Samsung Electronics (005930): Which Division Earns the Money, and Which Ticker You’re Buying (H1 2026)

Every screener agrees that Samsung Electronics is one of the largest companies on earth. Almost none of them agree on how large, what it earns, or even how many Samsung Electronics shares exist. Those are not opinions — they are four different arithmetic choices, and the Korean half-year filing settles all four. Here is what it says.

📌 Prices, market values and flow figures are Korea Exchange settled data for September 4, 2026. Financial and ownership figures are from Samsung’s half-year report for the six months to June 30, 2026, filed August 14, 2026. FX throughout: ₩1,345.99/$1 (Sep 4, 2026).

🏢 What Is Samsung Electronics, and What Does It Actually Sell?

Samsung Electronics is a Korean manufacturer that runs two very different businesses under one ticker: a semiconductor business that makes DRAM, NAND flash and logic chips, and a consumer-devices business that makes Galaxy phones, TVs and home appliances. It has been the largest company on the Korea Exchange for decades, and in the first half of 2026 it booked $226.9bn of revenue.

The company was incorporated on January 13, 1969 as Samsung Electronics Industry Co., listed on what is now the KOSPI on June 11, 1975, and took its present name on February 28, 1984. Its headquarters is at 129 Samsung-ro, Yeongtong-gu, Suwon, about 30km south of Seoul. (Source: DART half-year report, rcpNo 20260814003699, section I.1) A neutral company profile is on Wikipedia.

Internally it reports as four units, and you need the Korean names to read any filing:

• DS (Device Solutions) — semiconductors: DRAM, NAND flash, mobile application processors, image sensors, and the foundry that makes chips for other companies.
• DX (Device eXperience) — everything you can buy in a shop: Galaxy smartphones, TVs, monitors, refrigerators, washing machines, air conditioners, PCs, network equipment.
• SDC — Samsung Display, a majority-owned subsidiary making the OLED panels that go into smartphones (including some that are not Samsung’s).
• Harman — the Connecticut-based automotive-electronics and audio business Samsung bought in 2017.

Around the parent sits the group. As of June 30, 2026 the Samsung chaebol comprised 68 domestic affiliates, of which 18 are listed. Samsung Electronics is one of those 18 — it is not the holding company, and it does not own the others. That distinction matters enough that we give it its own section below. If the Korean market structure itself is new to you, start with our KOSPI vs KOSDAQ primer.

📊 Where Does Samsung’s Revenue Actually Come From?

Two-thirds of it now comes from memory chips. In the first half of 2026 the “memory” product line alone booked $145.3bn — 64.1% of consolidated revenue. Smartphones, the business most Western readers associate with the name, were 22.9%.

Source: DART half-year report rcpNo 20260814003699, section II.4 (매출 및 수주상황), filed Aug 14, 2026. Product-line figures include intersegment sales, so they do not sum to consolidated revenue of $226.9bn. Filed in won; converted at ₩1,345.99/$1 (Sep 4, 2026), the single rate used throughout this article.

The shift is recent and violent. Memory was $62.7bn of revenue in FY2024 (28.1% of the total) and $77.3bn across the whole of FY2025. In the first half of 2026 alone it was $145.3bn. The filing gives the reason in one line: memory average selling prices rose about 220% against the prior-year average. Smartphone ASPs rose 7%; TV ASPs fell 4%.

By region — this table is on a parent-company (separate) basis, which is how Korean filings present it — China is the largest destination, and has been in every period the filing shows:

Source: same filing, section II.4(3), which prints FY2024, FY2025 and H1 2026. Separate-company basis, not consolidated — the totals are smaller than group revenue by construction. China ranked first in all three periods (FY2024 $48.2bn, FY2025 $53.2bn, H1 2026 $65.8bn), ahead of the Americas each time ($45.6bn, $50.4bn, $52.5bn); the regional ranking has not changed, only the scale. Six months of 2026 already exceed twelve months of 2025 in China, in Asia & Africa and in the Americas. Converted at ₩1,345.99/$1.

🗓️ How Did Samsung Get Here?

Samsung spent its first two decades as a domestic appliance maker, bet the company on memory chips in the 1980s, and has been a cyclical semiconductor business wearing a consumer-electronics coat ever since.

Sources: DART half-year report rcpNo 20260814003699 (incorporation, listing, TV share, DRAM share); DART FY2025 annual report rcpNo 20260310002820. Share figures are third-party research data as cited by the company; the H1 2026 figures are the company’s own estimates.

