Wonik IPS Co., Ltd. (KOSDAQ:240810) earned ₩89.9bn of net income last quarter, up 225% from a year earlier. It also earned ₩18.4bn of operating income, down 49.6% from a year earlier. Both are true, and only one of them is about making semiconductor equipment.
The bridge between them is a ₩80.9bn finance income line. Korean press covering the August 6 disclosure attributed roughly ₩80bn of it to valuation gains on investment assets. To put that in scale: the company's entire finance income for all of fiscal 2025 was ₩34.6bn. One quarter produced more than double a full year.
This is not an accounting scandal. Mark-to-market gains on investment holdings are perfectly ordinary under IFRS, and a company with ₩1.07tn of equity and a long history of holding stakes in affiliates and suppliers will book them. But it does mean that anyone reading the headline number is reading the wrong number.
Start with the operating story, because that is the company.
Revenue in the second quarter was ₩216.5bn, against ₩242.2bn a year earlier. Down 10.6%. Two quarters before that, in the fourth quarter of 2025, revenue was ₩275.0bn. So the top line has fallen 21% from its recent high. Operating income went from ₩36.5bn to ₩18.4bn, and the operating margin from 15.1% to 8.5%. On a sequential basis things improved, revenue up 31.3% and operating income up 71.3% from a weak first quarter, but the first quarter was the weak one and the second quarter still sits well below where the business was a year ago.
Now the reported story. Net income of ₩89.9bn. First half net income of ₩112.0bn, up 388% year over year. Those are the figures that headline a screen, and they are almost entirely an artifact of the non-operating line. Pretax income for the quarter was ₩101.2bn against ₩18.4bn of operating income, and the difference is finance income.
Look at the first half combined and it becomes plainer. Revenue for the six months was ₩381.4bn against ₩366.4bn a year earlier, up about 4%. Operating income for the six months was ₩29.1bn against ₩29.1bn a year earlier. Flat, to within a rounding error. That is the underlying business: unchanged. Net income over the same comparison quintupled.
At ₩111,100 a share and 49,083,901 shares outstanding, the market cap is ₩5.45tn (~$3.9bn at roughly 1,400 won to the dollar, an approximate rate). Trailing twelve-month net income through June is ₩173.1bn, which puts the stock at about 31x. That looks unremarkable for a semiconductor equipment name in a hot cycle.
Swap the denominator. Trailing twelve-month operating income is ₩73.9bn. Against a ₩5.45tn market cap that is roughly 74x. Trailing revenue is ₩924.8bn, so the stock is close to 5.9x sales for a business whose operating margin was 8.5% last quarter. Book value is ₩1,073.6bn, so about 5.1x book.
Neither multiple is right on its own. The 31x understates because it capitalizes a one-time gain as if it recurs. The 74x overstates because it implicitly says the current operating margin is the steady state, which nobody who follows this company believes. What both numbers do agree on is that the market is not paying for trailing results. It is paying for 2027.
That is a defensible thing for the market to do with an equipment supplier. It is worth being explicit that it is what is happening, because the 31x headline invites the opposite conclusion.
There is a structural reason the operating margin swings this violently, and it is not gross margin.
Gross margin has been remarkably stable. It was 39.8% in fiscal 2017, 40.0% in the loss-making year of 2023, 42.1% in 2025, and 45.7% in the second quarter of 2026. Through a decade in which revenue ranged from ₩630.9bn to ₩1,232.3bn, the gross margin moved less than six points.
The operating margin over the same period went from 19.4% in 2017 to negative 2.6% in 2023 to 8.1% in 2025. The whole swing sits in the operating cost base between gross profit and operating profit. That gap was about ₩129bn in fiscal 2017. It has run between roughly ₩282bn and ₩315bn in every year since 2021, essentially flat regardless of what revenue did. Note that the SG&A line the company reports does not bridge gross profit to operating income on its own, so a large block of operating cost, development spending most obviously, sits outside it. The filings do not break it out in a way I can attribute precisely.
The arithmetic that follows is simple and it explains a lot. Against a roughly ₩300bn fixed operating cost base and a roughly 41% gross margin, Wonik IPS needs somewhere near ₩730bn of annual revenue just to break even at the operating line. Fiscal 2023 came in at ₩690bn and produced an ₩18.1bn operating loss. Fiscal 2024 came in at ₩748bn and produced ₩10.6bn of operating profit, barely positive. Fiscal 2025 came in at ₩910bn and produced ₩73.8bn.
Annualize the first half of 2026 and you get about ₩763bn, which lands just above that breakeven line. That is exactly what the ₩29.1bn of first-half operating income says.
The fair objection is that I am judging a project business on a calendar quarter. Equipment revenue is recognized on shipment and acceptance, and those events cluster. The second quarter can be light because tools were built but not yet delivered, and there is balance sheet evidence pointing that way: inventories rose to ₩458.1bn at the end of June from ₩272.7bn at the end of December, and current liabilities more than doubled to ₩414.8bn from ₩188.1bn. Local coverage cites brokerage estimates of a sharply larger third quarter as overseas shipments get recognized. Those are forecasts, not results, and I am not going to treat them as facts.
A second objection is that the investment gains are real money even if they are not operating money. They are. The portfolio is an asset the shareholders own, and a company that turns ₩80bn of unrealized appreciation into equity has genuinely become richer on paper. My point is narrower: unrealized valuation gains do not repeat on schedule, they can reverse, and capitalizing them at 31x is a mistake in either direction.
The third risk cuts against the bulls and the bears both. Wonik IPS sells deposition and etch tools into Korean memory fabs, with growing exposure to China. That customer base is concentrated, and the timing of any single fab's capex decision moves a full year of this company's revenue. Neither the fixed cost analysis above nor the multiple debate survives a customer pushing a fab out two quarters.
The third-quarter report, and specifically the operating income line rather than the net income line. If revenue steps up materially from ₩216.5bn and operating income moves well clear of the ₩18-27bn range it has run in for four quarters, the fixed cost base is being cleared and the operating leverage works in the right direction. If revenue steps up and operating income does not follow, something in the cost base has moved permanently.
Watch the finance income line in the same report too. A quarter that brings ₩80bn of valuation gain can bring a valuation loss the next time, and the market has not yet had to price that scenario for this stock.
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.