EWY Holdings: Two Stocks Are 43%, Not 50%

Not financial advice. TheGatBull explains Korean market structure for global investors; it does not recommend trades.

EWY — the iShares MSCI South Korea ETF — has 42.85% of the fund in its top two holdings. Not the 50% you have probably seen.

Open the holdings file iShares publishes every business day — as of August 13, 2026 — and you get Samsung Electronics at 22.82% and SK Hynix at 20.03%. Together, 42.85%. Several screeners and aggregator sites showed the same two names, as displayed on August 9, as SK Hynix 27.17% + Samsung 23.19% = 50.36%.

That is a 7-point spread. And one of those two numbers cannot exist — the arithmetic is below.

But the conclusion here is not “it is less concentrated than you feared, relax.” It is the opposite. 42.85% has now passed the combined weight of the S&P 500’s entire top ten (about 40%, press-reported). EWY gets there with two stocks.

What this means for EWY holders.
1. About $43 of every $100 in your EWY position sits in two memory-chip companies. That is not “Korea diversification” — it is exposure to the memory cycle.
2. If you had that number as $50, you were overstating it. The issuer’s own file says 42.85%, and that 7-point gap was worth roughly 1.19 percentage points in won terms in a single week. The math is below.

The worry Americans already have

Concentration — how much of an index sits in a handful of names — is not a new argument in the US.

The Magnificent Seven (Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, Tesla) account for roughly 33.8% of the S&P 500 as of June 2026 — press-reported; we have not matched this against S&P Dow Jones Indices’ own data. Widen to the top ten and it is reported at about 40%. Historically the top ten averaged 24%, with the last pre-2023 peak at 28% in 1970, per the same reporting.

American readers do not need those numbers explained. “I bought an index fund and it turned out to be a big-tech bet” has been standard retail-investor phrasing since 2024.

Take that worry across a border

In a single-country ETF, the same problem goes one step further.

The S&P 500’s denominator is 500 names. EWY’s is about 80. And a national stock market inherits that nation’s industrial structure. The US market spreads its top names across software, platforms, semiconductors and e-commerce. Korea’s top names sit in one industry.

We mapped that structure in an earlier piece — open EWY and you find two memory stocks, not “Korea.” Today’s question is the next one: exactly what percentage, and was the number your search turned up correct?

What the issuer’s own file says

iShares holdings, as of August 13, 2026 — every weight below uses that date.

#	Holding	Weight	Sector
1	Samsung Electronics (005930)	22.82%	Information Technology
2	SK Hynix (000660)	20.03%	Information Technology
3	SK Square (402340)	3.44%	Industrials
4	Samsung Electro-Mechanics (009150)	3.06%	Information Technology
5	KB Financial (105560)	2.24%	Financials

Top two: 42.85%. The gap to third place (3.44%) is about 17 points. That is not “the top is heavy.” That is “the top two and everything else are in different leagues.”

By sector the same fact reads differently — IT 48.59%, industrials 20.33%, financials 11.01%, utilities 0.32%. Which means a regulated utility in this fund is 0.32% — rounding error territory. It is why the company that fell 6.63% yesterday on its Q2 results is nearly invisible to an EWY holder.

Under the Gat. People call EWY “Korea exposure.” More precisely it is “43% memory exposure plus 57% of the rest of Korea.” Calling two things by one word is what causes the confusion. The fund has never hidden it — it posts the file every business day.

Why 50.36% cannot exist

This is the core of the piece.

We used the screener figure at first too (Samsung 23.19% + SK Hynix 27.17% = 50.36%). Then we checked it against the sector data on the same screen, and the sentence collapsed.

The two stocks together came out larger than the entire sector they belong to.

Figure	Value	Verdict
Screener: Samsung + SK Hynix	50.36%	⛔ Impossible — exceeds the sector it belongs to
iShares: entire IT sector	48.59%	✅ Primary source, issuer-published
iShares: Samsung + SK Hynix	42.85%	✅ Primary source, issuer-published

Both companies are classified in Information Technology. A part cannot be larger than the whole. 50.36% is arithmetically impossible. (The same verdict held a week earlier against the August 6 file, when the IT sector total was 46.71%.)

Where exactly the error creeps in — FX timing, as-of date lag, treatment of cash and derivative positions — we could not determine. Nor do we need to. The issuer publishes the original for free, every business day.

Under the Gat. The problem is not that the number is wrong. The problem is which direction it is wrong in. Fifty percent sounds scarier than forty-three, and a scary number pushes people out of a fund. Forty-three is scary enough — but fear is only useful when it is accurate.

Why these two should not be treated as one block is something we took apart separately in Samsung and SK Hynix are not the same trade — inside the fund they add to 43%, but their P&L structures diverge.