🌍 Is Samsung Actually a Global Leader, or Only a Korean One?

It leads or co-leads five separate global markets — DRAM, TVs, smartphone units, smartphone OLED panels and automotive digital cockpits — and in the first half of 2026 it gained share in all five at once.

Source: DART half-year report rcpNo 20260814003699, section II.1. Four of the five 2026 rows are explicitly labelled in the filing as the company’s own estimate (DRAM, TV and digital cockpit say “2026년 반기 시장점유율은 당사 추정치입니다”; the OLED-panel note says “2026년 2분기 시장점유율은 당사 추정치”). The smartphone row carries no such label — it cites TechInsights unit-share data with no estimate caveat, so treat it as differently sourced from the rest of the column. The 2024 and 2025 columns are the named research firms’ data as reproduced by Samsung. We have not independently verified any of them.

Hold that table next to the profit table below. Samsung gained share in every consumer market it competes in — and its consumer division still lost money in the second quarter. That is the whole article in two sentences.

📈 Is the Business Getting Better or Worse — and Why One Year of ROE Tells You Nothing

Better, by a margin that makes single-year ratios useless: operating profit in the first half of 2026 was $109.0bn, more than triple the whole of FY2025, on revenue that rose 98.7% year on year.

Sources: FY2024/FY2025/H1 2026 from DART half-year report rcpNo 20260814003699; FY2023 from DART FY2025 annual report rcpNo 20260310002820. *FY2023 carried a net tax credit, so net profit exceeded pre-tax profit. Two ROE rows, deliberately. Samsung states its own consolidated ROE in the management-analysis section of the FY2025 annual report (rcpNo 20260310002820): “ROE 10.8%” for FY2025, “전년 대비 1.8%p 증가” — implying 9.0% for FY2024 — plus 13.7% on a parent-company basis. Our row divides net profit attributable to owners of the parent by period-end controlling equity, which in a year of fast equity growth reads lower than an average-equity basis; the H1 figure doubles the half-year profit and is not a forecast. Where the filing states a number we print the filing’s. Converted at ₩1,345.99/$1 throughout, so the series shows business change, not currency change.

Two details in that table are worth more than the headline. First, the effective tax rate nearly tripled, from 8.6% to 22.5%. Samsung’s low-8% tax rate in 2024–25 came from Korean national-strategic-technology R&D and capex credits; those credits are largely fixed in size, so as profit exploded they stopped mattering. Second, almost none of the profit is paper. Pre-tax profit of $113.9bn sits only $4.9bn above operating profit — non-operating items are 4.3% of the total. Samsung booked $10.7bn of after-tax gains on its investment portfolio in the half, and every dollar of it went to other comprehensive income rather than to net profit, because those holdings are classified FVOCI. (Source: rcpNo 20260814003699, consolidated statement of comprehensive income — the ₩14,351,234m line is presented net of tax, as the statement’s other-comprehensive-income items are.) That is a genuine quality-of-earnings advantage over peers whose investment marks run through the income statement; we walked through the opposite case in our SK hynix reference article.

🔍 Why Does Half of Samsung’s Divisional Asset Base Produce 2.6% of Its Profit?

Because the asset split barely moves and the profit split swings violently: on the same roughly 50/50 divisional balance sheet, the non-semiconductor divisions contributed 53.4% of divisional operating profit in FY2024, 42.4% in FY2025 and 2.6% in the first half of 2026.

Samsung’s English quarterly earnings release publishes revenue and operating profit by business. It does not publish total assets by business, and it uses a different grouping (MX & Networks; VD/DA & Harman) that cannot be lined up against the DART divisions without doing the arithmetic yourself. The asset column appears in section II.7-라 (사업부문별 요약 재무 현황) of the Korean regulatory filing — and that filing prints three periods side by side, which is what stops the number being misread.

Source: DART half-year report rcpNo 20260814003699, section II.7-라, filed Aug 14, 2026 — the filing prints three adjacent columns — 제58기 반기 (H1 2026), 제57기 (full-year FY2025) and 제56기 (full-year FY2024). Only share-of-total is comparable across these columns, because FY2024 and FY2025 are twelve-month figures and H1 2026 is six months; asset shares are balance-sheet stocks and compare directly. Rows are the sum of the filing’s printed percentages and do not reach 100% because a small residual is unallocated.

