KOSPI 200 and KOSDAQ 150: How Korea’s Index Stocks Get Picked

Every June and December, a committee at the Korea Exchange decides which Korean companies a large slice of the world’s passive money is required to own. The rules it follows are published — in Korean, as two PDFs, on a page most foreign investors never find. This is what they say, and what our own trading data shows happens when the list changes.

Ask an English-speaking investor what the KOSPI 200 is and you will usually get a version of “the Korean S&P 500.” That is close enough for a headline and wrong in the way that costs money: it suggests the index is a passive description of the market. It is not. It is a list, maintained under a written rulebook, and when the list changes, money moves because it has to.

That rulebook exists in public. It is a Korean-language PDF called KOSPI 200 지수 방법론 (“KOSPI 200 Index Methodology”), filed in KRX’s index methodology library alongside a sibling document for the KOSDAQ 150. Everything below is read from those two documents, from KRX’s own index notice board, and from our own settled price database. Where we are inferring rather than quoting, we say so.

📊 Is the KOSPI 200 the Same Thing as the KOSPI?

No. The KOSPI is the capitalisation index of every common stock on Korea’s main board; the KOSPI 200 is a separate index built from 200 of them, chosen twice a year against published criteria, and it starts from a different base date and a different base number.

This is the single most common confusion, and it is easy to see why: the two move together almost all the time. But they are different objects with different rulebooks, and a reader who treats them as interchangeable will misread everything that follows — including why Korea’s derivatives, and its trading halts, are built on one and not the other.

The same relationship holds one level down. The KOSDAQ is the whole secondary board; the KOSDAQ 150 is a selected list of 150 names inside it. If you are still working out how the two boards differ in the first place, start with our Korean stock market 101 guide to KOSPI vs KOSDAQ and come back.

Source: KOSPI 200 지수 방법론 and KOSDAQ 150 지수 방법론, sections 6 and 9. Every clause cited in this article is read from two specific editions — the KOSPI 200 methodology dated December 2020 (14 pages) and the KOSDAQ 150 methodology dated December 2023 (13 pages), both dates printed on the cover. Both documents are archived with this article (see the method note), so you can check any quotation here against the page it came from.

For scale: on June 12, 2026 the 200 KOSPI 200 constituents carried a combined market capitalisation of roughly $4.09 trillion, and the 150 KOSDAQ 150 names roughly $227 billion — an eighteen-fold gap between Korea’s two benchmark lists. (Full market cap as published by KRX for each index, converted at the June 11, 2026 USD/KRW close of 1,517.38, the last settled rate in our data before the change date; see the method note.)

🧮 How Does a Stock Get Into the KOSPI 200?

A stock must first survive an eligibility filter, then clear a market-cap threshold inside its own industry group, then beat a turnover ranking, then pass a buffer test that deliberately favours the incumbent — and even then the Index Committee can decline to admit it.

The methodology is unusually explicit about its own intent. Section 5.1 states that the document was written “to exclude arbitrariness and separate judgement as far as possible.” The rest of section 6 is what that looks like in practice.

Step 1 — eligibility. The universe is the KOSPI constituents as of the screening reference date, excluding: administrative-issue and liquidation-trading stocks; REITs, ship investment companies, infrastructure investment funds and SPACs; any stock with a free-float ratio below 10%; and anything listed for less than six months. The KOSDAQ 150 list adds investment-caution-alert issues to the exclusions.

Step 2 — the screen itself. Two numbers decide everything: average daily market capitalisation and average daily trading value, measured over the trailing six months including the month the reference date falls in.

Step 3 — three rounds. This is where the two indices diverge, and the differences are not cosmetic.

Source: KRX methodology documents, sections 6.1–6.5. Percentages are the exchange’s own, quoted as written.

Round 2 is the one worth reading twice. The incumbent buffer (110% on the KOSPI 200, 120% on the KOSDAQ 150) is wider than the newcomer hurdle (90% and 80%). That asymmetry is deliberate: it makes the index sticky. A stock that is drifting down gets held for a cycle; a stock that is climbing has to clear a higher bar to displace it. Turnover is suppressed by design, which is good for tracking costs and bad for anyone hoping to predict entries purely from a market-cap table.

