Why SK Hynix Wants Its Memory Unit Listed in New York

SK Group Is Looking to Wall Street, Not Seoul

SK Hynix is preparing to take its NAND memory business public in the United States — and if the numbers being floated are accurate, this will be the single biggest semiconductor IPO in American market history.

Reuters reported this week that Solidaim, the company that grew out of SK Group’s 2020 acquisition of Intel’s NAND division, held investment bank proposal meetings with underwriters. The target: a valuation of up to $150 billion and a raise of roughly $15 billion. For context, ARM, the chip-design firm that listed in London in 2023, came in at $54 billion. Cerebras, the AI-chip startup that went public this year, reached $56 billion. Solidaim, if these figures hold, would dwarf both by nearly three times.

The deal would be a strategic choice as much as a financial one. SK Group has been gradually turning to American capital markets for reasons that go beyond convenience. It listed SK Hynix’s American Depositary Receipts on Nasdaq in July 2026. Bloomberg reported that SK is also weighing an ADR listing for Kioxia — the Japanese NAND manufacturer it now controls as the second-largest shareholder — to raise at least $10 billion more.

Each move signals the same calculation: Korea’s capital markets simply cannot absorb the scale of funding required to win in memory chips. The hyper-scale investment cycle — new fabs, R&D pipelines, capacity expansion — demands a pool of investor money that Seoul’s exchanges do not have in sufficient depth, especially for a business whose fortunes can swing sharply with commodity cycles.

The Four-Layer Overlap Problem

What makes Solidaim’s planned US listing particularly contentious in Korea is the corporate structure it sits inside.

SK Group’s semiconductor control pyramid runs like this: SK Holdings (listed in Korea) sits at the top, controlling SK Square (also listed in Korea), which controls SK Hynix (listed in both Korea and the US), which controls AI Company, which controls Solidaim. If Solidaim lists in the US, that makes four publicly traded entities sitting atop one another — what Korean market observers call an extreme case of “over-listing” or 중복 상장.

The concern is not merely aesthetic. SK Holdings and SK Square already capture significant earnings from their downstream stakes, particularly in SK Hynix. Adding Solidaim as a fifth (or sixth, if Kioxia’s ADR materializes) layer on top of that means a growing portion of Solidaim’s future value will flow to American and other foreign shareholders who own no part of the ultimate parent, SK Group.

The Korea Corporate Governance Forum issued a blunt assessment in August when the listing first surfaced, calling the strategy a way to evade Korea’s Fair Trade Act restrictions on multi-level pyramidal ownership structures — restrictions that are supposed to prevent exactly this kind of concentrated family-group control without proportional ownership. The forum warned that using foreign listings to extend the pyramid effectively exports the “Korea discount” to markets where investors have less recourse and fewer protections.

For SK Hynix shareholders specifically, the math is straightforward. When Solidaim sells shares to outside investors, SK Hynix’s indirect stake through AI Company gets diluted. That means less of Solidaim’s future profit shows up in SK Hynix’s own earnings. With SK Hynix’s share price already well below its 2021 peak — a drop driven by memory-chip price softness and broader market turbulence — any further dilution of the most profitable asset in its portfolio is likely to provoke resistance among domestic investors.

Why SK Is Proceeding Anyway

SK Group has not publicly confirmed the Solidaim listing timeline. But the group’s actions speak clearly: capital needs are urgent and the options at home are limited.

The Fair Trade Act does require 100% ownership when a listed company establishes or acquires a subsidiary, which forces SK to create overseas holding vehicles to bypass that rule. The group’s counterargument is pragmatic — a domestic entity would need to self-fund entirely for hyper-scale projects that no single Korean balance sheet can carry. Without external capital from public markets, the alternative is borrowing, which adds leverage at precisely the wrong point in the semiconductor cycle.

The timing of this push coincides with a deepening shortage in NAND supply. Data-center demand for enterprise SSDs continues to outstrip the capacity of incumbents, and Solidaim — with its heritage from Intel’s mature memory operations and its dedicated role in the server and cloud segment — is well positioned to capture margin in that gap. The group appears to believe the window for raising large sums in favorable conditions will not stay open indefinitely.

Chairman Choe Tae-won’s leadership style also bears on this calculation. He has consistently favored speed and scale in SK’s semiconductor strategy, often making moves that force Korea’s institutional investors and regulators to catch up. The Sk Hynix Nasdaq listing in July 2026 was one such move. Solidaim’s IPO would be another.

What Changes If the Deal Happens

A Solidaim listing would reshape several relationships at once.

Among investors, it would create a new, US-listed pure play on enterprise NAND — something the current SK Hynix ADR does not offer, since SK Hynix is half its story in DRAM and the other half is buried in a Korean corporate pyramid. Portfolio managers focused on AI infrastructure and data-center exposure could buy into Solidaim directly, without taking on SK Hynix’s governance concerns or SK Group’s broader conglomerate structure. That makes the US listing intelligible from a capital-market perspective even if it is painful for Korea-listed parents.

For SK Hynix, the risk is that the group’s crown jewel slowly loses economic weight relative to its listing status. The more equity Solidaim issues abroad, the thinner SK Hynix’s share of the downstream profit stream becomes. This is not theoretical — it mirrors what happens whenever a large subsidiary exits the parent’s consolidated earnings through a public offering.

For Korean market regulation, the episode is a stress test. The Fair Trade Act’s three-tier limit on pyramidal ownership is designed to prevent exactly the kind of opaque, multilayer control structure that SK is now constructing with foreign listings as a loophole. Whether regulators respond depends on whether they view Solidaim as a genuine business necessity or a structural arbitrage.

What Comes Next

The immediate next step is whichever investment bank Solidaim selects after its bake-off process. Deal terms, pricing, and timing will become clearer once that underwriting syndicate is locked in.

If the $150 billion valuation holds, Solidaim would not just be the largest semiconductor IPO in US history — it would be among the largest of any sector outside of tech mega-cap listings. That magnitude alone could change how Wall Street treats the company, giving it a pricing and liquidity profile that SK Hynix cannot replicate as a subsidiary.

The listing will also force a reckoning for SK Group’s governance. Investors who bought SK Hynix for its DRAM competitiveness and its relationship to SK’s broader strategy will want clarity on how much of Solidaim’s future value they are effectively ceding to American shareholders. Korea’s institutional investors, who already feel underrepresented in decisions about the group’s direction, are unlikely to accept a status quo that keeps deepening that gap.

The broader question is whether this is a one-off maneuver or a model. If SK successfully lists Solidaim and then moves forward with Kioxia’s ADR as Bloomberg suggests, the group would have built a two-legged American financing platform for its memory-business ambitions. Other Korean conglomerates watch closely. Each successful precedent makes it harder for regulators to justify blocking the next one.

The Real Bet

At its core, Solidaim’s US listing is a bet that American capital markets are now the primary venue where Korea’s semiconductor winners will be financed — not a fallback, but the main stage.

That bet makes sense given the cost of capital, the depth of investor interest in AI infrastructure, and the limitations of Seoul’s market. It does not make the trade-offs any easier. Every dollar raised in New York is a dollar whose claims on Solidaim’s future profits belong partly to investors with no tie to Korea. And every layer added to the corporate pyramid is a layer that increases the distance between the actual operators of SK’s memory business and the Korean shareholders who are told they own it.

The question for investors is not whether Solidaim will list, but what price they expect to pay for the convenience of having it done abroad.