With that caveat fixed in place, here is the level detail for the latest period:

Source: DART half-year report rcpNo 20260814003699, section II.7-라 (divisions) and the consolidated balance sheet and income statement in the same filing. Percentage columns are the filing’s own. Read the asset column carefully. The filing states that every line of this table — assets included — is presented including inter-division transactions, with common assets allocated across divisions on the company’s own basis. The four divisions therefore sum to $796.2bn of assets against a consolidated balance-sheet total of $564.3bn — 41% more. Revenue is inflated the same way but only by 8.9%, and operating profit is essentially clean (the four divisions sum to within 0.03% of the consolidated figure). So the percentages above are shares of a divisional total, not of the group balance sheet. On the consolidated total, DS’s $412.9bn is 73.2%. Converted at ₩1,345.99/$1.

The ratio is worth stating carefully because the denominator is inflated. On divisional assets, DS’s operating profit annualises to a 51.4% return (25.7% for the six months as reported); the other three divisions annualise to 1.5% (0.7% for the half). Those are not true returns on assets — the denominator includes intra-group items — but the two are computed identically, so the ratio between them, roughly 35 to 1, survives the distortion even if neither level does.

Subtract Q1 and you can see the quarter

The half-year filing reports cumulative figures. The Q1 filing (rcpNo 20260515002181) reports the same table for three months. Subtract one from the other and you get the second quarter by division, on Samsung’s own DART definitions:

Sources: divisional figures from rcpNo 20260515002181 (Q1) and rcpNo 20260814003699 (H1), section II.7-라 of each; consolidated figures from the H1 income statement in the same filing. The division sum reconciles to the consolidated total within 0.03% — the residual is intersegment elimination.

The DX division ran an operating loss of about $0.61bn in the second quarter of 2026. Korean and international outlets did report a first-ever quarterly loss at Samsung’s device business when the July 30 provisional results came out, so the fact is not new. What the subtraction adds is the division-level number on the same definition Samsung uses in its own filings, next to the asset base that produced it — and the asset base is the part you cannot get in English.

What the filing does and does not let you say about why

The filing supports a juxtaposition, not a causal chain: memory average selling prices rose about 220% against the prior-year average while smartphone prices rose 7% and TV prices fell 4%, and the company describes component-price inflation squeezing its consumer markets. It does not disclose what DS charges DX, so we cannot tell you how much of the DX result is internal pricing.

Here is what the company actually wrote, in Korean, in the same filing. On the smartphone market: “2026년은 주요 부품 가격 상승 부담으로 시장 규모가 11.2억대 수준으로 감소할 전망이나, AI 기능 개선, 폴더블 시장 확대 등으로 프리미엄 수요는 유지될 것으로 전망됩니다” — the market is expected to shrink to around 1.12 billion units in 2026 under the burden of rising key-component prices, though premium demand should hold up. On TVs: “패널가격 상승세도 지속 중이어서 수익성 확보에 어려움이 가중되고 있습니다” — the continuing rise in panel prices is adding to the difficulty of securing profitability. And on memory demand, from the DS commentary: “모바일/PC의 경우, 세트 가격 인상 영향으로 수요 둔화 움직임이 보이고 있으나, 서버향 DRAM/SSD, HBM의 추가적인 수요 증가세가 이를 상회할 것으로 보여…” — signs of slowing demand in mobile and PC as set prices rise, but server DRAM, SSD and HBM demand growth is expected to more than exceed it. (Source: rcpNo 20260814003699, section II.1; our translations)

Note the direction of the last one, because it cuts against the neat version of this story: Samsung’s own framing is that server demand overwhelms the consumer weakness, not that one division is bleeding the other. Its System LSI commentary is likewise written as a problem being solved — “메모리 가격 상승에 따른 스마트폰 수요 둔화를 고부가 제품 수주 확대와 신규 응용처 확보를 통해 극복하고 있습니다” (“is overcoming the slowdown in smartphone demand caused by rising memory prices through expanded high-value-added orders and new applications”).

So the honest statement is narrower than the tempting one. Rising memory prices are an industry-wide input cost for every phone and TV maker, Samsung included, and Samsung is on both sides of that trade. How much of DX’s second-quarter loss came from component inflation, how much from the won, how much from operating costs or one-off items, is not something these filings decompose — and we are not going to guess at it. What we can say is that the two halves of Samsung now move on opposite cycles, and that in this half the consumer half lost money while its own market shares rose.

That has a consequence for how you read “diversification.” On the divisional balance sheet, 47.6% of assets sit outside semiconductors — but in this period that half did not offset the memory cycle, it moved with the same input-cost shock. The same structural question, asked of the pure-play alternative, is in our SK hynix deep-dive; for the cycle itself, see Korea’s chip supercycle.

