraday prints went lower still. Then came two green sessions — July 10 closed +2.52% at 7,475.94 — and the framing flipped: chip exports were at record highs, foreigners had turned net buyers after 13 straight sessions of selling, and the index was back well clear of the -20% mark (7,041 on the June 2 close). Bottoms, the argument went, look exactly like this.
Worth pausing on that phrase, “exited the bear market.” A -20% drawdown from 8,801.49 means a close below 7,041 — and on a closing basis the index never got there in early July (its worst close was 7,246.79, or -17.7%). The bear market being “exited” was an intraday, loose usage. Which tells you something before we even get to the ledger: the level was never the structure. It was a round number the story hung on.
The data: three days later, a circuit breaker
| Date | KOSPI | What happened |
|---|---|---|
| Jun 2 | 8,801.49 | Record close |
| Jul 8 | 7,246.79 (-5.35%) | -17.7% from the record |
| Jul 9–10 | 7,475.94 (+2.52%) | “Exits bear market” · “Was 7,200 the bottom?” |
| Jul 13 | 6,806.93 (-8.95%, -669.01 pts) | Circuit breaker No. 7 of 2026 · intraday low 6,783.43 |
| Jul 14 | 6,856.83 (+0.73%) | Intraday low 6,448.86 (-26.7% from peak) before institutions bought it back · KOSDAQ sell sidecar triggered |
July 13 erased ₩546.1 trillion (~$363B) of KOSPI market capitalization in a single session. Samsung Electronics (KRX: 005930) fell 10.70%; SK hynix (KRX: 000660) fell 15.37%. Foreigners sold a net ₩1.70T (~$1.13B) and institutions ₩2.22T (~$1.48B), while retail investors absorbed ₩3.88T (~$2.58B). At 6,806.93 the index sat 22.7% below its record — back underneath the very line the “exit” headlines had just celebrated. (Part 1 takes that session apart tick by tick: the won moved 0.13%, refiners rose double digits, and two stocks did most of the damage.)
Where the bulls are right: the fundamentals did not break
Give the bottom-callers their due — the data is on their side here.
- Exports are setting records. Korea’s exports in the first ten days of July came to $29.8 billion, up 53.9% year on year. Semiconductors alone were $11.2 billion, up 193% — an all-time high, and 37.6% of everything the country shipped. The trade surplus was $6.4 billion.
- Korean brokerage commentary does not frame this as an earnings event. Local sell-side notes cited in the Korean press read it as profit-taking after a violent run, plus a supply-demand shock from leverage liquidation — not deteriorating fundamentals.
- The bid came back. On July 14, institutions bought a net ₩3.22T (~$2.1B) and foreigners ₩0.95T (~$0.6B), pulling the index off 6,448.86 to close green.
So the story “chip demand died” is not supported by the export data. What broke was not earnings. What broke was positioning — which is exactly where Part 1 left the argument. The open question is whether the positioning has finished breaking.
🎩 Under the Gat — A US reader thinks of a bottom as a price. In Korea a bottom is a procedure. Picture a brokerage building in Yeouido: on one floor the research team publishes a target price for where the index “should” stop falling, and a few floors below, the margin desk is force-selling the accounts that broke their collateral ratio overnight. The people calling the bottom and the people making it are not the same people. Watch the second group.
Where the claim breaks: on the day everyone asked about the bottom, forced selling hit a record
Here is the reversal. The bear-market-exit story looked at the index. It did not look at the ledger underneath it.
