HD Hyundai Electric Co., Ltd. (KRX:267260) has one of the strangest balance sheets in Korean heavy industry, and it is strange in a way that is entirely good until it isn't.
At June 30 the company carried ₩3,233.5bn of total liabilities. Of that, ₩3,077.7bn falls due within twelve months and ₩155.8bn does not. Ninety-five percent of what this company owes is current.
For a manufacturer of large power transformers, equipment that takes years to build and decades to depreciate, that is not the normal shape. Heavy equipment makers finance long-lived assets with long-dated debt. This one barely has any.
Track them from FY2017, the year the company was spun out of Hyundai Heavy Industries.
Non-current liabilities: ₩465.6bn in FY2017, ₩432.5bn in FY2019, ₩316.3bn in FY2021, ₩355.7bn in FY2023, ₩286.9bn in FY2024, ₩188.2bn in FY2025, ₩155.8bn at June 30, 2026. Down by two-thirds.
Current liabilities: ₩588.8bn in FY2017, ₩1,252.5bn in FY2021, ₩1,497.4bn in FY2023, ₩2,001.1bn in FY2024, ₩2,548.7bn in FY2025, ₩3,077.7bn at June 30. Up more than fivefold.
Long-term borrowing shrank while short-term obligations exploded, at a company that was simultaneously becoming enormously profitable. The obvious explanation is the right one: most of that current liability is not debt. It is money customers have already paid for equipment not yet delivered.
The summary financials do not itemise contract liabilities, and the notes to the half-year report are where the split sits. But nothing else produces this pattern. A company borrowing heavily would show the money in non-current liabilities. A company being pre-paid shows it in current ones.
The cash flow statement confirms it and dates the change precisely.
In FY2022, operating cash flow was negative ₩124.1bn against ₩133.0bn of operating income. In FY2023, negative ₩22.4bn against ₩315.2bn. The company was earning accounting profits and consuming cash, which is normal for a manufacturer building inventory into a lengthening order book.
Then FY2024: operating cash flow of ₩1,033.7bn against operating income of ₩669.0bn. Conversion above 150%. FY2025: ₩959.6bn against ₩995.3bn, roughly one for one. The first half of 2026 produced ₩480.3bn against ₩545.3bn of operating income.
Something changed in 2024, and it was not the company's collection department. It was that the queue for transformers got long enough that buyers would pay in advance to hold a place in it.
That is a supplier's market in its purest form. The customer takes the financing risk, funds the working capital, and waits.
The consequence for shareholders has been direct.
Dividends paid were ₩18.0bn in FY2023, ₩75.6bn in FY2024 and ₩221.3bn in FY2025. In the first half of 2026 alone the figure was ₩233.9bn, exceeding all of last year. That is a twelvefold increase across two fiscal years, and it is still accelerating.
Meanwhile share capital has been unchanged at ₩180,235,675,000 since FY2019. No new shares have been issued in seven years. Retained earnings went from negative ₩499.8bn at the end of FY2021 to positive ₩1,503.4bn at June 30, a swing of ₩2.00tn.
Worth noting that ₩233.9bn against a market capitalisation of ₩29.59tn is a dividend yield under 1%. The payout has grown extraordinarily and the share price has grown faster.
Doosan Enerbility offers the contrast. Both companies nearly failed. Both issued large amounts of equity to survive. Doosan's share capital rose from ₩596.8bn in FY2015 to ₩3,267.3bn by FY2023, a fivefold dilution. HD Hyundai Electric's rose from ₩51.0bn in FY2017 to ₩180.2bn in FY2019 and then stopped.
Today Doosan reports negative operating cash flow in every interim period and recently signed a ₩928bn contract in Oman with no advance payment at all. HD Hyundai Electric reports positive operating cash in every period shown and appears to be substantially pre-funded by its customers.
The difference is not management quality. It is that one of them makes something that is currently scarce and the other builds power plants in competitive tenders. Bargaining position determines who finances whom, and it shows up on the balance sheet before it shows up anywhere else.
Which is the risk, and it is worth being precise about the mechanism.
Customer advances are not permanent capital. They are a liability that gets extinguished by delivering equipment, and they get replenished only if new orders keep arriving with the same payment terms. If the queue shortens, two things happen at once. New advances slow, because customers no longer need to pay early to secure a slot. And existing advances continue to unwind as the backlog is delivered.
The result would be a period where operating cash flow falls well below operating income, exactly reversing what happened in FY2024. Revenue and profit could look fine while cash drains. That is what FY2022 and FY2023 looked like at this company, when operating cash was negative and operating income was positive.
None of that is a forecast. It is the arithmetic of what happens when advance funding stops growing, and shareholders paying thirteen times book value at the August 27 close of ₩821,000 should know which line moves first.
The current liability balance, quarter by quarter. It went from ₩2,548.7bn at December 31 to ₩3,077.7bn at June 30, so advances are still building. Growth stalling would be the earliest available signal that the shortage is easing, and it would appear months before pricing or margins moved.
The second thing is the relationship between operating cash flow and operating income. When cash conversion falls below one for two consecutive periods, the customer financing has stopped growing and the company is delivering against advances it took in earlier years. That is the point at which a very good balance sheet starts becoming an ordinary one.
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.