In FY2021, HD Hyundai Heavy Industries Co., Ltd. (KRX:329180) lost ₩800.3bn at the operating line on ₩8.31tn of revenue. An operating margin of negative 9.63%, and a gross margin of negative 2.52%, meaning the ships cost more to build than customers paid for them.
In the second quarter of 2026 the same company earned ₩1,039.9bn on ₩6.33tn of revenue. An operating margin of 16.42% and a gross margin of 21.42%.
Nothing about the yards changed. What changed is which contracts are being delivered.
Shipbuilding does not work like other manufacturing, and the difference matters enormously for how you read an income statement.
A shipowner orders a vessel two to three years before delivery, at a price fixed on the day the contract is signed. The yard then builds it, recognising revenue as construction progresses. Steel is bought along the way at prices nobody knew when the contract was priced, and labour costs whatever it costs by then.
So the margin on any given ship is set by two numbers separated by years: the price agreed at order, and the cost incurred at build. A yard has no ability to reprice a ship once the contract exists.
That explains FY2021 precisely. The vessels delivered that year were ordered in 2018 and 2019, at the bottom of a decade-long pricing depression when Korean yards were taking work at any price to keep docks full. They were then built through 2020 and 2021, when steel plate prices spiked. Contracts signed cheap, built expensive.
And it explains now. The vessels being delivered in 2025 and 2026 were ordered in 2023 and 2024, when newbuild prices had recovered sharply and Korean yards had regained pricing power in LNG carriers and large containerships. Contracts signed expensive, built into a calmer input cost environment.
The 16.42% is not news about how well the company is being run this quarter. It is the arithmetic consequence of decisions taken three years ago, working through.
The same is true in the other direction, which is the useful part. If you want to know what HD Hyundai Heavy will earn in 2028 and 2029, the income statement is nearly useless. What you need is the price of orders being taken in 2026.
That is knowable, partially, from the disclosures. The company filed five single sales contract announcements in the two months to late August, on July 1, July 9, July 14 as an amendment, August 10 and August 24. Korean rules require disclosure of contracts above a share of revenue, so the flow of those filings is a running indicator of order intake.
What the filings do not give you directly is price per vessel against cost, which is where the margin lives.
For completeness, the operating margin by fiscal year: 2.37% in FY2019, 0.39% in FY2020, negative 9.63% in FY2021, negative 3.20% in FY2022, 1.49% in FY2023, 4.87% in FY2024, 11.59% in FY2025.
Six years from the trough to the current level, and the improvement has been continuous since 2021 without a single reversal. That consistency is itself evidence the mechanism is the order book working through rather than anything cyclical about costs.
Revenue has grown alongside it, from ₩5.46tn in FY2019 to ₩17.58tn in FY2025, and first-half 2026 revenue of ₩12.25tn was 54.2% above a year earlier. Annualised, this year runs near ₩24.5tn. The yard is fuller than it has ever been and the ships in it are the profitable ones.
One transaction in the recent past deserves noting, because it cost existing holders a great deal.
Share capital went from ₩443,865,580,000 to ₩524,806,125,000 in the fourth quarter of 2025. At the ₩5,000 par value that is 16,188,109 new shares against roughly 88.8m existing, dilution of 18.2%.
Total equity over the same quarter rose from ₩6,465.0bn to ₩9,341.6bn, an increase of ₩2,876.6bn, against fourth-quarter net income of ₩489.6bn. So roughly ₩2.4tn came from the issuance.
Divide one by the other and the implied issue price is somewhere near ₩147,500 a share. The August 27 close was ₩470,000. The equity figure includes other movements so treat the derivation as approximate, but the direction is not in doubt: the company sold a fifth of itself at a small fraction of what the shares fetch today, months before the margin story became obvious.
That is the same pattern visible at Samsung SDI, which raised roughly ₩2tn in 2025 at the bottom of the battery cycle, and at Doosan Enerbility, whose share capital rose more than fivefold across its rescue years. Korean industrials raise equity when they need it, not when it is cheap for shareholders.
Two things.
The order price cycle turns. Chinese yards have been expanding capacity aggressively, and newbuild prices are set globally rather than by Korean yards alone. If contracts signed in 2026 are priced below those signed in 2023, today's 16% margin is the peak and 2029 will look like FY2024.
And the yard has to run. The company shut every plant for three days at the end of July after three serious accident disclosures in eighteen days, and two government work stoppage orders including one covering the entire panel factory assembly line at Gunsan remained in force after production resumed. A margin built on a full order book requires the docks to keep moving.
At the August 27 close the company is capitalised at ₩49.33tn against ₩10,542.9bn of equity, so 4.68 times book, and about 15.5 times annualised first-half earnings.
The single contract disclosures, and specifically whether the pace continues into the fourth quarter. Order intake now is 2029's revenue, and the flow of filings is the only public running count.
The second marker is the FY2026 annual report's order backlog figure, which Korean shipbuilders disclose in both value and vessel count. Backlog value rising while the count falls means prices are still going up. The reverse would mean the yard is buying volume again, and that is how the last cycle began.
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.