Between FY2023 and FY2025, LG Energy Solution, Ltd. (KRX:373220) spent ₩33.2tn buying property, plant and equipment. ₩9,923.1bn, then ₩12,399.0bn, then ₩10,833.9bn.
Over the same three years its revenue fell from ₩33,745.5bn to ₩23,671.8bn, a decline of 29.9%.
The company built a manufacturing base for a demand curve that did not arrive on schedule, and the consequences are visible in every part of the balance sheet.
Property, plant and equipment stood at ₩23,654.7bn at the end of FY2023. At June 30, 2026 it was ₩45,923.4bn, up 94%. It now represents 59.0% of the company's ₩77,877.7bn of total assets.
Divide revenue by plant. FY2023: ₩33.75tn of revenue against ₩23.65tn of PP&E, so 1.43 won of sales per won of fixed assets. FY2025: ₩23.67tn against ₩40.79tn, so 0.58.
Annualise the second quarter of 2026, where revenue was ₩7,560.2bn, and you get roughly ₩30.2tn against ₩45.92tn of plant. That is about 0.66, an improvement on FY2025 and still less than half the FY2023 level.
For a business whose cost structure is dominated by depreciation on that plant, utilisation is close to being the whole story on margins. A factory running at half the throughput it was designed for spreads the same fixed cost over fewer cells, and the gross margin follows.
Capital spending on that scale had to be funded, and it was not funded from operations.
Cash from operating activities was ₩4,444.2bn in FY2023, ₩5,111.7bn in FY2024 and ₩4,432.3bn in FY2025. Capital spending in each of those years exceeded it, by ₩5.5tn, ₩7.3tn and ₩6.4tn respectively.
Financing activities filled the gap: inflows of ₩4,354.7bn, ₩5,381.5bn and ₩6,285.9bn across the three years, plus ₩2,657.2bn in the first half of 2026.
The result is on the balance sheet. Non-current liabilities went from ₩10,126.5bn at the end of FY2023 to ₩28,785.7bn at June 30, an increase of 184%. Total liabilities are now ₩47,591.0bn against ₩30,286.6bn of equity, a ratio of 1.57 times. At the end of FY2023 the same ratio was 0.86.
Roughly a doubling of leverage in two and a half years, at a company that has since lost money for three consecutive quarters and whose retained earnings have gone to negative ₩723.4bn.
The correction has started, and it is substantial.
Purchases of property, plant and equipment in the first half of 2026 came to ₩3,271.7bn, against ₩6,130.5bn in the first half of 2025. Down 46.6%. Annualise it and FY2026 lands near ₩6.5tn, which would be 40% below FY2025 and 47% below FY2024.
That is the right decision. A company with 0.66 of revenue per won of plant does not need more plant; it needs demand for the plant it has. Cutting spending while utilisation recovers is exactly what a capital-intensive business should do at this point in a cycle.
It arrives late, which is the criticism. The demand slowdown in electric vehicles was visible through 2024, and FY2024 was the peak capital spending year at ₩12.4tn. The company added the most capacity in the year its revenue fell 24%.
The defence is real too. Much of that spending was committed under joint venture agreements with General Motors, Stellantis and Honda, where walking away is not a unilateral option. And US production tax credits pay per kilowatt-hour of domestic output, which changes the economics of a plant built in America even at modest utilisation. A company optimising purely for near-term returns would have built less and forfeited both the customer relationships and the credits.
One line is worth flagging because it is unusual and I cannot account for it from what I have.
Cash used in investing activities was ₩12,065.4bn in FY2024 and ₩10,881.3bn in FY2025, and ₩6,157.8bn in the first half of 2025. In the first half of 2026 the investing line was positive ₩767.8bn.
That is a swing of nearly ₩7tn, in a period when the company still spent ₩3,271.7bn on plant. So investing activities generated something in the region of ₩4tn from somewhere else. Cash rose from ₩3,779.3bn at December 31 to ₩7,170.3bn at June 30.
Candidates include maturing short-term financial instruments, asset or stake disposals, or proceeds from monetising US tax credits, which are transferable under American law. The summary financials do not itemise it and the cash flow note in the half-year report will. It is a large enough number that anyone modelling this company's liquidity should find out which it was, because a one-time disposal and a recurring credit monetisation have very different implications.
Utilisation, expressed as revenue against the plant balance. It reached 0.66 on an annualised second-quarter basis. Getting back toward 1.0 would require roughly ₩46tn of annual revenue against the current asset base, which is nearly double where the company is running. That is the gap the recovery has to close, and it will close through revenue rather than through writing down assets, unless management decides otherwise.
The second thing is the debt maturity profile in the half-year report. Non-current liabilities of ₩28.79tn were built up during a period of falling rates and rising optimism, and they will have to be refinanced. A company with negative retained earnings, three quarters of losses and 1.57 times leverage refinances on different terms than the one that borrowed the money in 2024.
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.