SK Square Co., Ltd. (KRX:402340) did not issue a single share this summer, and it still ended up with less of the only thing it owns. On July 10 its 20.5%-held affiliate SK hynix Inc. (KRX:000660, NASDAQ:SKHY) listed 177,900,000 American Depositary Shares on the Nasdaq Global Select Market at $149 each, raising roughly $26.5bn. Those ADSs sit on top of 17,790,000 newly issued Korean common shares, at a ratio of ten ADSs to one share. SK Square did not subscribe. Its stake went to about 20%.
Half a point of ownership is not a crisis. But the arithmetic of the transaction says something uncomfortable about where SK Square sits in the SK group, and the market has spent the seven weeks since telling us the price was wrong.
Work backwards from the offering. Ten ADSs at $149 is $1,490 per underlying Korean share. Local coverage put the implied won price at ₩2,249,751, a 2.72% premium to the reference price used at pricing. On August 27, SK hynix closed at ₩1,730,000. That's about 23% below where SK Square's ownership was cut.
Dilution below market is the normal complaint. This is the reverse, and it should have been the good version: existing holders gave up 2.5% of the company and received a premium for it. Seven weeks later the premium is gone and then some. Whether that's the memory tape or the digestion of $26.5bn of new supply, the filing itself doesn't say, and neither does anything SK Square has published. What the sequence does establish is that a very large block of hynix now sits with buyers whose entry price is well above the current one.
For SK Square the consequence is indirect but real. Its book claim on hynix shrank by half a point in exchange for cash it will never touch. The proceeds went into hynix's balance sheet for hynix's capex. SK Square got a smaller slice of a bigger pie and no dividend uplift for the trouble.
The structural point matters more than the half point of ownership. Before July, a raise of this size would have been a Korean market event, with the largest shareholder either writing a cheque or watching itself get diluted. SK Square, which had no realistic way to fund a proportional participation in a ₩40tn raise, would have been squeezed either way.
Instead the money came from US institutions, in dollars, through a structure that never asked SK Square for anything. That's a relief in the near term. It's also a demotion. A holding company whose relevance rests on being the anchor shareholder of a crown jewel is worth something when the jewel needs an anchor. When the jewel can print $26.5bn on Nasdaq in a single morning, the anchor is decorative.
There's a second-order effect on SK Square's own investor base that nobody at the company will put in writing. The stock has long carried a following built on the idea that you could buy SK hynix cheaper by buying its parent. That argument still works arithmetically. But a US investor who wanted hynix and could not easily reach the Korean market now has SKHY, which trades in dollars, settles in New York, and carries none of the holding-company complications. The pool of people who need SK Square as an access vehicle just got smaller. The pool who want it as a discount play is a different, more patient, and probably narrower group.
Here is the gap as of the August 27 close. SK Square traded at ₩1,068,000 across 131,923,998 shares, for a market capitalisation of ₩140.9tn, or roughly $101bn at an approximate rate of ₩1,400 to the dollar. SK hynix closed at ₩1,730,000 across 730,492,365 shares, for ₩1,263.8tn. Twenty percent of that is ₩252.8tn.
So the market values SK Square at about 56% of the market value of its hynix shares. The look-through claim is roughly ₩1,915,000 per SK Square share against a ₩1,068,000 price. Add the ₩1.32tn of cash on the June 30 balance sheet and the leftover portfolio, subtract ₩3.64tn of total liabilities, and the gap barely moves. It's a hynix discount with rounding attached.
Part of that discount is legitimate and US readers should not wave it away. If SK Square ever sold hynix shares it would owe Korean corporate tax on an enormous embedded gain, and the after-tax NAV is meaningfully below the gross number above. Korean holding companies also trade at structural discounts for reasons that predate this management team and will outlive it, including the absence of any credible path by which a minority holder forces a distribution of the underlying asset.
What is harder to defend is the direction. At the end of March, SK Square's net asset value was reported at about ₩123tn, of which the hynix stake accounted for roughly ₩118tn. Since then hynix has roughly doubled in market value and SK Square has not kept pace. A discount that widens while the underlying asset appreciates is the market saying it does not expect the gap to be monetised.
Worth noting on the book: SK Square carried ₩55.0tn of total equity at June 30, against that ₩140.9tn market cap. It trades at roughly 2.6x book and 0.56x look-through NAV at the same time, because the hynix stake sits on the books at equity-accounted cost plus accumulated share of profits rather than at market. Both ratios are correct. They just measure different things, and anyone screening SK Square on price-to-book is looking at the wrong number.
The bear case on my own framing is that the dilution is trivial and I've made too much of it. Half a percentage point of a stake that is compounding at the rate hynix's earnings currently are is noise. SK Square's equity-accounted share of hynix profit in Q2 2026 alone was large enough to swamp the arithmetic of a 2.5% share count increase within a single quarter. On that view the Nasdaq listing is straightforwardly good: it widened hynix's investor base, funded capex without leverage, and cost the parent almost nothing.
There's a governance-flavoured counter too. A larger, more international shareholder register at hynix may pressure the group toward cleaner capital allocation, and some of that pressure would flow up to SK Square. The August 19 announcement of a ₩40tn buyback and cancellation at hynix is at least consistent with a company that now answers to a broader audience. If SK Square does not sell into that buyback, its ownership percentage rises passively as hynix retires shares, which partially reverses July's dilution without costing SK Square a won.
And the cyclical risk cuts both ways. Memory pricing has been extraordinary. If it holds, the discount is a gift. If it turns, SK Square is a leveraged expression of a downcycle it cannot hedge, with a portfolio too small to cushion anything.
Two things, both dated. First, SK hynix's Q3 report, due in the autumn, will show the post-listing share count and the major shareholder table. That's where SK Square's exact stake gets confirmed rather than estimated, and where the effect of the ₩40tn buyback on the denominator starts to appear. The 17.79m new common shares were scheduled for additional listing on the KRX on July 29, so Q3 is the first clean quarter.
Second, SK Square's own Q3 disclosure of net asset value. The end-March figure of ₩123tn is stale by a factor of roughly two on the hynix line. If the updated NAV lands near the ₩250tn implied by the market and the share price does not move, the discount is the whole story and it isn't closing on its own. If management pairs that disclosure with a capital return sized against the gap rather than against the year's dividend income, that would be new information.
Neither number tells you what the stock does. Both tell you whether the July dilution was a footnote or a signal about who this company works for.
kstock reads DART every morning and writes up what moved — the contract, the buyback, the number that does not add up. The daily post and a Saturday roundup, by email.