The SK ecosystem - one chaebol, three ways to own the AI memory leader

The standard retail trade on "the SK AI story" is $SKM, SK Telecom's NYSE ADR. It is liquid, it is on a US exchange, it has SK in the name, and it roughly doubled off its 2025 base. The problem: SK Telecom does not own SK hynix. It has not owned a single share since November 2021, when the stake was spun off into a separate company. Most of the capital chasing "SK = HBM = AI" through SKM is buying the right re-rate for the wrong reason: SKM's actual AI asset is not memory at all, it is one of the earliest outside stakes in Anthropic.

The actual exposure sits elsewhere in the family tree. SK Group is a chaebol: a constellation of listed companies tied together by cross-shareholdings under one founding family, and the AI-supercycle asset everyone wants, $HXSCL (SK hynix, the global HBM leader), hangs off a holding company called SK Square that most US investors have never heard of. This piece maps the whole tree: who owns what, which of the three plausible tickers carries which exposure, and why the wrapper you pick changes the trade more than the thesis does.

Why it matters now

Three clocks are running at once. SK hynix filed confidentially with the SEC in March 2026 for a US ADR listing targeting end-2026, which would make the crown jewel directly buyable on a US exchange for the first time, see SK hynix is heading to a US listing. SK Square, the holdco that owns 20.5% of hynix, has nearly tripled and joined the KOSPI's million-won club as its NAV discount narrowed from 65.7% at end-2024 to roughly 42-45% by May 2026, making it the poster child of Korea's Value-Up corporate-reform program. And SK Telecom re-rated alongside both, on a narrative it only partially deserves.

The family tree

SK Group is South Korea's second-largest chaebol. The simplified ownership map, top down:

• SK Inc. (KRX 034730) sits at the top: the group holding company, controlled by the founding Chey family. It holds the controlling stakes in the major listed subsidiaries below it.
• SK Square (KRX 402340) is the technology-investment holdco, spun off from SK Telecom in November 2021. Its defining asset: a 20.5% stake in SK hynix, which makes it the chipmaker's largest shareholder. Around that core sit smaller stakes in semiconductor and platform businesses.
• SK hynix (KRX 000660 / $HXSCL OTC) is the crown jewel: the world's HBM market-share leader at roughly 60%, lead HBM3E and HBM4 supplier to NVIDIA, and the anchor of QA's DRAM / HBM Memory bubble.
• SK Telecom ($SKM, NYSE ADR) is Korea's dominant mobile carrier with ~45% market share. Since the 2021 spin-off it is a sibling of SK Square, not a parent of anything semiconductor.
• The supporting cast: SK Innovation (energy; its SK On unit makes EV batteries), SKC (semiconductor materials; its Absolics unit builds glass substrates for advanced packaging in Georgia, USA), SK Ecoplant (construction and, increasingly, datacenter builds).

The structural quirk that matters: in a chaebol, the economics and the control are not aligned the way US investors expect. SK Inc. controls SK Square; SK Square's value is mostly its hynix stake; but the market prices each layer separately, and every layer of holding-company wrapping historically traded at a steep discount in Korea. That discount is the second trade in this piece.

The crown jewel, briefly

SK hynix is covered in depth in its own explainer and in HBM is the tightest bottleneck in the AI cycle, so the short version: it leads the part of the memory market with the most pricing power (HBM, sold on multi-year contracted capacity to NVIDIA and the hyperscalers), it out-earns and out-margins $MU while trading at a lower forward multiple, and it has been the single biggest beneficiary of the memory supercycle on the KOSPI. Everything else in the SK complex trades, to a first approximation, as a derivative of this asset.

Trade 1 - own the asset: SK hynix directly

The cleanest expression. The primary listing is KRX 000660 in KRW; the US OTC ADR is $HXSCL, real but thin. The SEC-registered ADR listing targeted for end-2026 (est. $10-14B raise) is the event that turns this from an access-constrained trade into a one-click one.

