In the first half of 2026, Samsung Heavy Industries Co., Ltd. (KRX:010140) generated ₩2,306,934,409,142 of operating cash flow. Revenue for the same six months was ₩6,133,0 billion — so cash conversion was 37.6% of sales.

Purchases of property, plant and equipment over the same period: ₩96,960,957,139.

Ninety-seven billion won. Against a yard carrying ₩5,130.9bn of net plant on its books, in a company preparing to build some of the most complex offshore structures ever fabricated.

Where The ₩2.3tn Came From

Not from profit. Net income for the half was ₩322.9bn, so operating cash flow was seven times reported earnings.

The balance sheet says where the rest came from. Total liabilities rose from ₩10,853.8bn at the end of December to ₩13,249.0bn at the end of June, an increase of ₩2,395.2bn. Of that, ₩1,612.4bn landed in current liabilities.

At a shipyard, the dominant current liability is not bank debt. It is contract liabilities — money customers have paid against vessels and offshore units not yet delivered. Samsung Heavy won about $10bn of new orders through July against a full-year target of $13.9bn, and each of those contracts brought a down payment.

So the ₩2.3tn is substantially the customers' money, held against work Samsung Heavy has committed to perform. It is not free cash flow in any economic sense. It is a float, and it reverses as the ships and platforms get built.

That is not a criticism — it is how the industry works, and it is far healthier than the alternative. Operating cash flow was negative ₩1,693.0bn in FY2022 and negative ₩516.5bn in FY2023, when the yard was building through a backlog it had already been paid for. Collecting advances is the good half of the cycle.

The Capital Spending Number Is The Odd One

What is genuinely unusual is what the company is not doing with it.

Capital spending on property, plant and equipment: ₩216.3bn in FY2025, ₩173.4bn in FY2024, ₩153.4bn in FY2023, and ₩97.0bn across the first half of 2026. Against a net plant balance that has barely moved — ₩5,208.7bn at the end of FY2023, ₩5,115.9bn at FY2024, ₩5,107.1bn at FY2025, ₩5,130.9bn at 30 June 2026.

Annualise the first half and capex runs at under 4% of net plant. For a heavy fabrication yard with dry docks, cranes, workshops and a large steel-handling infrastructure, that is very likely below the depreciation charge — which would mean the asset base is shrinking in real terms even as the book value holds flat. I have not read the depreciation line; it is in the half-year cash flow statement as an add-back, and anyone assessing this should put the two numbers side by side.

The contrast with the order book makes it sharper. Samsung Heavy is targeting $8.2bn of offshore orders this year, including a $2.88bn floating LNG unit for Delfin Midstream, and pursuing two more awards of similar scale. Those are the most capacity-intensive products a yard can take. Building them in a facility that has received minimal investment for four years is either admirable discipline or a decision that gets tested in 2028.

Where The Money Actually Went

Three places, and none of them is the yard.

Financing activities consumed ₩1,009.5bn across the half, following outflows of ₩1,266.5bn in FY2025 and ₩599.2bn in FY2024. That is roughly ₩2.9tn of debt repaid in two and a half years, and it is the single best thing this company has done with its cash. Total liabilities have risen since December, but the increase is customer advances, not borrowing.

Investing activities consumed ₩1,269.4bn, of which only ₩97.0bn was plant. So about ₩1.17tn went into financial assets — deposits, securities or similar — which the summary statements do not itemise. That is money parked rather than deployed.

And the cash balance itself barely moved: ₩772.6bn at the end of December, ₩807.4bn at the end of June.

So a company that generated ₩2.3tn in six months ended the period with essentially the same cash it started with, having repaid ₩1tn of debt and moved ₩1.17tn into financial instruments.

Inventory Is Almost Nothing

One more line worth flagging, because it is unusual enough to look like an error.

Inventories were ₩497.8bn at 30 June, against first-half cost of revenue of ₩5,262.8bn — roughly seventeen days. Korean yards typically carry more. They also fell sharply once before: from ₩1,729.2bn at the end of FY2023 to ₩452.5bn at the end of FY2024, a drop of ₩1,276.7bn in a year when revenue rose.

That looks like a reclassification between inventory and contract assets rather than a physical destocking, which would be consistent with a change in how work in progress is presented. The half-year report notes would confirm it. Either way, anyone computing inventory turns for this company across the FY2023-24 boundary will get a meaningless number.

The Case That This Is Exactly Right

The defence of the capital discipline is strong and probably correct.

Samsung Heavy lost more than ₩7tn between 2015 and 2023 and still carries an accumulated deficit of ₩1,276.9bn. A company in that position should repay debt and rebuild equity before it builds anything else. Korean yards also have physical capacity that has been under-utilised since the last downcycle — the docks exist; they were simply empty for years. Filling existing capacity requires labour and working capital, not new cranes.

There is also a strategic argument. Adding shipbuilding capacity at the top of a cycle is how the industry destroyed itself twice. A yard that takes the cash and pays down debt instead is the one still standing when the cycle turns.

What Would Settle It

Depreciation against capital expenditure in the half-year statements. If depreciation is running near ₩300bn a year against ₩200bn of capex, the yard is consuming its asset base slowly and can continue for years. If depreciation is far higher, the under-investment has a deadline.

Second, whether capital spending rises once the offshore awards convert. An FLNG unit occupies dock space and fabrication capacity for years. If Samsung Heavy wins the second Delfin unit and the Canadian project without increasing capex, either the yard has more spare capacity than outsiders think, or something is being deferred.

Third, the ₩1.17tn now sitting in financial assets. Money parked in deposits at a company with an accumulated deficit and no dividend is money awaiting a decision. What that decision turns out to be — debt repayment, capacity, or eventually shareholders — is the most consequential capital allocation call this management will make.