Two definitions first, because the whole argument lives in them. A Korean margin loan (신용거래융자) has to stay backed by collateral worth roughly 140% of the loan — the collateral ratio. Break it, and the broker sells the position for you. Separately, brokers publish a forced-liquidation ratio: how much of the market’s unsettled buy credit (미수금 — the cash you owe at T+2 for stock you already bought) gets force-sold rather than paid. Two different ledgers, but they move together, because both are retail leverage and both get liquidated the same way. The ratio is the visible fever chart; the ₩38.6 trillion is the body underneath it.
| The claim (Jul 9–10 headlines) | The data | |
|---|---|---|
| Bear market | “Exited” (7,475.94) | Jul 13: 6,806.93 = -22.7% from the peak, right back in |
| Bottom signal | Two up sessions | Forced liquidation as a share of unsettled buy credit: 1% in January → 5.1% in June → 10.2% on July 9 — the highest reading of 2026, more than double June’s |
| Leverage | Not mentioned | Margin loan balance ₩38.63T (~$25.7B) — a record, as of June 24 (the July balance is the number to watch). KOSPI-only margin: ₩19.9T (~$13.2B) in January → ₩29.2T (~$19.4B) in June |
| Dry powder | “Retail has cash to deploy” | Customer deposits at brokerages: ₩139.7T (~$92.9B, Jun 4) → ₩107.1T (~$71.3B, Jul 9) — ₩32.6T (~$21.7B), 23.3%, gone in five weeks |
| Volatility | “Normalizing” | VKOSPI 96.94 on June 29, per KRX data cited by Financial News — Korea’s version of the VIX, its highest reading since the index launched in 2009 |
| Circuit breakers | “A freak event” | 7 of the 13 in KRX history happened in 2026 |
And then July 14 delivered the punchline: the intraday low was 6,448.86, or 26.7% below the record, and the retail investors who had absorbed ₩3.88T of stock the day before sold ₩4.14T (~$2.8B) into the bounce. That is what deleveraging looks like from the inside. Some of that selling is voluntary; some of it is not the investor’s decision at all — when the collateral ratio breaks, the broker sells.
Bandaemaemae is not a margin call
Koreans have a word for the mechanism: bandaemaemae (반대매매) — literally “opposite trade,” in practice forced liquidation. In the US, a margin call usually opens a window: brokers typically give you a few days to wire cash, even though the agreement you signed lets them liquidate without calling you at all. In Korea, when the collateral ratio on a margin loan breaks, the brokerage sells your shares at the next session’s open — priced down near the daily limit, regardless of your wishes. On July 9 that forced selling reached ₩142.2 billion (~$95M) in a single day, and the forced-liquidation ratio hit 10.2% — its highest reading of 2026, from 1% in January.
Then the loop closes: forced selling pushes prices down, lower prices break the collateral ratio in the next account, and that account is sold at tomorrow’s open.
Two structural facts make the swings larger than the news alone would justify. First, index concentration: Samsung Electronics and SK hynix effectively are the KOSPI, which is why one bad chip day becomes an index event — we took that concentration apart in what’s actually inside EWY. Second, the catalyst that had been holding chip sentiment up — SK hynix’s Nasdaq ADR listing (the filing, line by line) — was spent the moment the ADR actually began trading, right as the peak-chip debate reignited (the earlier round of that fight).
The verdict
“7,200 was the bottom” — rejected. Three days later: -8.95%, a circuit breaker, and -22.7% from the record.
“The fundamentals broke” — also rejected. Chip exports just printed an all-time high, +193% year on year.
What’s left: this was a leverage event, not an earnings event — the same conclusion Part 1 reached from the tape, now confirmed in the ledger. Which means the bottom will be set by the progress of the deleveraging, not by a number on the index. So far, all the data confirms is that the liquidation has started.
Four checkpoints to watch: (1) the margin loan balance — does ₩38.6T (~$25.7B) actually shrink from its June 24 record? (2) the forced-liquidation ratio — does 10.2% roll over? (3) leveraged-ETF assets — Korea’s domestic 2x SK hynix funds fell about 32% on July 13, and New York started listing 2x and inverse ETFs on the ADR the same week (Part 1 has the tickers). If money flows straight back into them, the amplifier has been reloaded rather than discharged. (4) SK hynix’s Q2 report on July 29 (how a Korean earnings release actually works — and why this company publishes no preliminary number), where consensus operating profit is near ₩65T (~$43B) while Korea Investment & Securities models ₩60.4T (~$40B), about 7% below. That number, not the narrative, settles the peak-chip argument.