And the original does not sit still

It moved while we were writing this. The top two were 41.38% as of August 6. They are 42.85% as of August 13 — 1.47 points in one week.

Which is exactly what should happen: the two names rose 18.83% and 15.68% over that stretch, and in a market-cap-weighted index, a stock that rises gains weight automatically. Concentration is not a fixed figure. It is a variable that updates daily — and right now it is updating upward. That is why quoting one screenshot for months is a bad habit.

What is that 7-point gap worth in a week?

Left abstract, it does not land. Convert it into this week’s numbers.

Friday August 7 close to Friday August 14 close, KRX regular-session 15:30 closes (not alternative-venue aggregates):

Instrument	Week
Samsung Electronics (₩231,000 → ₩274,500, about $194)	+18.83%
SK Hynix (₩1,422,000 → ₩1,645,000, about $1,160)	+15.68%
KOSPI (6,258.77 → 6,977.94)	+11.49%
KOSDAQ (798.81 → 864.65) — Korea’s smaller-company index, its Nasdaq analogue	+8.24%

Won-dollar conversions use ₩1,418.3 per dollar, the Seoul 15:30 close on August 14. The won firmed 0.39% over the same week.

Now we apply each of the two weightings ourselves — this is our own calculation, not a figure either source publishes — to get what the pair contributed, in won terms.

Weighting used	Combined contribution
iShares primary (42.85%, Aug 13)	+7.44 pp
Discarded screener figure (50.36%)	+8.63 pp
Difference	1.19 pp

“pp” is percentage points. In plain terms: if you had modeled EWY’s week using the screener’s 50.36%, you would have credited these two stocks with about 1.19 percentage points more of the fund’s move than they actually carried. On a $10,000 position, that is roughly $119 of performance attributed to the wrong place, in five trading days.

1.19 percentage points in one week. That is what the 7-point error actually weighs.

Three limits, stated plainly: the weighting date (Aug 13) and the return window (Aug 7–14) do not fully overlap, so this is an approximation under fixed weights rather than a realized attribution; the math is in won and carries no FX effect; and we could not source what the remaining 57.15% contributed, which means “the two stocks made X% of EWY’s gain” is a ratio claim this piece cannot support — only the absolute contribution above.

One more thing. Shift the weekly window to Monday, August 10, and the picture gets far more dramatic: KOSPI +10.77% against KOSDAQ +1.19%, a gap of 9.58 points. But that gap exists only because the shifted window pushes KOSDAQ’s +6.97% on August 10 outside it — the day a buy-side sidecar fired, the mechanism that suspends program-trading orders for five minutes when futures spike. So we do not use it. Choosing a start date to make a gap look bigger is not analysis. Last week’s divergence is covered in the weekly scoreboard.

Two ways to read the same number

The same 42.85% reads in opposite directions. Neither reading is wrong.

	Reading A — “that is good”	Reading B — “that is dangerous”
Premise	Korea’s real center of gravity is semiconductors. An index should reflect reality	Two stocks in the same industry, on the same cycle. Almost no diversification benefit
This week	The pair delivered +7.44 pp in won terms. You would not have received that without holding them	The same size arrives in reverse when the cycle turns. The last week of July did exactly that
Implication	EWY is an honest product. Korea simply looks like this	EWY is closer to a sector ETF wearing a country ETF’s label

Hold the US anchor up again. Not a perfect parallel, but the direction holds.

• S&P 500 top ten ≈ 40% (press-reported) — 10 out of 500
• EWY top two = 42.85% (issuer-published, Aug 13) — 2 out of about 80

Three caveats. First, the denominators differ by more than six times. Second, EWY tracks MSCI Korea 25/50, whose rules cap the combined weight of positions above 5% at 50% — meaning this concentration is already inside a rule-managed band (methodology document not read directly). Third, US concentration is spread across seven to ten companies in software, chips, retail and EVs. EWY’s is two companies, one industry, one inventory cycle. The correlation is different.

What this says about the Korean market

That was the fund. This is the market.

EWY’s 43% is not something iShares built. It is the output of replicating MSCI Korea 25/50, and that index simply transcribes the market-cap distribution of Korean equities. So 43% is not product design. It is the output of a country’s industrial structure.

Which means this number should be read as a Korean economic indicator, not fund research. The government can expand dividends and buybacks through its Value-Up program, regulators can tighten leveraged products, and the index’s center of gravity will not move. It moves when those two companies’ memory prices move.

This week proved it. Samsung +18.83%, SK Hynix +15.68%, KOSPI +11.49%. KOSDAQ +8.24%. What lifted the index was not policy or liquidity. It was two companies.

