Ecopro BM Co., Ltd. (KOSDAQ: 247540) earned ₩18.0 billion of operating income in the June quarter and finished it with a net loss of ₩0.45 billion. Nothing dramatic happened in between. Finance costs of ₩25.9 billion outweighed finance income of ₩10.5 billion, a small amount of other expense took the rest, and a ₩2.2 billion tax charge on ₩1.75 billion of pretax income, which happens when profitable subsidiaries pay tax while the group as a whole barely breaks even, finished the job.
That single line is the company in miniature. Korea's largest cathode maker, supplier of high-nickel material to Samsung SDI and SK On, is running plants that cost more to finance than they currently produce. Over the last four quarters, finance costs of roughly ₩110 billion have absorbed 84% of ₩131 billion in operating income. And on June 30 the board asked shareholders for ₩1.2 trillion.
Understanding why a 3.1% operating margin isn't quite as bad as it sounds, and why it also isn't fine, requires knowing what the revenue line is. Cathode material is sold on a formula: the cost of the nickel, cobalt, lithium and aluminium in the product, passed through, plus a conversion fee for turning it into powder. When metal prices fall, revenue falls with almost no effect on profit. When they rise, the reverse.
That's why revenue went from ₩6,900.9 billion in 2023 to ₩2,766.8 billion in 2024 to ₩2,531.6 billion in 2025 without the company shrinking by anything like that much. Volumes did fall, and the EV slowdown was real, but a 63% revenue decline in two years is mostly the lithium price. June-quarter revenue of ₩576.7 billion was down 26% from a year earlier and down 5% from March.
So the number that matters isn't revenue. It's what the conversion fee earns against the assets built to collect it. Property, plant and equipment stood at ₩3,377.2 billion at the end of June, up from ₩2,525.0 billion in September 2024 and ₩1,003.5 billion at the end of 2022. Annualise the June quarter's ₩18.0 billion of operating income and you get about ₩72 billion against a ₩3.4 trillion plant base. That's roughly a 2% return on the factories, before interest.
Depreciation on assets that big doesn't care about utilisation. This is the whole problem, and it explains the shape of the last three years better than any story about EV demand: Ecopro BM built for a 2023 volume forecast and is running against 2026 volumes.
On June 30 the board approved a rights offering of 9,900,990 new shares against 97,830,434 outstanding, about 10.1% dilution, at an expected price of ₩121,200 for total proceeds of ₩1.2 trillion. The record date is September 4, warrants trade from September 28 to October 2, the final price is fixed October 12, existing holders subscribe October 15 and 16, payment is October 23, and the new shares list November 5. Parent company Ecopro Co. has resolved to take part.
The use-of-proceeds table is the interesting part. Facilities take ₩150.0 billion. Working capital takes ₩135.0 billion. Debt repayment takes nothing at all. And ₩915.0 billion, more than three quarters of the raise, is allocated to the acquisition of securities issued by other corporations.
The filing doesn't say which corporations. In a group structured like this one, with cathode and precursor and recycling and an Indonesian nickel operation spread across affiliates, the plausible reading is recapitalising subsidiaries that carry the overseas plants. Whatever the destination, shareholders being asked for ₩1.2 trillion are being told that ₩915 billion of it leaves the operating company, and that none of it retires the debt whose interest just erased the quarter's profit.
The Financial Supervisory Service asked for a correction to the June 30 registration statement on July 14, and amended versions have been filed on July 24, August 7 and August 18. Regulators send prospectuses back often enough that this isn't scandalous on its own. Three rounds on a deal this size is still worth knowing about.
Here's the part that hasn't been widely noticed. The ₩1.2 trillion figure is the share count times ₩121,200, and ₩121,200 was calculated off the market price on June 29 with a 20% discount already applied. The share count is fixed. The price isn't.
Ecopro BM closed at ₩118,200 on August 28, which is already below the expected issue price, and the final price will be struck at a 20% discount to the prevailing price in early October, floored at 60% of the reference. Hold the stock at ₩118,200 and the final price lands somewhere near ₩95,000, which raises about ₩940 billion rather than ₩1.2 trillion. The gap has to come out of one of the three uses, and given the sizes, it comes out of the ₩915 billion.
There's a reflexive quality to this that shareholders should sit with. The plan is sized off a share price, the offering itself pressures the share price, and short selling has been barred for subscribers since July 1 precisely because everyone knows how that dynamic works.
Operating cash flow was negative ₩72.2 billion in the June quarter and negative ₩88.9 billion in March, so negative ₩161.0 billion across the half against ₩39.0 billion of operating income. Inventories rose from ₩602.8 billion at the end of 2025 to ₩791.9 billion in June, an increase of ₩189.1 billion in six months while revenue was falling.
That build has an optimistic reading and a pessimistic one. Lithium prices have been rising and the company says higher prices are now being reflected in selling prices, which means inventory bought cheap gets sold dearer, and a rising metal price mechanically lifts the value of what's on the floor. The pessimistic reading is that material accumulates when customers take less than planned. The half-year report doesn't settle which it is.
Meanwhile current liabilities of ₩2,121.7 billion exceed current assets of ₩1,681.8 billion by ₩439.9 billion, cash sits at ₩467.1 billion, and financing brought in ₩263.4 billion in the June quarter alone. This is a company that needs the October money.
The bull case is not stupid and it's roughly this: the spending is done, the earnings aren't. Capital expenditure was ₩1,023.8 billion in 2024, ₩418.1 billion in 2025, and only ₩161.7 billion in the first half of 2026. The plants in Hungary and Canada exist, they're positioned for North American and European content rules, and the fixed cost is already in the depreciation line. A recovery in volumes drops through at very high incremental margin because the conversion fee has almost no variable cost attached once the building is paid for.
There's contracted demand behind it. The 2023 agreement to supply NCA cathode to Samsung SDI runs from 2024 to the end of 2028 and was valued at around ₩44 trillion, and Samsung SDI and SK On together take about 94.7% of sales. The portfolio has widened past high-nickel into mid-nickel, LMR, LFP and sodium-ion chemistries, which addresses the criticism that the company bet everything on one formulation. And the Indonesian nickel smelter disruption that hurt the June quarter has been repaired.
That's what 146 times trailing earnings is actually paying for. Not the earnings. The installed base, and a view about when it runs full.
October 12 is the date. The final issue price fixed that day determines whether the raise brings in ₩1.2 trillion or something closer to ₩940 billion, and the company will have to say which of the three stated uses absorbs the difference. Watch too what proportion of the offering existing shareholders actually take up, because parent Ecopro's participation is announced and everyone else's is not.
Then the third-quarter report. Two lines in it: cost of revenue against revenue, which shows whether rising lithium prices are reaching the margin the way management says, and operating cash flow, which has now been negative for two straight quarters. A third would matter more than any headline about capacity.
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.