• What you get: pure HBM/DRAM/NAND exposure, no holdco wrapper, no telecom attached.
• What you pay: memory-cycle beta in full, plus dilution from the primary raise when the ADR prices, plus KRW exposure.
• Access today: KRX via a broker that supports it (see /stack/ibkr for the mechanics), or the thin OTC line. Track it live on /stocks/hxscl.

Trade 2 - own the wrapper: SK Square and the closing discount

SK Square is the leveraged-but-laggy version of the same trade. Its 20.5% hynix stake dominates its net asset value, and the stock historically traded at a Korea-typical holdco discount of well over half. Then two things happened at once: hynix went vertical, and Korea's Value-Up program (the corporate-reform push modeled on Japan's, pressuring listed companies to close price-to-book gaps via buybacks, dividends, and governance changes) gave the market a reason to re-rate the wrapper itself.

The result is the standout chart of the Korean market: SK Square nearly tripled, crossed the symbolic 1,000,000-won mark, and climbed to the #3 market cap on the KOSPI, behind only Samsung Electronics and SK hynix itself. The NAV discount narrowed from 65.7% at end-2024 to 51.5% at end-2025 to roughly 42-45% by spring 2026.

The mechanics of the trade are worth being explicit about, because they cut both ways:

• When the discount narrows, SK Square outperforms hynix (you earn the asset's move plus the re-rate). That is what the last 18 months were.
• When the discount is done narrowing, SK Square becomes hynix-with-extra-steps: same downside in a memory downturn, plus governance and conglomerate risk, minus the liquidity.
• There is no clean US wrapper. SK Square trades in Seoul (KRX 402340); US OTC volume is negligible. This is a KRX-access trade, full stop.

Trade 3 - the mislabeled one: SK Telecom

Here is the correction this article exists to make: SK Telecom has owned zero shares of SK hynix since November 2021. The stake went to SK Square in the spin-off. Any pitch that says "buy SKM for the hynix exposure" is describing a company that stopped existing five years ago.

What SKM actually is: Korea's largest carrier, roughly $14.6B market cap, a 3.7% dividend yield, forward P/E around 16, revenue roughly flat, beta 0.67. A classic defensive telecom, with one asset that is anything but: a stake in Anthropic, started with a $100M check in August 2023 and topped up in the $965B Series H that closed in May 2026. SKT's CEO confirmed both the top-up and the intent to hold through Anthropic's IPO (targeted H2 2026) in Tokyo on June 10. At the Series H mark, Morningstar describes the Anthropic position as dominating SK Telecom's valuation, on a company whose entire market cap is ~$14.6B.

The relationship runs deeper than a passive stake. SKT joined Anthropic's invite-only Project Glasswing in June 2026, gaining early access to Claude Mythos, the gated cybersecurity model previously limited to the US government and select firms (SK hynix and Samsung also received access). And per the Korean coverage of the Tokyo remarks, Anthropic is actively seeking compute capacity that SKT intends to supply from its AI-datacenter buildout. The full proxy math, alongside AMZN, GOOGL, and the funds, is in How to invest in Anthropic.

So the correction cuts both ways: SKM is not the hynix play retail assumes, and it is also not "just a telecom." It is an Anthropic proxy wearing a carrier's clothes. If you bought it for AI-memory exposure you got the wrong asset; if you bought it for the Anthropic stake you got one of the few public wrappers that exists. Track it on /stocks/skm.

The supporting cast

Two group companies earn a mention in an AI-supercycle map:

• SKC / Absolics: SKC's Absolics subsidiary is building glass-core substrates for advanced chip packaging in Covington, Georgia, with US CHIPS Act support. Glass substrates are a credible next step for the packaging layer that HBM-heavy accelerators are straining, the same substrate-bottleneck territory as the T-Glass story. Early, capex-heavy, optionality rather than earnings.
• SK Innovation / SK On: EV batteries and energy. Strategically central to the group's balance sheet (and its capital-raising decisions), mostly irrelevant to the AI-memory thesis except as a reminder that the chaebol allocates capital across the whole family, not just the jewel.