And then a committee decides anyway

Section 5 of the methodology carves out an exception to its own machinery. Between the screening date and the change date, “significant corporate events, or highly unusual and complex situations this methodology does not cover” can arise; when they do, the KRX Index Committee — established in May 1994, up to seven members, mostly outside experts, on two-year terms — reviews whether the stock is a fit. The methodology then binds KRX to two things: section 5.4 requires it to publish the background and the handling in advance on the KRX index site, and section 5.5 requires the details to be preserved permanently as an electronic record.

There is also a plainer hook, one sentence long and easy to miss. Section 6.1.2 of the KOSPI 200 methodology — section 6.1.4 in the KOSDAQ 150 document — says that “any other stock deemed unsuitable as a constituent may be excluded from the eligible universe” (our translation of 그 밖에 구성종목으로 적합하지 않다고 인정되는 종목은 심사대상에서 제외할 수 있습니다). No corporate event is required. No test is specified. It sits inside section 6, the mechanical part, and it is the clause that makes the arithmetic conditional.

This is not theoretical. The May 22, 2026 notice announcing the June changes carries a footnote, under its KOSDAQ 150 item, stating that two KOSDAQ names — Taihan Fiberoptics and Fadu — will not be included in this periodic change following deliberation by the Index Committee (our translation; the notice is in Korean). The November 18, 2025 notice carries the same construction, also under the KOSDAQ 150 item, for three names: Wonik Holdings, Samhyun and — again — Fadu. Two of those three are instructive: Wonik Holdings and Samhyun were both admitted at the very next review, in the June 2026 list. Fadu was not, and has now been left out of six consecutive reviews.

It is not a KOSDAQ-only habit. The November 23, 2023 notice applied the same footnote under its KOSPI 200 item to Youngpoong Paper Mfg (006740), the stock at the centre of that autumn’s price-manipulation case. Be precise about what that notice does and does not say, because the difference matters. Its entire text on the point is one line: “Youngpoong Paper Mfg (006740) was not included in this periodic change, following deliberation by the Index Committee” (our translation). It does not say the stock had cleared the screen, and we have not found a KRX document that does; the widely repeated account that it was a qualifying candidate comes from press coverage of the case, not from the exchange's own notice board. So we are not going to lean on it.

We do not need to. The claim the notices support on their own is narrower and sturdier, and section 6.1.2 is what carries it: the published arithmetic tells you who is eligible, not who gets in. KRX reserves the right to exclude a named stock, on either board, and the methodology attaches no test to that right. Across the three notices quoted above, six such exclusions are named — one on the KOSPI 200 in November 2023, three in November 2025 and two in May 2026, all on the KOSDAQ 150, covering five distinct companies because Fadu appears twice. If you are reconstructing a future list from a market-cap table, that is the gap you cannot close, and no amount of spreadsheet work will close it.

🗓️ When Does the KOSPI 200 Rebalance, and When Is the Next One?

Twice a year. The methodology defines the change date as the trading session immediately after the final trading day of the June and December KOSPI 200 futures contracts, with constituents decided by the Index Committee in May and November.

Note what that means: the index’s maintenance calendar is set by its own derivative. The futures contract does not follow the index schedule; the index schedule follows the futures contract. And KRX’s contract specification fixes the last trading day of KOSPI 200 futures as the second Thursday of the contract month — and the specification adds, in brackets, that if that Thursday is a holiday the last trading day is brought forward — successively, to the preceding trading day, never pushed back. That bracket does not bite in December 2026, but it will eventually, and it is the first thing to check in any future cycle.

The KOSDAQ 150 methodology defines its change date the same way — by reference to the KOSPI 200 futures calendar, not to any KOSDAQ instrument. The two indices are reviewed together, announced in the same notice, and changed on the same day.

Here is the actual record, pulled from KRX’s index notice board rather than reconstructed:

Source: KRX index notice board, notices posted on the dates shown. Change dates are quoted from the notice text.