Same company, different wiring
This is not a foreign story for US investors. The stock that fell 15.37% in Seoul on July 13 is the same company whose ADR debuted on July 10 — under the when-issued ticker SKHYV, switching to the permanent Nasdaq: SKHY on July 14 — and Seoul Economic Daily reported that ADR trading at roughly a 16% premium to the Seoul shares at one point (a ratio-adjusted comparison: each ADR represents a fraction of a Seoul-listed share). Part of that gap is mechanical — New York trades after Seoul closes, so the ADR had not yet repriced Monday’s 15.37% drop. The rest is demand. Which is which only becomes clear once both markets have traded through the same news. Same asset, two markets, two very different sets of plumbing.
| US (NYSE / Nasdaq) | Korea (KRX) | |
|---|---|---|
| Market-wide halt | 7% / 13% / 20% on the S&P 500 | Three stages: 8% and 15% halt trading for 20 minutes each; 20% ends the session |
| Single stock | LULD, 5-minute pause | VI (“volatility interruption”) — the same idea as LULD: one stock moves too far, too fast, and it stops trading continuously for 2 minutes |
| Futures dislocation | — | Sidecar: if index futures move ±5% for one minute (±6% on the KOSDAQ), the automated basket orders that arbitrage futures against stocks are switched off for five minutes — the KOSDAQ’s fired on July 14 |
| Retail leverage | Margin call — broker discretion, usually a few days | Margin loan + bandaemaemae — mechanical, next session’s open, near the daily limit |
| Triggered in 2026 | — | 7 circuit breakers |
Not a perfect parallel — US brokers can also liquidate without notice, and Korean investors can post additional collateral before the deadline. The difference is default behavior, not legal power.
🎩 Under the Gat — I don’t call this panic. Panic is an emotion; bandaemaemae is a circuit. Declaring a bottom before the circuit finishes running is like opening the washing machine mid-spin — the drum doesn’t care that you’re ready. Watch the margin balance, not the round number.
FAQ
Was 7,200 the bottom for the KOSPI?
No. The index closed at 6,806.93 on July 13 and traded as low as 6,448.86 on July 14. That does not make the current level a bottom either — the indicator to track is the margin balance, not the index level.
What does retail leverage have to do with the halts?
Margin debt hit a record ₩38.6T (~$25.7B) on June 24, and Korean brokers liquidate it mechanically — so a fall becomes forced selling, which becomes a bigger fall. That amplifier is why the index gaps rather than drifts. (For the halt mechanics and the count, see Part 1.)
Did Korean chip fundamentals actually deteriorate?
Not in the export data — chip exports hit $11.2B in ten days, +193% YoY, a record. Local sell-side commentary reads the drop as profit-taking plus leverage liquidation. The caveat: SK hynix Q2 estimates have started coming down (₩60.4T / ~$40B at Korea Investment & Securities vs. ~₩65T / ~$43B consensus).
What is bandaemaemae?
Forced liquidation of margin positions by the broker at the next session’s open, priced near the daily limit, without the investor’s consent. Faster and more mechanical than a US margin call. It hit ₩142.2B (~$95M) in one day on July 9.
What should I watch next?
July 29 — SK hynix’s Q2 report. And weekly: the margin loan balance and the forced-liquidation ratio.
