Between the start of FY2023 and the end of FY2025, LG Chem, Ltd. (KRX:051910) spent ₩41,235.5bn buying property, plant and equipment. Broken out: ₩12,959.9bn, then ₩14,614.9bn, then ₩13,660.7bn. Another ₩4,141.6bn went out in the first half of 2026.
To put ₩41tn in proportion, total group equity at the end of FY2022 was ₩37,481.1bn. LG Chem spent more than its entire book value on new plant in three years.
Revenue over the same period went from ₩55,043.7bn in FY2023 to ₩48,699.8bn in FY2024 to ₩45,932.2bn in FY2025. Down 16.6%.
The balance sheet records what the money bought. Property, plant and equipment was ₩29,662.7bn at the end of FY2022. At the end of June 2026 it was ₩61,793.2bn. It more than doubled.
Divide revenue by plant and the picture is stark. In FY2023, LG Chem generated ₩1.41 of revenue for every won of net property, plant and equipment. In FY2025 it generated ₩0.81. Annualize the second quarter of 2026 and the figure recovers to about ₩0.92, still well short.
Most of that spending is not chemicals. LG Chem consolidates LG Energy Solution, and the bulk of the ₩41tn went into battery cell plants — the North American joint ventures and wholly owned sites, plus European capacity, plus cathode material facilities LG Chem built itself. That was a deliberate bet, made when every forecast said electric vehicle demand in the United States would compound through the decade and when American industrial policy was paying manufacturers to build there.
The demand did not arrive on the schedule the plants were built to. That is not hindsight applied unfairly; it is what the impairments say.
In the fourth quarter of 2025 LG Chem reported an operating loss of ₩413.3bn and a pretax loss of ₩2,489.5bn. The ₩2.1tn gap between those two lines is not an operating problem. It is asset writedowns, which the company has acknowledged taking in that quarter.
For the full year, group net income came to a loss of ₩977.1bn against operating income of ₩1,180.9bn. A company can report positive operating income and a large net loss only when something below the line is very wrong, and here what was wrong was the carrying value of plant built for volumes that had not materialized.
Retained earnings tell the cumulative story: ₩19,650.6bn at the end of FY2023, ₩16,999.3bn at the end of FY2025, ₩16,720.6bn at the end of June 2026. Nearly ₩3tn of accumulated profit has been consumed since the peak.
Operating cash flow across FY2023-25 was ₩7,536.5bn, ₩7,012.3bn and ₩8,233.9bn — ₩22,782.7bn in total. Against ₩41,235.5bn of capital spending, that leaves roughly ₩18.5tn to be found somewhere else.
It was found in debt. Total liabilities went from ₩30,492.7bn at the end of FY2022 to ₩63,338.9bn at the end of June 2026. Non-current liabilities went from ₩14,033.1bn to ₩35,764.8bn, and rose ₩6,757.8bn in the second quarter of 2026 alone.
And it was found by cutting the dividend. Dividends paid were ₩1,104.1bn in FY2023, ₩367.3bn in FY2024 and ₩226.6bn in FY2025 — down 79% in two years. First-half 2026 shows ₩186.1bn.
So the sequence over three years reads: spend ₩41tn, borrow ₩18tn, cut the dividend by four-fifths, impair the assets, report a loss. That is the whole story of the capital cycle in one company.
The first half of 2026 contains a clue about how the next chapter gets funded. Investing activities show a net outflow of only ₩316.3bn despite ₩4,141.6bn of capital expenditure. Something on the order of ₩3.8tn came back in from investing.
Management has said proceeds from selling down its LG Energy Solution stake will support growth investment and shareholder returns. That is the most consequential decision on the table. LG Chem's control of the battery maker is the reason the group is valued the way it is, and monetizing it to pay for capital spending already incurred is a very different act from monetizing it to fund something new.
Selling a stake in your best asset to service the balance sheet you built acquiring capacity for that asset is a circle worth watching closely.
Take the other side properly, because it is strong.
In 2022 and 2023 there were exactly three non-Chinese battery cell makers of scale. Customers were signing multi-year supply agreements and demanding local capacity, American policy was paying for domestic plants, and a manufacturer that did not build would have watched CATL and BYD take the entire market. The capacity had to be built ahead of demand because that is how the automotive supply chain works — a carmaker does not award a platform to a supplier who says the factory will be ready later.
Building early and eating a few years of underutilization is a defensible price for holding a position in an industry that will be very large. If EV penetration resumes at anything like the trend of 2020-2023, LG Chem owns plants that would take four years and ₩40tn to replicate, on a balance sheet that has already absorbed the cost.
The second-quarter numbers support the patient reading. Operating income of ₩599.6bn against a first-quarter loss, net income positive at ₩66.8bn after two loss-making quarters, gross margin at 21.1% — the best in the eight quarters on file. Revenue of ₩14,175.9bn was up 24.2% year on year. If that continues, the asset turnover ratio starts recovering on its own without another won of spending.
Revenue per won of property, plant and equipment. It fell from 1.41 to 0.81 and is around 0.92 annualized on the second quarter. That single ratio is the honest scorecard for a ₩41tn capital programme, and it needs to get back above 1.0 before anyone can say the money was well spent.
The second number is capital expenditure itself. Management has talked about prudent capex in 2026. First-half spending of ₩4,141.6bn annualizes to about ₩8.3tn, which would be roughly 40% below the FY2024 peak. If the full year comes in near that, the building phase is genuinely over and free cash flow turns positive for the first time since 2022.
The third is what happens to the LG Energy Solution stake. How much gets sold, at what price, and whether the proceeds go to debt reduction or to shareholders. That decision will tell you what the board thinks the last three years actually cost.
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.