💰 What Do You Actually Pay for Samsung Today — and Which Market Cap Is Right?

Neither of the two market caps you will see quoted is wrong; they answer different questions. Common shares alone are worth $1,109.8bn. Add the listed preferred class at its own price and Samsung’s listed equity is worth $1,224.0bn.

Samsung Electronics has two listed share classes, not one and not three. We checked this two ways: the full KRX daily snapshot for September 4, 2026 contains exactly two Samsung Electronics lines (005930 and 005935), and the share-total schedule in the half-year filing lists exactly two classes (보통주 / 우선주).

Source: Korea Exchange Open API, settled daily quotations for 2026-09-04 (data.krx.co.kr). Share counts cross-check to the half-year filing’s share-total schedule, rcpNo 20260814003699.

There are two ways to get this wrong, and they push in opposite directions. Count only the common and you understate Samsung’s listed equity by $114.2bn — more than the entire market value of Samsung Electro-Mechanics. Count all 6.65bn shares at the common price and you overstate it by $38.1bn, because the preferred trades at a 25.0% discount to the common. The preferred is 9.3% of total listed value. We worked through the same arithmetic, with July 2026 prices, when correcting our earlier Samsung Electronics valuation note — that piece is the point-in-time record and carries the peer-multiple discussion; this one is the structural view.

The same “which one” problem hits the P/E harder:

Net profit attributable to owners of the parent. Trailing twelve months = FY2025 (₩44.261tn) − H1 2025 (₩12.962tn) + H1 2026 (₩118.371tn) = ₩149.670tn, or $111.2bn. Won figures are shown here so the sum can be checked directly against the filing. All components from rcpNo 20260814003699. Price-to-book on controlling equity of $419.8bn (₩565.06tn at June 30, 2026) is 2.64× on the common alone, 2.92× on both classes.

One more structural item the ratios hide, and here is the arithmetic in full because the filing does not present this subtotal:

Every input line is taken from the consolidated balance sheet in rcpNo 20260814003699; the subtotals are ours, not the filing’s, and another analyst could reasonably draw the operating/non-operating line elsewhere. Converted at ₩1,345.99/$1.

Under our house valuation standard, that $164.5bn is the wedge between a “basic” value that excludes non-operating assets and a “diluted” value that includes them. Strip them out and the operating business is running on roughly $255bn of equity — which is one reason division-level returns look so different from group ROE.

🏷️ Which Samsung Ticker Are You Actually Buying?

The Korea Exchange lists 26 securities whose name begins with “Samsung,” and three of them have nothing to do with the Samsung group. Samsung Electronics is one listed company among 18 listed Samsung affiliates, and it does not own most of the others.

We verified this from the roster, not by absence. The half-year filing states that the Samsung group had 68 domestic affiliates as of June 30, 2026, of which 18 are listed, and its cross-shareholding schedule names them: Samsung Electronics, Samsung C&T, Samsung Life, Samsung Fire & Marine, Samsung Biologics, Samsung Epis Holdings, Samsung SDI, Samsung SDS, Samsung Electro-Mechanics, Samsung Heavy Industries, Samsung E&A, Samsung Securities, Samsung Card, Samsung FN REIT — and four that do not carry the Samsung name at all: S-1 (에스원), Cheil Worldwide (제일기획), Hotel Shilla (호텔신라) and Multicampus (멀티캠퍼스). Samsung Pharm, Samsung Climate Control and Samsung Publishing are not on that roster. (Source: rcpNo 20260814003699, section IX; group status separately filed at rcpNo 20260831000250)

So the name cuts both ways: three KRX tickers say “Samsung” and are not in the group, and four group members do not say “Samsung” at all.

Market values from the Korea Exchange Open API, settled quotations for 2026-09-04, converted at ₩1,345.99/$1. Group membership verified against rcpNo 20260831000250. Abbreviated list — the KRX also carries preferred lines for several of these plus four Samsung Securities SPACs.

One practical note for a non-Korean investor. Unlike SK hynix, Samsung Electronics has no liquid US exchange listing: as we established when we mapped the dollar prices, Samsung’s US quotes are thin over-the-counter and GDR lines, not a tradable ADR, and the OTC line can sit well away from the converted Seoul price in either direction. If a US screener is showing you a dollar figure for Samsung, read Samsung & SK hynix stock price in USD before you act on it; the mechanics of buying the Seoul line are in our access guide, and the full list of Korean names that do have real US listings is here.

