Skip to content

KOSPI down 16% in October: does Korea bounce? 53.3% vs 53.7%

A monthly candlestick chart of Korea's KOSPI, KRX:KOSPI, from December 1985 to October 2026 on a log scale, on a white ground. A short mark sits over each month the index fell 10% or more: orange when it rose the next month, grey when it fell, blue for October 2026, its next month still open. A dashed line marks June 2026's record close. A large orange headline reads: After a 10% fall, the KOSPI rose next 53.3%. A box asks: does Korea's market bounce after a 10% month? It shows 53.3%, then bars of rose and fell: after a 10% month 16 of 30; after the others 53.7%, a coin flip; falls before 2006, 8 of 20; falls since 2006, 8 of 10.
The KOSPI monthly, December 1985 to October 2026, log scale: a mark over each month it fell 10% or more, orange when it rose the next month and grey when it fell; the dashed line is June 2026's record close.

Korea’s KOSPI (KRX:KOSPI) almost tripled from June 2025 to May 2026, then fell 22.19% in July and 16.19% in October. October’s close, 5,730.74, sits 37.1% under June’s record of 9,114.55. Does Korea’s market bounce after a month like that? We checked every month since 1985.

What we measured

The KOSPI’s monthly closes, December 1985 to October 2026: 491 months. It closed 10% or more lower in 31 of them. For each of those months we asked two things: was the index higher a month later, and three months later, more often than after the other months?

We wrote both tests and the bar down before pairing any month. A rule needed a higher share, fewer than 5% of 10,000 random draws doing as well, and the same lean in both halves of the record.

What we found

After a 10% month, the KOSPI rose the next month 53.3% of the time (16 of 30), against 53.7% after the other months: a coin flip. Three months on it was higher 53.3% of the time (16 of 30), against 59.2%.

The typical next month was +1.12%, against +0.45% after the others. Three months on it was +2.47%, against +1.70%. The typical month leans up a little, but the shares say coin flip.

TestAfter a 10% month, higher (%)After the other months (%)Random draws as low (of 10,000)
The next month53.3 (16 of 30)53.75,527
Three months on53.3 (16 of 30)59.23,257

Could it be luck?

We drew 30 months at random, 10,000 times. For the next month, 5,527 draws had 16 rises or fewer. For three months on, 3,257 had 16 or fewer. Neither test meets the bar we set first, so there is no rule either way.

Two halves that disagree

We split the record before the run: 1985 to 2005 and 2006 to 2026. Before 2006 the KOSPI rose the month after only 8 of its 20 falls, 40.0%, against 51.6% after the other months of those years. The Asian crisis of 1997 and 1998 is among them.

Since 2006 it rose after 8 of 10, 80.0%, against 55.6%. Three months on, the halves split the same way. So the record cannot say which half October 2026 belongs to.

This year’s falls

This year holds three of the KOSPI’s seven deepest months since 1985: July is third, March fifth and October seventh. After March it rose 30.61% the next month. After July it rose 3.40% the next month and was 13.11% lower three months on. October’s next month is still open.

MonthKOSPI that month (%)
October 1997-27.25
October 2008-23.13
July 2026-22.19
May 1998-20.08
March 2026-19.08
April 1990-18.60
October 2026-16.19

On the chart

The KOSPI monthly from December 1985 to October 2026, on a log scale, where the same percentage move takes the same height. A short mark sits over each 10% month: orange when it rose the next month, grey when it fell, blue for October 2026, whose next month is still open. The dashed line is June 2026’s record close.

What to do with it

A 10% month in the KOSPI has not said much about the next one: it rose about as often as after any month. Treat a bounce as a coin flip, not a given.

How to check it yourself

  1. Take the KOSPI’s monthly closes and mark each month it closed 10% or more lower.
  2. For each, see whether it closed higher the next month and three months later, and do the same after every other month.
  3. Compare the shares, then split the record in two and see whether both halves agree.

What this does not tell you

Why the index fell (each crash had its own story), single Korean stocks, the won, or how big the next move is. Monthly closes, past data.

Observations, not recommendations.

We’re Signal Pilot Labs. We build the Elite Seven, seven indicators for TradingView, and we post what we measure, including the parts that didn’t work.

See what we build at signalpilot.io.


Past measurement described after the fact. Nothing here is a recommendation or an indication of future results.