SK Innovation Co., Ltd. (KRX:096770) reported ₩3,487,296,471,000 of operating income for the second quarter of 2026. In the same quarter of 2025 it reported an operating loss of ₩417,568,625,000.
That is a swing of nearly ₩3.9tn in a single year, on revenue of ₩29,157.2bn. It is the best operating quarter in the company's history and it beat expectations by a wide margin; the shares rose 6.5% on the day.
Net income for the quarter was ₩73,463,246,000.
Two percent of the operating profit reached the bottom line.
Pretax income was ₩569.7bn, so ₩2,917.6bn disappeared between the operating line and the pretax line. Then income tax took ₩521.1bn, an effective rate above 90% on what was left.
The finance lines account for much of it: finance income of ₩858.7bn against finance costs of ₩2,394.2bn, a net charge of ₩1,535.4bn. The remainder sits in other non-operating items that the summary statements do not name. Press coverage of the results described a non-operating hit of around ₩2.6tn, and the company separately disclosed a derivative trading loss on 30 July, the same day it released the preliminary results.
None of this is new. It is the defining feature of SK Innovation's accounts and it is why the company is so widely misread.
Look at three years of it. Finance costs were ₩6,145.2bn in FY2023, ₩6,360.9bn in FY2024 and ₩6,362.6bn in FY2025. Over the same three years operating income was ₩1,880.6bn, ₩355.7bn and ₩448.7bn.
In FY2025 the company's gross financing charges were fourteen times its operating profit.
Finance income partly offsets — ₩5,232.4bn, ₩4,647.3bn and ₩4,769.3bn across the same years — because SK Innovation holds enormous foreign currency assets and liabilities and both sides revalue. But the net charge has been ₩0.9tn to ₩1.7tn a year, and the gross figures swing violently: finance income was ₩3,348.5bn in the first quarter of 2026 and ₩858.7bn in the second, while costs went ₩4,290.6bn then ₩2,394.2bn.
An investor modelling this company from operating income and applying a normal tax rate will be wrong every single quarter. The balance sheet, not the refinery, sets the reported result.
Total liabilities stood at ₩64,173.5bn at 30 June against equity of ₩37,809.5bn. Non-current liabilities alone are ₩29,129.1bn. That is the machine generating the finance line.
Strip the noise and the operating improvement is real and worth understanding, because it comes from three different places.
SK On, the battery unit, swung to an operating profit of ₩821.8bn. Two things drove it: expanded sales in Asia, and customer compensation payments. The second deserves a flag — compensation from a customer, presumably for volumes contracted and not taken, is money that arrives once. It is not a run-rate.
More durable is the unwinding of BlueOval SK, the joint venture with Ford. Completing that restructuring is expected to save roughly ₩500bn a year in depreciation and interest. That is a permanent reduction in the cost base and it flows through every future quarter.
Lubricants had a strong quarter, which matters more than its size suggests because base oil margins have been unusually good across the industry.
Refining, the business most people think of when they hear the name, was weaker sequentially as oil prices fell from June. That is worth holding onto: the quarter that produced record group operating profit was not a refining quarter.
Across the first six months, operating income was ₩5,649.5bn against an operating loss of ₩462.2bn a year earlier. Revenue was ₩53,369.3bn, up 31.9%. Net income for the half was ₩969.6bn.
So the half converted 17% of operating profit into net income. That is better than the second quarter's 2%, because the first quarter had a large positive swing in finance income. It is still a long way from normal.
For scale: FY2025 produced ₩448.7bn of operating income for the entire year and a net loss of ₩5,436.4bn. The first half of 2026 alone produced twelve times the operating income of all of 2025.
The most encouraging line in the cash flow statement is one that gets no attention. Financing activities were an outflow of ₩3,842.0bn across the first half of 2026, on the cumulative basis Korean interim statements use. In FY2025 financing was an inflow of ₩2,345.7bn; in FY2024 an inflow of ₩7,327.1bn; in FY2023 an inflow of ₩9,490.4bn.
After three years of borrowing on an extraordinary scale to build battery capacity, SK Innovation is repaying.
Capital spending has collapsed alongside it. Purchases of property, plant and equipment were ₩11,238.1bn in FY2023, ₩10,027.9bn in FY2024, ₩5,367.2bn in FY2025 and ₩1,814.4bn across the first half of 2026. The building phase is over.
If that continues, the finance cost line eventually falls, and the gap between operating income and net income narrows. That, rather than any refining margin, is the mechanism that would re-rate this company.
Three objections deserve stating.
First, the quarter contained one-offs on both sides. Customer compensation helped SK On; a large non-operating charge hurt the group. Netting a good one-off against a bad one does not produce a run rate.
Second, the operating improvement leans on a battery business that has lost money for most of its existence and has just written off a great deal of plant. One profitable quarter, partly compensation-driven, is not a turnaround established.
Third, refining — still the largest business by revenue — softened during the quarter as crude fell. If oil keeps falling into the second half, the segment that has historically funded everything else gets weaker precisely as the battery segment is being asked to stand on its own.
The honest counter to all three is that the direction of every important line is now correct at the same time: operating income up, capex down, debt being repaid, and the loss-making joint venture unwound. That has not been true at this company since 2021.
Finance costs in the third quarter. They ran ₩2,394.2bn in the second and ₩4,290.6bn in the first. A third-quarter figure below ₩2tn, with financing outflows continuing, would show the deleveraging reaching the income statement rather than just the balance sheet.
Second, SK On's operating result excluding compensation income. The company should be asked for it directly. A battery business earning several hundred billion won a quarter on product sales alone is a different asset from one that broke even and received a cheque.
Third, whether capital expenditure stays near the first half's ₩1.8tn annualised pace. Two more quarters at that level would make free cash flow comfortably positive for the first time in five years, and would finally give this company an earnings figure that means something.
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.