Samsung Heavy Industries Co., Ltd. (KRX:010140) earned ₩535.8bn in FY2025 and ₩322.9bn across the first half of 2026. Second-quarter operating margin reached 10.1%. The company paid a performance bonus this year for the first time in twelve years.
Retained earnings at 30 June 2026: negative ₩1,276,928,668,233.
Two good years have not undone what came before them, and the balance sheet is the only place the full story is still written down.
The loss record, year by year. FY2015: negative ₩1,212.1bn. FY2016: negative ₩138.8bn. FY2017: negative ₩340.7bn. FY2018: negative ₩388.2bn. FY2019: negative ₩1,315.4bn. FY2020: negative ₩1,492.7bn. FY2021: negative ₩1,452.1bn. FY2022: negative ₩627.4bn. FY2023: negative ₩155.6bn.
Nine consecutive years. More than ₩7.1tn in total.
Operating income was negative for eight of those years, from FY2015 through FY2022, turning positive only in FY2023 at ₩233.3bn. So this was not a financing problem or an accounting quirk. The yard sold ships and offshore units for less than they cost to build, repeatedly, for the better part of a decade.
The cause is well documented across the Korean industry: offshore production and drilling contracts signed in the early 2010s at prices that assumed engineering scope the yards had underestimated, followed by an oil price collapse that led customers to cancel or defer. Under IFRS, an expected loss on a contract is recognised in full the moment it becomes probable, which is why the losses arrived in large lumps rather than gradually.
Follow the share capital line and you can watch the survival sequence.
It was ₩1,155.0bn at the end of FY2015. It rose to ₩1,950.6bn by the end of FY2016 and to ₩3,150.6bn by FY2018 — two capital increases, in which existing shareholders were diluted to keep the company funded through the losses.
Then, in 2021, it fell to ₩880.1bn, where it remains. That is a capital reduction: the accumulated deficit was partly absorbed by writing down paid-in capital, which is the standard Korean mechanism for cleaning up a balance sheet without a formal restructuring.
Anyone who held Samsung Heavy shares from 2015 to 2022 was diluted twice and then had their capital written down. The stock's per-share history across that period cannot be read straight for the same reason.
The deficit peaked at negative ₩2,168.7bn at the end of FY2023. It was negative ₩2,136.0bn at FY2024, negative ₩1,602.8bn at FY2025, and negative ₩1,276.9bn at 30 June 2026.
So roughly ₩892bn has been earned back — about two fifths of the peak deficit — in two and a half years.
At the first half's run rate of ₩322.9bn per six months, the remaining ₩1,276.9bn takes about two more years to clear. That would put retained earnings back above zero somewhere in 2028, thirteen years after the losses began.
Total equity is ₩4,710.4bn, which is comfortable, and total liabilities of ₩13,249.0bn give a ratio of 2.81 — higher than either Korean peer but not alarming for a yard funded largely by customer advances. Operating cash flow of ₩2,306.9bn across the first half is the strongest evidence that the business itself is now sound.
Here is the practical consequence, and it is the reason to care about a historical accounting balance.
Korean companies may only distribute out of distributable profit, which is determined on the separate rather than consolidated financial statements and is reduced by accumulated losses. I have not read Samsung Heavy's separate accounts, so I cannot state the legal position precisely. But a consolidated deficit of ₩1.28tn is a strong indication that the separate position is also constrained, and the payout record is consistent with it: dividends paid were zero in FY2023 and FY2024, and ₩1,083,416,198 in FY2025 — about ₩1bn, against ₩535.8bn of net income.
That is a rounding error, and it is the single clearest difference between Samsung Heavy and its two Korean peers, both of which now pay substantial dividends. An investor buying HD Korea Shipbuilding or Hanwha Ocean gets a cash return. An investor buying Samsung Heavy gets the earnings retained to fill a hole dug a decade ago.
Which is not necessarily a bad trade. It does mean the total return has to come entirely from the share price.
The counter-argument is reasonable and probably right.
An accumulated deficit is a record of the past, not a claim on the future. Samsung Heavy's cash generation, order book and margins today have nothing to do with contracts signed in 2012. The company holds $35.9bn of backlog, won about $10bn of new orders through July against a $13.9bn full-year target, and is earning 10% operating margins on work it priced recently.
The deficit also has one genuine benefit: accumulated tax losses. FY2024 carried an income tax credit of ₩369.4bn as deferred tax assets were recognised, and while FY2025 returned to a normal charge of ₩114.7bn, carried-forward losses reduce cash tax for years. A company that lost ₩7tn does not pay much tax on its first few trillion of recovery.
And the balance sheet risk is genuinely gone. Equity has grown from ₩3,409.2bn at the end of FY2023 to ₩4,710.4bn, and the yard is collecting advances faster than it spends them.
The first meaningful dividend. Whenever Samsung Heavy declares one, it will be the clearest possible signal that the accumulated deficit has cleared or that the board has found a way around it — most likely a capital reserve reduction, the mechanism several Korean companies have used this year to convert paid-in surplus into distributable reserves. Watch the annual general meeting agenda for exactly that resolution.
Second, the pace of the recovery. Two more halves at ₩322.9bn would put the deficit near ₩630bn by mid-2027. Any slowdown pushes the dividend further out, and at this company the dividend is the whole shareholder-return question.
Third, and most important: whether the offshore contracts now being signed behave differently from the ones signed in 2012. The deficit exists because the last offshore cycle was mispriced. The company is entering another one. Gross margin on offshore revenue in 2027 and 2028 is the number that tells you whether anything was learned.
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.