The HBM leader is bringing its ~72%-margin memory business to Nasdaq on July 10 — ticker "SKHY," a ~$29.4B raise. Here’s the breakdown of all three Korean memory names, and why the US listing is the catalyst that re-rates the whole group.

SK Hynix is coming to Nasdaq. July 10, ticker "SKHY," ~$29.4B — the event that re-rates the group.

This is the news that makes the trio actionable for a US account. SK Hynix confidentially filed with the SEC in March, and on June 24 announced an ADR issuance worth up to ~$29.4B (45.45 trillion won), with trading set to begin on Nasdaq July 10 under "SKHY" (BofA, Citi, Goldman, JPMorgan leading). In scale it would eclipse Alibaba’s 2014 listing and Saudi Aramco’s 2019 IPO. The company’s own stated rationale is blunt: the listing lets its "true corporate value be properly evaluated" and valued in line with US peer Micron. Proceeds fund the Yongin fab cluster (online 2027) and a first US plant — a $4B packaging facility in Indiana. For you specifically: today these are KRX-only (buyable via IBKR direct), but after July 10 SK Hynix becomes a one-click US-listed name.

Three ways to own Korean memory. The leader, the comeback, and the holdco discount.

The elegance of the trio is that each expresses the same AI-memory tailwind differently: SK Hynix is the purest, highest-quality leader (and the one with the hard catalyst); Samsung is the cheaper, more diversified laggard staging an HBM4 recovery; SK Square is the financial-engineering angle — a holding company you can buy at a discount to the very SK Hynix shares it owns. Pick your risk profile.

The HBM king, about to be valued like one. Best-in-class economics at a fraction of US multiples.

SK Hynix is the highest-quality memory name on earth right now. It owns the HBM market (~58% share, NVIDIA’s primary supplier across HBM3E/HBM4 and the Vera Rubin platform), prints a 72% operating margin, and just became Korea’s most valuable company — yet trades at a ~6-8x forward P/E versus Micron’s richer multiple. That gap is the entire bull thesis: the Nasdaq listing is explicitly designed to close it. Long-term supply agreements (with price-protection clauses) are also reducing the historical cyclicality discount, which is why brokers are calling the memory valuation framework "rewritten." The catch: it’s up 300%+ this year, so you’re buying a parabolic move into a known event — and memory is still cyclical underneath the re-rating story.

The head-to-head — and the winner. Same supercycle, same NVIDIA customer — but SK Hynix is the better business at half the multiple.

SK Hynix’s entire listing pitch is one sentence: "value us like Micron." So the comparison isn’t academic — it is the thesis. Micron ($MU) just posted a blowout fiscal Q3 2026 (reported June 24), which gives a clean, current benchmark. Here’s the full breakdown:

What the numbers say. These are two of the most profitable hardware businesses on the planet, riding the same wave. Micron’s latest quarter was the more explosive print — revenue +346% to a record $41.5B, an 84.6% gross margin, $28.2B net income, and a guide to $50B next quarter — helped by a broader DRAM/NAND mix and U.S. accounting. SK Hynix’s edge is HBM dominance and margin leadership at the operating line — a 72% operating margin (above NVIDIA’s 65%), ~58% HBM share, and status as NVIDIA’s primary HBM partner. Both have shifted toward multi-year fixed-price contracts (Micron’s 16 "Strategic Customer Agreements" covering ~half of revenue; SK Hynix’s LTAs) — the structural change that is de-rating the old "memory is cyclical" discount for both.

The valuation gap is the whole point. On comparable, elite economics, SK Hynix trades at ~6-8x forward earnings while Micron trades at ~12x — roughly double. That’s the inefficiency SK Hynix is listing on Nasdaq to attack: put the same business in front of US investors, on the same exchange as Micron, and let the multiple converge. If SK Hynix re-rated even partway to Micron’s multiple on similar numbers, that’s meaningful upside before any further earnings growth — which is exactly why the July 10 debut is a genuine catalyst and not just a ticker change. Winner: SK Hynix — better HBM position, lower multiple, and a re-rating event dated to the calendar. You’re buying the #1 HBM franchise at roughly half Micron’s P/E. The bear counter: Micron’s ~12x isn’t demanding either — the whole group is cheap because the market still prices in a cyclical top, and if memory rolls over, both compress together.

The cheaper giant, catching up. HBM4 qualified, a buyback brewing, more than just memory.

Samsung is the value/comeback leg. It was the HBM laggard — late to qualify with NVIDIA — but that’s turning: it began mass-producing HBM4 in February and, on June 5, NVIDIA’s Jensen Huang confirmed Samsung (alongside SK Hynix and Micron) passed HBM4 qualification for the Vera Rubin platform. It holds ~21% HBM share (tied with Micron for second). Two things sweeten the value case: a rumored ~$65B (90 trillion-won) buyback that sparked a sharp rally, and the fact that Samsung is far more diversified than SK Hynix — memory plus a leading foundry, displays, and the world’s #1 smartphone business. So you get HBM upside with less single-product concentration, at an even cheaper multiple (~6-7x forward).

The trade-off is the mirror image of SK Hynix: that diversification dilutes the pure-HBM exposure (memory is a smaller share of the whole), and Samsung carries its own baggage — historical strike/labor risk and a foundry business that trails TSMC. There’s also chatter that Samsung could pursue its own US ADR listing after watching SK Hynix — a potential second catalyst for the group. The pick if you want memory upside with a margin of safety and less concentration; the laggard if you want pure HBM torque.

Own SK Hynix at a discount. The NAV-discount angle — same playbook as the ACMR thesis.

SK Square (402340) is the financial-engineering way to play this. It’s an investment holding company spun off from SK Telecom, and its single most important asset is a ~20% stake in SK Hynix (it’s the largest SK Hynix shareholder). Because holdcos routinely trade below the market value of the stakes they own — the classic NAV discount — SK Square can be a leveraged, discounted way to own SK Hynix: when Hynix runs, the value of Square’s stake runs with it, and any narrowing of the discount adds a second layer of return.

This is the exact structure we dissected in the ACMR thesis: asset-backed, discounted, with a catalyst that could close the gap. Here the potential catalyst is obvious — the SK Hynix Nasdaq listing crystallizes and globalizes the value of Square’s core holding, which could pressure the discount to narrow. The honest caveats are identical too: holdco discounts persist for real reasons (tax on selling the stake, the conglomerate "Korea discount," capital-allocation uncertainty), so the discount is real value but not freely accessible, and Square also holds other, smaller assets that muddy the pure-Hynix exposure. The highest-torque, most complex way to own the trio — a discounted, leveraged proxy on SK Hynix for those who like the holdco game.

How to play the trio. Leader, value, or leverage — with the listing as the shared catalyst.

The clean framing: SK Hynix is the core (best business, the hard Nasdaq catalyst, and after July 10 a one-click US name), Samsung is the value/diversified leg (cheaper, HBM4 recovering, buyback optionality, possible second ADR), and SK Square is the high-torque proxy (own Hynix at a discount, ACMR-style). All three are uptrending into the listing, and the US debut is a genuine re-rating catalyst — SK Hynix’s own words are that it wants to be valued like Micron. The honest risk that applies to all three: this is still cyclical memory, every name has had a massive run (SK Hynix +300% YTD), and a "sell-the-news" reaction around July 10 is a real possibility. The listing is the catalyst and potentially the sentiment peak — own the trend, but respect that you’re buying parabolic moves into a known date.