LG Energy Solution, Ltd. (KRX:373220) lost money before tax in each of the last three quarters. Negative ₩476.1bn in the fourth quarter of 2025, negative ₩858.6bn in the first quarter of 2026, negative ₩127.8bn in the second.
In each of those quarters it also recorded an income tax expense. ₩296.4bn, then ₩85.4bn, then ₩200.8bn.
Add it up and the company generated ₩1,462.6bn of pretax losses and paid ₩582.6bn of tax on them. Net losses came to ₩2,045.1bn. The tax line made the result 40% worse rather than cushioning any of it.
That is unusual enough to be worth explaining, and the explanation matters for how you read every future quarter.
Take the annual series first. FY2023: pretax income of ₩2,043.5bn, tax of ₩405.5bn, an effective rate of 19.8%. FY2024: ₩348.9bn of pretax income, ₩10.3bn of tax, 2.9%. FY2025: ₩414.1bn of pretax income, ₩333.3bn of tax, 80.5%.
Then the quarters above, where the rate stops being definable because the denominator is negative.
A company whose effective tax rate goes from 20% to 3% to 80% to undefined in three years is not experiencing changes in tax policy. It is experiencing changes in where its profits and losses sit.
The first is geography. LG Energy Solution's manufacturing is spread across Korea, Poland, China and, increasingly, the United States, where it runs joint venture plants with General Motors, Stellantis and Honda in Ohio, Tennessee and Michigan. Those US entities receive production tax credits under American law, worth a fixed amount per kilowatt-hour of cells manufactured domestically, and at times those credits have exceeded the group's entire reported operating profit.
A US entity generating taxable income pays US tax. A Korean parent generating losses gets no relief from that, because tax systems do not net across borders. So the consolidated income statement can show a group loss and a group tax charge simultaneously, because the two arise in different countries.
The second is deferred tax recognition. When a company records losses, it normally books a deferred tax asset, on the basis that the losses will shelter future profits. Accounting standards allow that only when future taxable profit is probable. A company that has lost money for several consecutive periods struggles to make that case, and auditors then require it to stop recognising the asset, or to write off assets already booked.
The FY2025 pattern fits that. An 80.5% effective rate on positive pretax income of ₩414.1bn suggests a large non-deductible item or a write-down of previously recognised deferred tax assets, and the latter is the more common cause at a company whose profitability was deteriorating through that year.
I want to be clear these are inferences. The tax reconciliation note in the half-year report sets out exactly which items produced the charge, and anyone doing real work on this company should read it rather than take my reasoning.
If the second explanation is doing most of the work, the implication is more interesting than it first appears.
Unrecognised tax losses do not disappear. They sit unused, and when the company returns to sustained profitability they shelter that profit, producing an effective tax rate well below the statutory one for several years. The accounting treatment reverses: the deferred tax asset gets recognised, often in one large credit.
So the current arithmetic understates economic performance, and a future recovery will overstate it. Anyone building a model that applies a 22% Korean tax rate to forecast pretax profits in 2028 will be wrong in both directions, and the error is large.
The size of the unrecognised balance is disclosed in the notes. It is the single most useful number for anyone trying to value the eventual recovery.
Underneath all this, the actual business improved in the most recent quarter.
Operating income was negative ₩207.8bn in the first quarter of 2026 and positive ₩113.3bn in the second. Revenue rose from ₩6,555.0bn to ₩7,560.2bn, up 15.3% sequentially and 35.8% against the ₩5,565.4bn of the second quarter of 2025. Gross profit was ₩1,545.3bn, a margin of 20.4%, against 18.0% in the previous quarter.
So the company sold considerably more, at better gross margins, and made money at the operating line, and still reported a ₩328.6bn net loss. The full gap runs through net finance costs of ₩211.2bn and the ₩200.8bn tax charge.
That divergence is the practical reason to care about the tax line. The operating business and the reported bottom line are telling different stories, and for the moment the bottom line is the less informative one.
Three things.
The US production tax credits are a policy variable. They have been the largest single support to this company's reported profitability, and any narrowing of eligibility or of the credit rate would hit both the operating line and, through the profitability of the US entities, the tax position.
The finance costs are growing with the balance sheet. Non-current liabilities went from ₩17,285.3bn at the end of FY2024 to ₩28,785.7bn at June 30, up 66.5% in eighteen months, funding a property, plant and equipment base that now stands at ₩45,923.4bn. Net finance costs of roughly ₩200bn a quarter on that debt are not going away, and they consume the operating profit the business is currently capable of generating.
And retained earnings have gone negative, from ₩2,364.5bn at the end of FY2023 to negative ₩723.4bn at June 30. A company with an accumulated deficit has constraints on what it can distribute under Korean law, which matters if the market ever expects a dividend.
The tax reconciliation note in the FY2026 annual report, and specifically the balance of unrecognised deductible temporary differences and unused tax losses. That number tells you how much sheltered profit is waiting, and whether the current charges are a geographic mismatch, a recognition decision, or both.
The nearer marker is the third quarter, in late October. If operating income holds positive and the tax charge falls toward zero, the recognition issue is easing and reported net income will start converging on the operating result. Another quarter of large tax expense against a small operating profit would mean the mismatch is structural, and reported earnings will keep understating this business for as long as the losses sit in the wrong country.
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.