Under the Gat. Buying a Korean index is less like buying the Korean economy and more like buying leverage on the memory cycle. And the ETF did not create that fact. The Korean stock market is shaped that way, and MSCI merely wrote it down. If there is something to be annoyed at, it is the industrial concentration, not the product.

How to approach it — the exposure ladder

If your conclusion is “43% is too much,” there are three options. None of these is a recommendation. We are describing structure.

1. One click today — buy EWY (NYSE Arca) as it is, accepting 42.85% in the top two as of August 13. Expense ratio 0.59%.
2. Separate the two names — SK Hynix has a US-listed ADR (Nasdaq: SKHY). An ADR is a US-listed certificate representing shares of a foreign company; it trades in dollars in an ordinary brokerage account, exactly like a domestic stock. Samsung Electronics has no US-listed ADR — only over-the-counter routes, which many US brokerages restrict or price poorly, so treat it as effectively unavailable for most retail accounts. Adjusting the two independently is therefore only half possible. That asymmetry is itself a feature of accessing Korea.
3. If you want only the other 57% — there is effectively no US-listed product that gives Korean exposure while excluding EWY’s top two. That is not a dead end so much as information: demand for a “diversified Korea” product has not been filled.

What to put on a watchlist: the iShares holdings file updates every business day. It moved from 41.38% to 42.85% in the past week alone. If the top two clear 45%, the distance to the MSCI 25/50 cap narrows too. Reading the original is the practical conclusion of this piece.

The short version

• EWY’s top two holdings are 42.85% (iShares, as of Aug 13, 2026). The widely quoted 50.36% exceeds the issuer’s own IT sector total of 48.59% and therefore cannot exist.
• 42.85% has passed the S&P 500’s entire top ten (about 40%, press-reported). Two stocks now outweigh ten.
• From August 7 to 14, the pair contributed roughly +7.44 pp in won terms on EWY’s weighting. Using the wrong figure would have shown +8.63 pp — a 1.19-point error in one week.
• Concentration is not fixed. It rose from 41.38% to 42.85% within the same week. Stocks that rise gain weight on their own.
• This structure is not a product flaw. It is the output of Korea’s industrial structure. Memory prices move the index; policy does not.
• The issuer’s original is free and updated every business day. The reprocessed version is on the first page of search results. Reversing that order is the point of this piece.

FAQ

What is EWY’s exact top-two holdings weight?

As of August 13, 2026, iShares’ official holdings file shows Samsung Electronics 22.82% plus SK Hynix 20.03% = 42.85%. Holdings update every business day — a week earlier, on August 6, the same pair was 41.38% — so the current figure is best checked on the issuer’s page directly.

Why do some sites show about 50%?

We could not determine the cause. But that figure (50.36%) exceeds the same issuer’s published IT sector total of 48.59%, which makes it arithmetically impossible — a part cannot exceed the whole it belongs to.

Is 43% more concentrated than US indexes?

The comparison needs matching terms. The S&P 500’s top ten is reported at about 40% (press-reported, not matched against primary index data). EWY’s top two is 42.85%. The comparison only holds when the number of holdings is stated alongside it.

Can I own EWY and reduce the semiconductor weight?

SK Hynix can be adjusted separately through its US-listed ADR (Nasdaq: SKHY), but Samsung Electronics has no US-listed ADR. So there is no standard route for adjusting both symmetrically.

Did the KOSPI pass 7,000 this week?

Only intraday. On August 14 it opened at 6,995.67 and reached an intraday high of 7,010.86, but closed at 6,977.94. When citing a period high, state whether it is a closing or an intraday figure.

Where do these numbers come from?

Holdings, sector weights and net assets are iShares’ own disclosure (primary, August 13, 2026). August 14 KRX closes were pulled directly from a price feed and cross-checked by back-calculation against the prior session’s confirmed closes. The Seoul FX close is press-reported. S&P 500 concentration figures are press-reported and not matched against the index provider’s primary data.

Sources

• iShares MSCI South Korea ETF (EWY) — fund holdings, as of Aug 13, 2026 — issuer-published, primary
• Korea Exchange — KRX regular-session closes, the basis for every Seoul figure here
• Yahoo Finance chart API — KOSPI (^KS11), KOSDAQ (^KQ11), Samsung Electronics (005930.KS), SK Hynix (000660.KS); retrieved Aug 14, 2026
• MSCI — Korea 25/50 Index methodology — benchmark rules, not read directly
• Etoday (Aug 14, 2026) — Seoul FX close, 15:30
• Forbes — S&P 500 weight in the Magnificent Seven passes 30% (June 2026) — press-reported

This article is for information only and is not investment advice. All investment decisions and their consequences rest with the reader. Figures are Seoul closes (KRX regular session, 15:30 KST) unless labeled otherwise; fund holdings are as of the date stated.

---