A company with ₩328.5bn of quarterly revenue reported ₩19.24tn of operating income. That is SK Square Co., Ltd. (KRX:402340) in the second quarter of 2026, and the ratio is not a typo or a one-off gain. It's what happens when a Korean investment holding company puts its equity-accounted share of an affiliate's profit inside the operating line, and the affiliate is SK hynix Inc. (KRX:000660) in the middle of the strongest memory cycle anyone has traded.
The number is real accounting. It is also nearly worthless as a guide to what SK Square can pay out, refinance against, or reinvest, and the gap between those two statements is the most useful thing to understand about this stock.
Start with what SK Square actually operates. Revenue of ₩328.5bn in Q2, ₩300.3bn in Q1, and ₩1.41tn for the whole of FY2025, down from ₩2.37tn in FY2022 as the group sold or transferred out its platform businesses. Property, plant and equipment on the June 30 balance sheet totalled ₩106.1bn. Inventories were ₩10.9bn. This is a small services and investment operation attached to an enormous shareholding.
Now the operating line: ₩19,235.4bn for the quarter. A business with ₩328.5bn of revenue cannot produce that from trading. The only source available is the equity-accounted share of associates' profits, overwhelmingly the roughly 20% holding in SK hynix, which SK Square classifies within operating income because investment in affiliates is its stated business. Back-solve it and a 20% share of ₩19.24tn implies affiliate net profit somewhere near ₩96tn for the quarter. I have not checked that against hynix's own filing and equity-method adjustments would move it, so treat the figure as order of magnitude rather than a number to quote.
Net income for the quarter came in at ₩18.68tn after ₩992.7bn of tax. Add Q1's ₩8.37tn and the first half printed ₩27.05tn of profit. For scale, SK Square's entire market capitalisation on August 27 was ₩140.9tn, roughly $101bn at an approximate ₩1,400 to the dollar. Six months of reported earnings equalled about a fifth of the company's market value.
If you stopped there you would conclude the stock is absurdly cheap. The cash flow statement is where that conclusion falls apart.
Cash from operating activities was negative ₩49.1bn after the first quarter and positive ₩260.8bn by the end of the half. So the six-month total is ₩260.8bn. Against ₩27.05tn of net income, that's under one percent.
The reason is structural rather than suspicious. Equity-method income is a bookkeeping entry that increases the carrying value of the investment. It does not move money. Money moves when the affiliate declares a dividend, and dividends from an associate are a fraction of its earnings. SK hynix's stated policy for 2025 through 2027 returns more than half of cumulative free cash flow, but a large share of that has been running through buybacks rather than cash dividends, and buybacks return nothing at all to a holder who doesn't sell.
Watch the cash balance to see the effect. On December 31, 2025, SK Square held ₩1.31tn of cash and equivalents. On June 30, 2026, it held ₩1.32tn. Total equity over the same six months went from ₩27.98tn to ₩55.01tn. The company added ₩27tn of net worth and roughly ₩9bn of spendable cash.
The financing line tells you what was actually distributed. Cash used in financing was ₩555.0bn across the half, against ₩204.25bn of dividends paid over the same period. That dividend figure is a genuine change: the equivalent line was ₩1.8bn in the first half of 2025, and ₩1.8bn for all of FY2025. It also lines up with the ₩200bn cash dividend SK Square flagged in its 2026 shareholder return plan. The consolidated statement doesn't separate payments to non-controlling interests, so don't treat ₩204.25bn as a precise parent-only figure.
To fund all of that, SK Square leaned on other sources. Non-current liabilities rose from ₩1.55tn to ₩2.93tn across the half. Investing activities threw off ₩234.3bn, consistent with the disposal programme. Net cash still fell ₩60.0bn over the six months before currency effects. A company reporting ₩27tn of profit ran its cash balance flat.
None of this is an argument that the equity-method line is misleading in one direction. It swings. In FY2023 SK Square reported an operating loss of ₩2.33tn and a net loss of ₩1.31tn, on revenue of ₩2.01tn. That was the memory downcycle running through the same mechanism in reverse. In FY2022 operating income was ₩16.1bn on ₩2.37tn of revenue, essentially flat, because hynix earned very little that year.
So the sequence over four fiscal years reads ₩16bn, then negative ₩2.33tn, then ₩3.92tn, then ₩8.80tn, on revenue that fell every single year. Nothing about the operating business explains any of it. Anyone modelling SK Square by forecasting its revenue is modelling the wrong company, and anyone alarmed by the 2023 loss was reacting to a mark, not a cash event.
The practical translation for a US reader: think of SK Square less like a diversified conglomerate and more like a closed-end fund with one dominant position, a small operating stub, and an income statement that reports its holding's earnings rather than its own. The fund analogy breaks in one important way, which is that a closed-end fund marks to market and SK Square does not. Its hynix stake sits on the books at cost plus accumulated share of profits, which is why the company trades at roughly 2.6x reported book value while simultaneously trading around 44% below the market value of its hynix shares.
The reasonable pushback is that cash conversion is the wrong test for a holding company. Nobody buys SK Square for its dividend yield. They buy it as a claim on a hynix position, and that claim compounds through retained earnings at the affiliate whether or not a won ever reaches Seoul. On that view the equity-method line is the honest number and the cash flow statement is a distraction.
There's something to that, and the ₩5.89tn transfer of capital reserves into retained earnings that SK Square completed this year cuts in the same direction. That move doesn't create cash. What it does is expand distributable profit under Korean company law, which is the legal constraint on how much a company may return. A holding company that clears its distributable-profit ceiling is preparing to return more than its cash flow alone would suggest, funded by borrowing against an asset that keeps appreciating.
The counter to the counter is simple enough. Borrowing against an appreciating stake works until the stake stops appreciating, and this stake is a memory business. Non-current liabilities already nearly doubled in six months. A downcycle that compresses hynix's dividend while SK Square carries more debt is the scenario where the distinction between accounting profit and cash stops being an academic point.
The Q3 cash flow statement, filed in November, answers the question the earnings headline can't. The specific line is cash from operating activities. If it stays in the low hundreds of billions of won while reported net income runs in the tens of trillions, the gap is permanent and structural, and the payout ceiling is set by hynix's dividend rather than hynix's profit. If it steps up materially, hynix has shifted its return mix toward cash and SK Square's distribution capacity changes with it.
The second thing to track is the dividends-paid line against SK Square's own stated plan. ₩204.25bn in a single half against ₩1.8bn for all of last year is a real break in behaviour. Whether it repeats in Q4, and whether the company funds it from received dividends rather than from new borrowing, is what separates a policy change from a one-time gesture.
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.