The December 2026 date is a derivation, so here is the check. The second Thursday of December 2026 is the 10th, which makes the next session Friday the 11th. Now run the same rule backwards over every row in the table. Second Thursdays: December 14, 2023 → Friday the 15th. June 13, 2024 → Friday the 14th. December 12, 2024 → Friday the 13th. June 12, 2025 → Friday the 13th. December 11, 2025 → Friday the 12th. June 11, 2026 → Friday the 12th. The rule reproduces all six published change dates exactly, which is why we are willing to put December 11 in a table. And neither December 10 nor December 11 is a 2026 KRX non-trading day, so the holiday bracket in the futures specification does not move anything this cycle.

The date that actually matters is in October

The methodology sets the screening reference date as the last trading day of the month two months before the month containing the change date, and the review period as the trailing six months including that month.

For a December 11, 2026 change, that puts the reference date on Friday, October 30, 2026 (October 31 is a Saturday, and the exchange’s only October closures this year are the 5th and the 9th), with a review period running from May through October 2026.

📈 What Actually Happens to a Stock on Rebalance Day?

In the June 2026 change, the volume showed up on the session before the date the notice named — and every one of the four stocks being added to the KOSPI 200 then lagged a market that rose 10.45% over the next two sessions.

This is the part nobody publishes in English, because it requires the constituent list and a full-market price history side by side. We have both, so we measured it.

Method. For every stock we take the median daily share volume over the 30 sessions from April 8 to May 21, 2026 — a window that ends the day before the announcement, so it cannot be contaminated by the event. Every figure below is that day’s volume divided by that stock’s own pre-announcement median. Prices and volumes are settled Korea Exchange closes held in our own coverage database (2,768–2,773 listed names per session). The control groups are the constituents that were neither added nor removed: 196 names on the KOSPI 200 and 134 on the KOSDAQ 150. Membership is taken from KRX’s own published constituent lists for June 11 and June 12, 2026 — which, usefully, also confirm the change date from a second direction: the lists are identical on June 10 and June 11, flip on June 12, and are identical again on June 15.

Median multiple of each stock’s own pre-announcement median volume, rounded to two decimals. Source: our own calculation from settled closing data; constituent lists as published by KRX for each session (archived with this article — see method note).

Three things fall out of that table, and one of them is a warning about the other two.

First, the announcement did nothing. On May 22 the eight affected KOSPI 200 names traded at 0.74× their normal volume, below the 0.87× of the wider index. Publishing the list moved nothing. The calendar did the work, not the news.

Second, the concentration is on June 11, not June 12. That is consistent with how tracking works — a fund that must hold the new list when the index is computed on the 12th has to be positioned by the close of the 11th — but we are describing sequence, not proving mechanism.

Third — and this is the warning — June 11 was also the quarterly futures-and-options expiry, which is the obvious explanation for the 196 unchanged names also running hot at 1.41×. We cannot separate the two events, and we are not going to pretend otherwise. What the control groups do let us say is the size of the excess. On the KOSPI 200, the changed names ran at 2.11× the unchanged names’ multiple on June 11, and by June 12 that had fallen to 1.24×. The KOSDAQ 150 is the sharper picture, because its own 134 untouched names only reached 1.71× on the same session: against that control, the 16 deletions ran at 4.16× and the 16 additions at 2.12×.

There is a second control worth running, and it cuts the other way from what you might expect. Widen the lens past the index: on that same June 11, the median multiple across all 794 KOSPI-listed names in our database was 0.85×, and across the whole market, 2,608 names, 0.62×. The rest of the market was quieter than normal that day. Whatever lifted volume on June 11 did not lift the market — it lifted the index constituents, which is what expiry-and-rebalance positioning would look like and is not what a general risk event would look like. That strengthens the reading; it still does not separate the expiry from the rebalance, because both act on the same 200 names. If you want the mechanics of expiry day itself, our guide to circuit breakers, sidecars and options expiry covers the witching-day pattern.

