Hyundai Engineering & Construction Co., Ltd. (KRX:000720) generated ₩1,009.4bn of operating cash flow in FY2021. Since then:
FY2022: negative ₩143.5bn. FY2023: negative ₩714.7bn. FY2024: negative ₩118.8bn. FY2025: negative ₩748.3bn. First half of 2026: negative ₩1,677.1bn.
Four full years and one half, all negative. The first six months of 2026 consumed more than twice what all of FY2025 did.
Over the same period the company reported cumulative net income of roughly ₩918bn, once FY2024's ₩766.2bn loss is netted against the other years. So the accounting profit and the cash have gone in opposite directions for half a decade.
Trade receivables tell most of the story. They were ₩2,015.8bn at the end of FY2022, ₩3,378.7bn at FY2023, ₩5,319.2bn at FY2024, ₩6,842.3bn at FY2025 and ₩7,236.2bn at 30 June 2026.
Up 259% in three and a half years, while revenue rose 46%.
As a share of revenue, receivables went from about 9.5% in FY2022 to 22.0% in FY2025, and annualising the first half of 2026 takes them near 28% — roughly a hundred days of sales outstanding at a company whose revenue is falling.
For a Korean builder there are two categories inside that. Domestic apartment work generates receivables from reconstruction associations and developers, collected as buildings complete and buyers pay instalments. Overseas plant and infrastructure work generates unbilled construction receivables — 미청구공사, or contract assets — where work has been performed but the client has not yet certified it for payment.
The second category is the dangerous one. An unbilled receivable is the contractor's own assessment that it has earned money the customer has not agreed it owes. When those balances build up, one of two things is happening: either billing is simply lagging, which resolves, or there is a scope disagreement, which resolves in a negotiation and often at a discount. The difference does not show up until it does.
The summary statements do not split the two. The half-year report notes do, and that is the single most useful page in the document for anyone holding this stock.
There is a rhythm worth understanding before reading any interim figure from this company.
In FY2025, cumulative operating cash flow was negative ₩1,209.2bn at March, negative ₩1,889.2bn at June and negative ₩2,395.7bn at September. The full year came in at negative ₩748.3bn, which means the fourth quarter brought in about ₩1,647bn. FY2024 shows the same shape: negative ₩707.4bn through nine months, negative ₩118.8bn for the year.
Korean construction collects heavily in the fourth quarter, as projects hit completion milestones and annual settlements are made. So a first-half burn of ₩1,677.1bn is not the annual figure and should not be annualised.
It is still the largest first-half burn on file, and it is happening while revenue falls 13.5%. Collecting less on a smaller revenue base is the wrong combination.
The gap has been financed. Financing activities brought in ₩366.0bn in FY2023, ₩733.7bn in FY2024, ₩434.2bn in FY2025 and ₩403.6bn in the first half of 2026.
Cash and equivalents nonetheless fell from ₩5,130.4bn at the end of FY2024 to ₩4,812.7bn at FY2025 and ₩3,488.9bn at 30 June 2026 — down ₩1,641.5bn in eighteen months.
And throughout, the company kept paying dividends: ₩94.8bn in FY2023, ₩94.9bn in FY2024, ₩68.9bn in FY2025, and ₩116.8bn already in the first half of 2026.
Paying a dividend out of borrowed money while operating cash flow is negative is a defensible choice for a company with a strong balance sheet and a temporary working capital swing. It is a less defensible one in year five.
The balance sheet is genuinely sound, to be fair. Total equity was ₩11,217.8bn at 30 June against total liabilities of ₩17,411.3bn, a ratio of 1.55, and the company holds ₩3.49tn of cash. This is not a solvency discussion. It is a discussion about whether reported profit is being converted into anything.
The strongest defence is structural, and it is real.
Construction is a working-capital business, and a contractor growing its order book must fund the gap between doing work and being paid for it. Hyundai E&C won ₩22,823.0bn of new orders in the first half of 2026, up 36.4%, and now carries a backlog above ₩104tn. Mobilising on new projects consumes cash before any of it comes back.
Overseas plant work is worse in this respect than domestic housing, because payment milestones are further apart and certification takes longer. A deliberate shift toward large overseas projects — which is what management says it is doing — mechanically extends the cash cycle even if every contract is profitable.
There is also a Korean-specific factor worth checking rather than assuming. Domestic builders provide credit support to project financing vehicles for apartment developments, and those guarantees have been the sector's main source of stress since 2022. Hyundai E&C's exposure is disclosed in its filings; the summary statements do not carry it, and it belongs in any assessment of this company's cash position.
The split between billed and unbilled construction receivables in the half-year report. If the ₩7,236.2bn is mostly billed and simply awaiting payment, this is a timing story with a known end. If the unbilled portion has grown faster than the total, the company is recognising revenue on work its clients have not yet agreed to.
Second, fourth-quarter collections. The pattern says a large inflow arrives in Q4. If the full-year 2026 operating cash flow lands closer to breakeven than to FY2025's negative ₩748.3bn, the first-half burn was mobilisation on new orders. If the full year is worse than FY2025 despite lower revenue, the receivable build is not resolving.
Third, the dividend decision for FY2026. A board that raises the payout while operating cash flow is negative for a fifth year is making a statement about its confidence. A board that holds or trims it is making a different one, and after the ₩1,775.9bn operating loss booked in the fourth quarter of 2024, the second would be the more reassuring signal.
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.