Korea Electric Power Corporation (KRX:015760, NYSE:KEP) lost ₩4,716.1bn in 2023. It then earned ₩3,622.0bn in 2024 and ₩8,666.7bn in 2025, with operating profit of ₩8,364.7bn and ₩13,490.6bn in those two years. Roughly ₩21.9tn of operating profit across twenty-four months, at a company that had spent the previous three years borrowing to cover a gap between what electricity cost and what it was allowed to charge.
Here is what happened to the debt.
Total liabilities were ₩202,450.2bn at the end of 2023. ₩205,445.0bn at the end of 2024. ₩205,604.5bn at the end of 2025.
They went up.
This isn't a story about waste, and it's worth being clear about that before drawing conclusions. KEPCO's cash flow statement shows where the recovery went.
Operating cash flow was ₩1,522.2bn in 2023, ₩15,876.1bn in 2024 and ₩20,880.2bn in 2025 — an enormous swing, and exactly what you'd expect from a utility whose tariff finally caught up with its costs.
Capital expenditure over the same three years was ₩13,908.4bn, ₩14,216.0bn and ₩15,834.1bn.
So 2023 produced free cash flow of negative ₩12,386.2bn, which is the year the balance sheet took the damage. 2024 produced positive ₩1,660.1bn. 2025 produced positive ₩5,046.0bn.
Two positive years, ₩6.7tn of cumulative free cash flow, against a liability stack of ₩205tn. That is not enough to move the number, and it wasn't going to be. Add roughly ₩1tn of dividends for 2025 and the free cash flow barely covers the payout.
Property, plant and equipment rose from ₩187,751.5bn at the end of December to ₩192,398.3bn at the end of June — ₩4,646.7bn in six months. The money is going into transmission, distribution and generation assets. It is a real grid being built, not a hole.
The first half of 2026 is where the trend reversed rather than merely stalled.
Total liabilities rose from ₩205,604.5bn to ₩210,716.2bn. Up ₩5,111.6bn in six months, after two years of going roughly nowhere.
The composition is the part to look at. Current liabilities rose ₩5,882.3bn to ₩72,989.6bn, while non-current liabilities fell ₩770.6bn to ₩137,726.6bn. Inside that, other current financial liabilities — short-term borrowings and the current portion of long-term debt — went from ₩45,939.1bn to ₩50,707.8bn, up ₩4,768.7bn, while the non-current equivalent fell from ₩83,995.7bn to ₩82,735.3bn.
So KEPCO added debt and shortened it at the same time. Roughly ₩50.7tn of financial liabilities now fall due within twelve months, against cash of ₩2,622.1bn and other current financial assets of ₩4,567.2bn.
For a company with an implicit sovereign relationship — the government and the Korea Development Bank together hold a controlling stake — this is a funding pattern rather than a solvency issue. KEPCO bonds are among the most liquid Korean credit instruments and the market for them does not close. But it does mean a growing share of a ₩133tn financing stack reprices at prevailing rates within the year, and it means the maturity cushion that a utility normally maintains has thinned.
Capital expenditure in the first half was ₩9,571.3bn. Annualise it and you get roughly ₩19.1tn, against ₩15,834.1bn for all of 2025 — an increase of about 21%.
Operating cash flow in the half was ₩8,545.6bn. So free cash flow was negative ₩1,025.7bn, the first negative half since the crisis years, and it came in a period when the company was still comfortably profitable.
That is the whole tension in this balance sheet. KEPCO's operating profit fell 16.6% year on year while its capital programme grew 21%. The grid build — transmission for new industrial demand, connections for renewables, replacement of ageing distribution — is not discretionary and is not slowing. The tariff that pays for it is decided annually by the government, and was not raised in a way that covered the first half's fuel costs, let alone the incremental capex.
KEPCO has said it aims to bring its bond issuance ratio below twice its debt-to-capital ratio by the end of 2027, using a combination of operating profit growth, tariff increases and disciplined capital spending. The first-half numbers move the wrong way on two of those three.
At ₩34,800 the market capitalisation is ₩22,340.3bn against ₩50,675.8bn of equity attributable to the parent. That's 0.44x, and it reads as a company priced for disaster.
Set the debt next to it. Financial liabilities of roughly ₩133.4tn, cash and financial assets of roughly ₩12.2tn, so net financial debt near ₩121tn — about 5.4 times the entire market value of the equity. On top of that sit ₩29,149.8bn of non-current provisions, principally nuclear decommissioning and spent fuel obligations that accrue for decades.
At that leverage, the equity is a thin sliver on top of a very large asset base, and small changes in the value of the assets or the profitability of operating them move it a great deal. A 0.44x book multiple on a balance sheet levered five times is not obviously the same bargain as a 0.44x multiple on an unlevered one. It's an option, and it's priced like one.
There's also ₩12,895.2bn of deferred tax assets in there — about a quarter of parent equity — carried on the expectation of future taxable profits. And ₩12,708.6bn sits under a line the summary balance sheet labels simply as other capital, unchanged for the period and unexplained in the figures available here. Between them those two items are half the book value a buyer is discounting.
Utilities are supposed to be levered. A regulated network with predictable demand and an implicit state guarantee can carry debt that would destroy an industrial company, and KEPCO's assets — the physical grid of a developed economy — are about as durable as collateral gets.
The capital programme is also the strongest argument for owning the equity rather than against it. Korea is building data centres and semiconductor fabs that need enormous incremental power, and every megawatt of that demand has to cross KEPCO's wires. A utility spending ₩19tn a year to serve demand growth is a utility with a growing rate base, which is the mechanism by which regulated networks compound. If the tariff eventually reflects the cost of that investment, the leverage that looks alarming becomes the reason the equity multiplies.
The qualifier is the same as everywhere else with this company: whether the tariff reflects the investment is a political decision, not a regulatory formula.
Other current financial liabilities in the third-quarter report against ₩50,707.8bn. If that keeps climbing while the long-term line falls, KEPCO is funding a growing capital programme on short paper going into a tariff decision it doesn't control — the least comfortable combination available.
Second, capital expenditure against ₩9,571.3bn for the half. Whether the full year lands near ₩19tn or gets pulled back toward 2025's ₩15.8tn tells you how much discipline management actually has over the build.
Third, total liabilities against ₩210,716.2bn at the year end. Two profitable years didn't reduce them. A third profitable year that also doesn't reduce them would mean the recovery has bought stability and nothing more.
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.