The stocks that do not fit

Group medians hide individuals, so here is every name, with no rounding up of the inconvenient ones — and, in the last column, the only comparison that means anything: what the index itself did over the same three sessions.

Volume multiple versus each stock’s own April 8 – May 21, 2026 median. Price changes are settled closes. The benchmark is the KOSPI 200 itself (1,231.54 on June 11 → 1,360.26 on June 15) because all eight of these are KOSPI 200 changes; the KOSDAQ 150 rose 3.96% over the same two sessions and is the right yardstick for that board’s names. Index levels: KRX published closes. Source: our own calculation; constituent lists as published by KRX.

Now read that last column, because the first two are a trap. The KOSPI 200 rose 10.45% in the two sessions from June 11 to June 15. Against that, all four additions lagged the index — by between 2.5 and 18.5 percentage points — and the only two names in the whole group that beat the index were deletions: GS E&C at +5.01pp and Sebang Global Battery at +0.87pp. Median relative performance was −11.86pp for the four names bought into the index and −4.63pp for the four thrown out. In a roaring two-day rally, the stocks passive money was obliged to buy were the ones that lagged it.

Two sessions is a single window, so we widened it rather than stop at the one that flatters the story. Measured from the June 11 close against the KOSPI 200 over the same span, the four additions’ median relative performance was −5.41pp to June 12, −11.86pp to June 15, −11.87pp to June 16, −20.31pp to June 30 and −8.89pp to July 31. The additions lag in all five windows — including the one ending July 31, by which point the index itself was down 15.00%, so they lagged on the way up and on the way down. What is not stable is which names beat the index: two do at June 12, two at June 15, none at all by June 30, and four by July 31, two of them additions. So the durable finding is the underperformance of the additions, not the identity of the winners.

GS E&C is doubly awkward for the tidy story: it never showed the volume signature either. At 1.41× it sat within a rounding error of the control group’s own 1.41× — 1.4135× against 1.4099×, which is a coincidence rather than a typo, and we mention it because the two figures print identically in the tables above.

One naming trap before you check any of this against a screen: “OCI” here is OCI Co., Ltd. (456040), the operating company created in the 2023 split, not OCI Holdings (010060), its listed parent. Both sit in the KOSPI 200 — the parent was already a constituent on June 11 — and only the operating company was added in June 2026.

And one name breaks the timeline entirely. GC Corp.’s heaviest session in this window was May 21 at 6.15× — the day before the announcement, followed by 4.22× on the notice day itself. Whatever was moving that stock, it was not the index calendar.

⚙️ Why Are Korea’s Futures, Options and Sidecars Built on the KOSPI 200?

Because the KOSPI 200 was built for them. The methodology’s own opening section says the index was developed in June 1994 to serve as the underlying for futures, options and ETFs — it is a derivatives instrument that also happens to be a benchmark, not the other way round.

That single sentence explains a set of facts that usually get presented as unrelated trivia:

• Why the rebalance calendar is keyed to futures expiry. Section 7.1 does not pick a convenient Friday; it defers to the futures contract’s last trading day. An index designed around its derivative reschedules itself around the derivative.
• Why a sidecar watches futures, not the index. Korea’s five-minute program-trading freeze keys off KOSPI 200 futures — the tradable contract — while the market-wide circuit breaker watches the broad composite index. Two brakes, two different reference points, and the tradable one is the derivative. That split comes from KRX’s market rules rather than from the index methodology quoted elsewhere in this article; our halt mechanics guide sets out the thresholds.
• Why 200 and not, say, 50. A derivatives underlying has to be broad enough to be a credible hedge and narrow enough to be replicable at low cost. On June 12, 2026 the 200 names covered 93.4% of total KOSPI market value — you can hedge the market with a list you can actually buy.
• Why the index has no cap. A capped index is a worse hedge for anyone whose actual exposure is uncapped. KRX’s answer is to publish capped variants as separate products rather than cap the flagship.