Sources
- Seoul Economic Daily (July 13, 2026) — KOSPI 6,806.93 (-8.95%, -669.01 pts), first close below 7,000 in two months; ₩546.119T market cap erased; Samsung -10.70%, SK hynix -15.37%; foreign -₩1.7047T, institutional -₩2.2193T, retail +₩3.8829T; seventh circuit breaker of 2026
- Seoul Economic Daily (July 14, 2026) — chip-peak debate and “leverage shock”; SK hynix ADR catalyst fades; Korea Investment & Securities Q2 operating profit estimate ₩60.4T vs. ~₩65T consensus; SKHY ADR reported at a ~16% premium to Seoul
- Asia Business Daily (July 14, 2026) — July 14 rebound: close 6,856.83 (+0.73%), intraday low 6,448.86, institutions +₩3.2159T and foreigners +₩0.9523T vs. retail -₩4.1424T; KOSDAQ sell sidecar
- Korea Customs Service via Seoul Economic Daily — July 1–10 exports $29.8B (+53.9%); semiconductors $11.2B (+193%, record), 37.6% of exports; trade surplus $6.4B
- Etoday (July 13, 2026) — Korea Financial Investment Association data: margin loan balance ₩38.6328T (June 24, record); KOSPI margin ₩19.8549T (Jan) → ₩29.2346T (Jun); forced-liquidation ratio 1% (Jan) → 5.1% (Jun) → 10.2% (July 9); customer deposits ₩139.6948T (Jun 4) → ₩107.1279T (Jul 9)
- Newdaily Economy (July 13, 2026) — forced liquidation ₩142.2B on July 9
- Financial News (July 14, 2026) — 7 of the 13 circuit breakers in KRX history occurred in 2026; VKOSPI 96.94 (June 29), highest since the index launched in 2009
- Asia Business Daily (“Was 7,200 the Bottom?”, July 9) · Korea JoongAng Daily (July 10, +2.52%)
Disclaimer: This article is for information and education only and is not financial advice. It does not declare a market bottom or top. Index levels, margin balances, and earnings estimates change quickly — verify current figures before acting. ₩/$ conversions use USD/KRW 1,503.4 (the July 13, 2026 onshore close), the same basis as Part 1. The author may hold positions in securities mentioned. Do your own research.
Frequently Asked Questions
Was 7,200 the bottom for the KOSPI?
No. On July 13, 2026 the KOSPI fell 8.95% to close at 6,806.93, giving up the 7,000 line, and on July 14 it traded as low as 6,448.86 intraday before closing at 6,856.83. That does not make the current level a bottom either — the indicator to track in Korea is the margin balance, not the index level.
How does Korean retail margin debt amplify a market fall?
Margin loans hit a record ₩38.6 trillion (~$25.7B) on June 24, 2026. When an account’s collateral ratio breaks, the Korean broker sells the position at the next session’s open, priced near the daily limit — no negotiation. That forced selling pushes prices lower, which breaks the next account’s collateral ratio, which is liquidated the following morning. The fall becomes mechanical rather than emotional.
Did Korean chip fundamentals actually deteriorate?
Not in the export data. Chip exports in the first ten days of July hit $11.2 billion, up 193% year on year — an all-time high. Korean sell-side commentary cited in the local press attributes the sell-off to profit-taking and leverage liquidation rather than earnings damage, though Korea Investment & Securities cut its SK hynix Q2 operating profit estimate to ₩60.4 trillion (~$40B) versus a consensus near ₩65 trillion (~$43B).
What is bandaemaemae (반대매매)?
Forced liquidation. If a Korean retail investor’s margin collateral ratio breaks, the broker sells the shares at the next session’s open, priced down near the daily limit, regardless of the investor’s wishes and with no negotiation. It is faster and more mechanical than a US margin call, and it generates additional sell orders in a falling market. Forced-liquidation volume hit ₩142.2 billion (~$95M) on July 9.
When does SK hynix report Q2 2026 earnings?
July 29, 2026. That is the day the peak-chip debate gets settled with a number instead of a narrative: consensus operating profit sits near ₩65 trillion (~$43B), while Korea Investment & Securities models ₩60.4 trillion (~$40B), about 7% below.
This article is for informational purposes only and is not financial advice. TheGatBull may earn a commission from some links at no cost to you — see our disclosure and full disclaimer.