The numbers

| Vehicle | Listing | What it is | The catch | | --- | --- | --- | --- | | SK hynix | KRX 000660 / HXSCL (OTC) | ~60% HBM share, the asset itself | Thin US access until the ADR lists (end-2026 target) | | SK Square | KRX 402340 | 20.5% of hynix, ~42-45% NAV discount | Seoul-only, discount can stall or widen | | SK Telecom | SKM (NYSE) | ~45%-share Korean carrier, 3.7% yield, fwd P/E ~16, pre-IPO Anthropic stake | Owns no hynix; the AI asset is the Anthropic mark, unrealized until its IPO | | SK Inc. | KRX 034730 | Group holdco above everything | Double-layer discount, governance-driven |

The bull case

• The asset is real. Roughly 60% of HBM, the lead supplier on NVIDIA's current and next memory generation, in the tightest-supply corner of the AI stack.
• Two re-rating engines, not one. The hynix earnings cycle plus the Korea Value-Up discount compression. SK Square's 18-month chart is what it looks like when both fire together.
• The access unlock is dated. A liquid US hynix ADR targeted for end-2026 widens the buyer base for the whole complex and gives every wrapper a cleaner reference price.
• The discount still exists. Even after the rally, SK Square's NAV gap of roughly 40%+ leaves room if Value-Up pressure persists; Japan's analogous reform cycle ran for years, not quarters.

The bear case

• Memory is still a cycle. Every layer of this tree is, ultimately, a claim on DRAM/NAND/HBM pricing. When the cycle turns, the holdco discount historically widens at exactly the wrong moment, leverage in both directions.
• Chaebol governance is the discount's reason. Cross-shareholdings, family control, and capital allocation that serves the group (battery capex, rescue financings) before minority shareholders. Value-Up is policy pressure, not a law of nature.
• Dilution overhang. The hynix US listing is backed by a large primary raise; new supply near the listing can cap the complex.
• FX and access friction. KRW exposure on every vehicle except SKM, and two of the four tickers are realistically Seoul-only for US retail.
• The narrative-fit risk runs both ways. The same flows that bought SKM for exposure it does not have can leave it just as fast when the label peels.

How to access

• Want the asset: SK hynix via KRX (000660) if your broker supports it, the thin HXSCL OTC line as the placeholder, or wait for the US ADR. Live data: /stocks/hxscl. The practical US-listed proxy for the same memory thesis remains MU, see Micron explained.
• Want the discount trade: SK Square, KRX only. This requires real KRX access (/stack/ibkr covers which account setups can reach Seoul).
• Want the Anthropic proxy (or the dividend telecom): SKM on the NYSE, eyes open about what it is, a carrier whose dominant AI asset is a pre-IPO stake. Proxy math in How to invest in Anthropic. Live data: /stocks/skm.
• Want the basket view: the Korea, US-accessible watchlist tracks the Korean names a US brokerage account can actually hold.

What to watch

• The hynix US ADR terms: final size, the new-share component, the symbol, and timing into end-2026. The single biggest event for every wrapper in the tree.
• SK Square's NAV discount: whether it holds in the 40s, keeps compressing on Value-Up momentum, or re-widens on a memory wobble. The discount is the trade.
• HBM4 share retention: hynix versus Micron and Samsung on the next NVIDIA generation; covered in the memory-bubble work, it sets the asset's earnings path.
• Value-Up policy cadence: index inclusion rules, tax incentives, and whether holdco reform survives Korea's political cycle.
• The Anthropic IPO (targeted H2 2026): the event that converts SKM's dominant unlisted asset into a marked, sellable position, and the single biggest catalyst SKM does not control.
• SKM earnings 2026-08-05: watch how much of the re-rate the actual telecom numbers can support, and the scale of the AI-datacenter segment against the narrative.

---

Live data on these tickers: /stocks/hxscl and /stocks/skm - price, bubble correlation, related research.

Bubble context: /bubbles/memory - the DRAM / HBM Memory cluster the crown jewel anchors.

Adjacent reading: SK hynix is heading to a US listing, Micron (MU) explained, and HBM is the tightest bottleneck in the AI cycle.

QuantAbundancia is educational research. Nothing here is investment advice. See /disclosures.