That last point has a consequence worth stating plainly. On June 12, 2026, Samsung Electronics and SK Hynix together were 55.05% of KOSPI 200 market value; on September 18, 2026 they were 54.58%. (KRX’s published constituent file carried 201 names on that September session — one mid-cycle addition, Hanwha Machinery & Service Holdings, had joined since June with nothing yet removed.) Both figures are on a full market-cap basis — the index itself weights by free float, which we cannot reproduce because KRX sells the float ratios as a licensed data product and adjusts them each June and December. The direction of the point survives the caveat: this is a two-stock index wearing a 200-stock coat, and any product tracking it inherits that. If you are choosing between Korean index products, that concentration question belongs next to the fee question — our comparison of Korea Value-Up ETFs works through a different index with a different answer (it caps single stocks at 15% at each rebalance), which is a useful contrast.

🔬 How Is the KOSDAQ 150 Different in Practice?

It is a much thinner slice of its market: on June 12, 2026 the KOSDAQ 150 held 60.9% of total KOSDAQ market value, against the KOSPI 200’s 93.4% of the KOSPI — and its rulebook is written to produce exactly that.

Compare the first-round thresholds in the table above. The KOSPI 200 sweeps up names until it has 85% of each industry group; the KOSDAQ 150 stops at 60%. These are not the same quantity as the market-share figures — one is a per-industry-group cumulative cut-off inside the selection process, the other is the finished index measured against its whole board — but they point the same way, and the outcome is consistent with the design.

The practical consequences for a foreign investor:

• A KOSDAQ 150 tracker is not “the KOSDAQ.” Roughly 40% of the board’s market value sits outside it — 39.2% on June 12, 2026. A KOSPI 200 tracker is a much closer proxy for its own board.
• The KOSDAQ 150 has a small-cap purge the KOSPI 200 does not. Section 6.4.5 (“소형주 제외”) lets KRX drop a candidate whose average market cap ranks below 300th among all KOSDAQ common stocks, and fill the seat with the largest remaining name that clears the liquidity test. It bites at the review, not mid-cycle — but it means a KOSDAQ 150 seat is not held by inertia alone. The KOSPI 200 methodology has no equivalent clause.
• It has a thinner safety net. Five reserve names per industry group against the KOSPI 200’s ten, which matters because reserves are what fill a seat when a constituent is suspended or delisted mid-cycle.
• Its own concentration is milder but real. On June 12, 2026 the three largest KOSDAQ 150 names by market cap were Alteogen (5.25%), Ecopro BM (4.83%) and Ecopro (4.56%) — all on a full market-cap basis, not index weights — nothing like Samsung’s 30% of the KOSPI 200 on the same basis, but a bio-and-battery tilt that a buyer should know they are taking.

One more asymmetry that catches people out: membership of these indices is not the only list KRX maintains that decides where your Korean stock can trade. A separate eligibility list governs Korea’s alternative trading venue, and it moves on its own schedule — see why a Korean stock can disappear from Nextrade.

🚧 What Is Changing in the Rulebook Right Now?

The industry-group classification that drives the entire selection process is being replaced, and KRX has said so in writing.

Every threshold in the selection table is computed inside an industry group: 85% of the group’s market cap, 110% of the group’s incumbent count, ten reserves per group. Change the group boundaries and you change who gets in, without changing a single percentage.

The index notice board carries a notice dated August 18, 2026 inviting index users to a September 1 briefing on — quoting the title — “the introduction of a KRX industry classification system and improvements to flagship index methodology,” explicitly for the purpose of gathering feedback. The current methodology, in section 6.2, classifies eligible stocks into ten (KOSPI 200) or eleven (KOSDAQ 150) industry groups “with reference to GICS” — the Global Industry Classification Standard that KRX’s own footnote attributes to S&P and MSCI. That clause is what is in play.

On September 22, 2026 KRX put the rest of it in writing. A press release titled “Korea Exchange introduces its own industry classification system (KRICS)” — notice no. 260915767, with a 28-page methodology attached — states four things plainly. (KRX publishes no deep link to an individual notice, so to pull it up yourself: open the press-release board, switch to the Korea Exchange tab, and search the title for 산업분류, which returns it as the only hit.)