🏛️ Who Controls Samsung Electronics, and What Do Minority Holders Get?

The controlling bloc holds 19.69% of the common stock, but more than half of that stake — 10.11% of the company — is restricted from voting on ordinary business at general meetings, with a carve-out that lets it vote on director elections and changes to the articles. Chairman Jay Y. Lee personally owns 1.67%.

Source: DART half-year report rcpNo 20260814003699, section VII.1 (largest shareholder and related parties) and section VI.3-나 (voting rights). Common shares only — the same parties also hold 545,816 preferred shares, which carry no vote in any case. The rows sum exactly to the total. “Restricted” means the shares fall within the 590,994,010 common shares the filing identifies as barred from voting under Korea’s Monopoly Regulation and Fair Trade Act, with a partial carve-out for director elections and articles amendments; a further 208,051 shares are restricted under the Insurance Business Act.

That structure is the Korea Discount in miniature, and it cuts both ways. The family’s own direct holding is 4.41%. Control runs through the affiliate chain — Samsung C&T and Samsung Life — rather than through direct ownership, which is exactly the arrangement Korea’s reform agenda has been aimed at. But the filing also shows the family selling: Hong Ra-hee’s stake fell by 15,000,000 shares during the half, disposed of in off-hours block trades. If you want the mechanics of Korean holding-company chains, our Korea Discount explainer and Value-up guide cover the ground.

So what did the August 21 shareholder-return announcement actually commit to?

One thing: a ₩15tn ($11.1bn) buyback of 53,285,968 common shares whose stated purpose in the filing is employee stock compensation — not cancellation. The ₩90–110tn ($67–82bn) headline is labelled, in the filing itself, as forward-looking estimate.

The Korean fair-disclosure filing (rcpNo 20260821800763) sets out the sequence plainly:

• The 2024–2026 policy returns 50% of three-year cumulative free cash flow, with a regular dividend of ₩9.8tn a year. ₩29.3tn ($21.8bn) was executed across 2024–25.
• The 2026 remainder is estimated at ₩90–110tn, “subject to change depending on 2026 operating results and investment.” The FCF base excludes memory long-term-agreement prepayments and employee stock-compensation outlays.
• About ₩30tn ($22.3bn) of cash dividend is planned for Q3 2026, with details to be fixed at a board meeting in late October 2026.
• Everything else — “cash dividend and treasury share purchase/cancellation considered together” — is deferred to a board meeting in late January 2027.

The separate board resolution filed the same day (rcpNo 20260821000616) is the concrete one: 53,285,968 shares, ₩14,999,999,992,000, on-market purchases between August 24 and November 21, 2026, sized at the August 21 close of ₩281,500. Its stated purpose is share-based compensation. That filing also closes out the company’s two earlier “rumour response” disclosures about a reported ₩90tn ($66.9bn) buyback, confirming only the ₩15tn ($11.1bn) employee-compensation programme.

The treasury ledger in that filing tells you what has actually happened to Samsung’s own shares in 2026:

Source: rcpNo 20260821000616, treasury-holdings schedule. The bridge closes exactly. The entire preferred treasury holding of 13,603,461 shares was cancelled in the same April action, leaving zero. The shares cancelled in April were bought under the 2025 programme; the FY2025 annual report states that the ₩5.3tn ($3.9bn) cancelled was the whole of the 2025 purchases excluding the portion earmarked for employee compensation.

So: Samsung cancelled 1.24% of its common stock in April — using shares bought under the 2025 programme — and then authorised two fresh buybacks. Both 2026 buyback resolutions state the same purpose, and it is not cancellation.

Sources as linked. Filed amounts are ₩7,174,300,000,000 (March, sized at the Mar 17 close of ₩193,900) and ₩14,999,999,992,000 (August, sized at the Aug 21 close of ₩281,500), together ₩22.17tn, converted at ₩1,345.99/$1. The March filing names the purpose explicitly: “2025년 10월에 도입한 「성과연동 주식보상(PSU)」과 성과인센티브(OPI, LTI) 지급 등 주식기준 보상에 사용할 목적” — for the performance stock unit scheme introduced in October 2025 and for OPI/LTI incentive payments. Buying periods were Mar 19–Jun 18 and Aug 24–Nov 21, 2026.