• The KRX Industry Classification Methodology has been enacted, effective September 23, 2026 (KRX press release, notice no. 260915767).
• Publication of the classification results begins Monday, October 26, 2026.
• The top level is nine sectors: energy & chemicals, materials, industrials, mobility, information technology, financials & real estate, consumer, health care, and media & content.
• The classification will be applied to “flagship indices such as the KOSPI 200 and KOSDAQ 150” — in the release’s own words, “within the second half of this year.”

Note what the release does not say. It gives no date for revising the index methodology itself. The widely reported November target for that revision comes from an unnamed exchange official quoted in the Korean financial press, not from the release — so treat “nine sectors, October 26, second half of 2026” as documented, and “November” as reported. As of this writing we have found no revised index methodology, in the index methodology library or anywhere else, and the two editions we hold — dated December 2020 and December 2023 — both classify by GICS reference. Until a newer one appears, section 6.2 is the live rule. Expect the clause number to move when it does: a revision that replaces the classification will almost certainly renumber the sections around it, so check the heading, not the number.

📚 Lingo Check

🎯 Why It Matters for K-Export Stars

This site exists on a simple premise: the documents that explain Korea’s market are published, and they are published in Korean. The KOSPI 200 methodology is a clean example. It is eleven pages, freely downloadable, and it answers questions that English-language coverage answers with guesses — why the rebalance lands when it does, why the incumbent gets a wider buffer than the challenger, why a committee can override an arithmetic screen, why the flagship index has no concentration cap.

The second half of the premise is that we should check the document against the tape rather than stop at the quote. That is why the volume section exists, and why it ends with a table that argues against the obvious trade. An index change is a scheduled, disclosed, non-discretionary flow — which is exactly the kind of event that looks like an edge until you measure it against the index and find that over those two sessions the only two names to beat the market were the ones being thrown out — and that widening the window changes which names win but never rescues the ones being bought.

Method note. Volume multiples are each stock’s daily share volume divided by its own median daily volume over the 30 trading sessions from April 8 to May 21, 2026, a window that ends before the May 22 announcement. Group figures are medians of those per-stock multiples, not ratios of summed volume, and are rounded to two decimals. Constituent membership comes from KRX’s own published constituent files for June 10, 11, 12 and 15, 2026 and for September 18, 2026; we diff consecutive sessions to identify additions and removals, which returns exactly the four-for-four KOSPI 200 change and the sixteen-for-sixteen KOSDAQ 150 change that the May 22 notice announced. Those files are archived alongside this article so the tables can be rebuilt — and so the December 2026 review can be measured the same way. The two methodology documents every rule in this article is quoted from are archived with them: KOSPI 200 지수 방법론, cover date December 2020, 14 pages, and KOSDAQ 150 지수 방법론, cover date December 2023, 13 pages. Section numbers cited here (5.1, 5.4, 5.5, 6.1.2, 6.2, 6.4.5, 7.1 and the rest) are the numbering in those editions; KRX does not version its methodology library in a way that lets us assert these are the newest, so the dates are the anchor. Index market-share and concentration figures use KRX’s published market-capitalisation totals for each index and each board on the stated date; they are full market cap, not free-float adjusted, so they are not index weights. Relative-performance figures compare a stock’s percentage change with the KOSPI 200’s own percentage change over the identical two sessions and are stated in percentage points. Price changes and index levels are settled closes. Dollar figures are converted once, at the June 11, 2026 USD/KRW close of 1,517.38.

Disclosure. The author holds a Korean ETF that tracks the KOSPI 200, and a position in d’Alba Global, one of the June 2026 additions discussed above. Neither is a recommendation, and this article contains no view on either.

Disclaimer: This article is for informational and educational purposes only and is not investment advice, an offer, or a solicitation to buy or sell any security. Index rules and schedules can change; figures are dated where given and were accurate as of September 22, 2026. Always verify current index methodology with the Korea Exchange and consult a licensed professional before investing.