Together that is $16.47bn of authorised buying in 2026, all of it earmarked for staff compensation, against $9.8bn of treasury purchases actually settled in the first half (cash-flow statement, 자기주식의 취득 ₩13,248,625m). Whether the January 2027 board turns the remaining return pool into cancellation or dividend is, on the evidence of the filings, genuinely undecided — and it is the single biggest open question for a minority holder.

There is a reason the company needs the shares. The same half-year report discloses 67,407,770 potential common shares arising from stock-based compensation arrangements, against 983,798 in the equivalent period a year earlier — a 68-fold increase, and about 1.15% of the issued common count. Samsung introduced a performance-linked share unit scheme in October 2025, and it transferred 12,063,951 treasury shares to staff in the first half of 2026 alone. (Source: rcpNo 20260814003699, earnings-per-share note) A buyback that offsets employee issuance is a real use of cash, but it is not the same thing as retiring stock, and a reader should not count it twice.

Meanwhile the dividend rail has not moved:

Source: DART half-year report rcpNo 20260814003699, section III.6. Payout is on net profit attributable to owners of the parent. The payout ratio collapsed because the policy fixes the regular dividend in won terms while earnings rose ninefold — that gap is precisely what the ₩90–110tn ($67–82bn) announcement is meant to close.

One governance detail that will not appear in any English summary. The same filing asks whether the company can set the dividend record date after the dividend amount is decided — the fix Korea’s 2024 dividend-procedure reform asked listed companies to adopt, so that investors know what they are buying before the record date. Samsung’s answer for both annual and quarterly dividends is “X” (no), and its plan is recorded as “under review” (검토중). For the largest company on the exchange, in the year of the largest shareholder return in Korean corporate history, that is a striking blank.

🌏 Is Foreign Money Buying or Selling Samsung?

Selling, in every window we can measure — but the size of the selling depends so heavily on which window you pick that the number is nearly meaningless on its own.

Source: our own investor-flow database, built from Korea Investment & Securities Open API daily investor data. Method: each day’s net share count multiplied by that day’s settled close, then summed — an approximation, not execution prices. Converted at ₩1,345.99/$1. Foreign ownership of the common stock stood at 46.7% on Sep 4, 2026.

Look at the 10-day and 20-day rows. Ten sessions show seven times the selling of twenty sessions — because inside the 20-day window foreigners were heavy buyers in mid-August, at prices between ₩255,000 and ₩275,000, and then sold in late August at ₩257,000. A writer who wanted to tell you “foreigners are dumping Samsung” would quote the 10-day figure; one who wanted the opposite story would quote the 20-day. Both are true; neither is the whole picture. And note what the price did: over those same 20 sessions the shares rose 10.6% while foreigners were net sellers throughout. Foreign selling is not a price forecast, and we will not present it as one. We publish these side by side for exactly that reason, and track the market-wide version weekly in our Foreign Flow Watch. The general mechanics of who moves Korean stocks are in our flows explainer.

📚 Lingo Check

🎯 Why It Matters for K-Export Stars

Samsung Electronics is the most-covered stock in Asia, and the wires do their job well. The record profit, the device division’s first quarterly loss, the ₩90–110tn return headline and the fact that the ₩15tn buyback is for employee compensation were all reported in English within hours, accurately. We are not claiming otherwise, and this article says so where it matters.

What the English record does not contain is the denominators. The quarterly release gives profit by business and no assets by business. The three-period comparison that turns “97.4% of profit” into “97.4% this half, 46.1% two years ago” is printed in the Korean filing and nowhere else. The share-class arithmetic, the treasury bridge, the roster that tells you three KRX tickers named “Samsung” are not Samsung, and the schedule showing that roughly half the control bloc votes only on a restricted set of resolutions — all of it is regulatory text, filed in Korean, that no one translates because no one has to.

That gap is not a scoop; it is a language barrier, and it is the only durable edge a site our size has. The English release is a summary written for an audience. The Korean 반기보고서 is a document written for a regulator, and regulators ask for the boring columns. This piece is the Samsung half of a pair — the SK hynix half is here — and the same question, asked of both, produces opposite answers about what “diversification” is worth.

Disclaimer: This article is for informational and educational purposes only and is not investment advice, a recommendation, or an offer to buy or sell any security. We do not publish price targets or ratings. All figures are point-in-time as of the dates stated and were sourced from official filings (DART), the Korea Exchange, and our own investor-flow database; prices, multiples and flow figures go stale quickly. Market-share figures for 2026 are the company’s own estimates as labelled in its filing and have not been independently verified. Do your own research and consider consulting a licensed adviser before investing. See our corrections policy.