On August 7 LigaChem Biosciences Inc. (KOSDAQ:141080) filed a material-management disclosure saying it had secured sole development and third-party out-licensing rights to LNCB74, a B7-H4 antibody-drug conjugate. Read quickly, that sounds like a company taking control of an asset it believes in. Read the filing, and the sequence is the other way round.
LNCB74 was a joint programme with NextCure, Inc., a US biotech founded in September 2015. The filing gives the reason the joint development stopped, and it isn't clinical: NextCure went through a reverse merger and a change of control. LigaChem then contracted to take the programme on alone. The structure of what it signed is spelled out — an upfront, then development, regulatory and sales milestones, then a royalty — and one line settles who's on which side of it. All the technology fees at each stage are non-refundable and are amounts the company pays. LigaChem is the payer.
The size is withheld under a confidentiality clause, but not entirely. The filing states that the total contract value is 10% or more of consolidated equity at the FY2025 year end, and gives that equity as ₩541,120,311,269. So the commitment is at least ₩54.1bn.
For a US reader who hasn't followed this field: an antibody-drug conjugate is three things bolted together. A monoclonal antibody that recognises a protein sitting on the surface of a cancer cell. A cytotoxic payload that is far too poisonous to inject on its own. And a chemical linker holding them together, designed to stay intact in the bloodstream and release the payload only once the antibody has been pulled inside a tumour cell. Enhertu and Trodelvy are the versions Americans have heard of. B7-H4 is a target found on several solid tumours and relatively sparse on healthy tissue, which is why more than one company is chasing it.
LigaChem's business is the linker and payload chemistry, not the antibody and not the commercial drug. It builds ADCs around antibodies it licenses in, then licenses the finished candidate out to a larger company that runs the expensive trials. That model has produced deals with Johnson & Johnson, Amgen and Ono, and it means the company's revenue is a series of upfronts and milestones rather than product sales. FY2025 revenue was ₩141.6bn. FY2023 was ₩34.1bn. There is no run rate here, only events.
The vulnerability in that model is exactly what just happened. When you don't own the antibody and don't own the trial, your rights depend on a counterparty staying itself. NextCure's corporate structure changed and the programme LigaChem had co-developed became something it had to negotiate for. The company got the right answer — sole control, plus the right to license it to a third party — and it is paying for the privilege of not losing an asset it helped build.
Korean disclosure has a habit worth knowing about here. When a listed company withholds a contract's financial terms, it does not simply go silent. The filing still has to state where the amount sits relative to a threshold — in this case that it exceeds 10% of consolidated equity — because that threshold is what triggered the disclosure obligation in the first place.
The result is a floor rather than a figure. An investor gets ≥₩54.1bn and nothing more precise. A US company in the same position would file the agreement as a redacted exhibit, and the reader would see the contract's structure with dollar amounts blacked out. The Korean version inverts that: no document, but a lower bound you can compute. Neither is obviously better. But an American reading this filing and concluding the amount is unknown has missed the one quantitative thing in it.
The same filing carries a candid risk block: milestones may never be paid if the project doesn't progress, and the in-licence can terminate early on clinical or regulatory failure with no penalty owed. That cuts both ways. The ₩54.1bn floor is a ceiling on the downside too, since most of it is contingent.
The timing is what makes it uncomfortable. Second-quarter revenue was ₩5,532,886,956 — ₩5.5bn — against ₩32.6bn in the same quarter last year. First-half revenue of ₩41.4bn is down 51% from ₩84.2bn. The operating loss for the half was ₩109.4bn, which is already larger than the ₩106.5bn lost across all of FY2025.
Where that loss comes from is worth spelling out, because the summary lines don't show it. Take Q2 revenue of ₩5.5bn, subtract ₩2.4bn of cost of revenue and ₩5.7bn of SG&A, and you get about negative ₩2.6bn. Reported operating income was negative ₩72.0bn. The ₩69.4bn difference is research and development, which for a clinical-stage ADC developer is the entire business. Across FY2025 the same arithmetic implies roughly ₩216.1bn of R&D against ₩141.6bn of revenue.
So the company spends more on research than it earns, by design, and it funds that from the balance sheet. Operating activities consumed ₩106.1bn in the first half. Cash and equivalents were ₩84.1bn at the end of June, down from ₩153.5bn a year earlier, but that line understates the position badly — current assets were ₩426.1bn, most of it in financial instruments the company has been rolling into cash as it needs them. Investing activities produced ₩73.7bn of inflow in the half for exactly that reason. Current assets fell ₩95.3bn in six months, which is the honest burn number, and on that pace the existing base lasts something over two years.
Equity has taken the hit: ₩616.0bn at the end of FY2024, ₩541.1bn at FY2025, ₩441.0bn in June. Retained earnings have gone from ₩311.8bn in March 2025 to ₩102.0bn.
On July 24 the company completed a ₩170bn issue of series 6 convertible bonds: bearer, unsecured, privately placed, and zero-coupon. The bondholder gets no interest at all and is paid entirely in the conversion option.
What's telling is the paper trail. The board first resolved on the issue on June 25, then filed revised terms on July 16, then again on July 22, before payment settled on July 24. Three versions of one deal inside a month. The filings run in pairs — a convertible bond decision alongside a paid-in capital increase decision each time — so the ₩170bn is one leg of something larger. Korea Biomedical Review has reported that Korea's semiconductor fund is backing the company to the tune of $323m in its first biotech investment, which would put the equity leg well above the convertible. The filings I've read cover only the ₩170bn.
At ₩94,900 the market capitalisation is ₩3.51tn, which is 7.97 times the ₩441.0bn of equity and cannot be assessed on earnings because there aren't any. The shares are 56% below the ₩213,500 fifty-two-week high.
None of this says the science is failing. In July the company disclosed a milestone on LCB97, an L1CAM-targeting ADC partnered with Ono, triggered by first patient dosing, and press reports put it at ₩14.1bn or more. That lands in Q3, not Q2, which is part of why Q2 looked as thin as it did. A platform company with partners at J&J, Amgen and Ono, taking full control of a B7-H4 asset while ADC dealmaking is still active, is not a company in retreat. Taking sole rights also means it keeps the whole economics if it re-licenses LNCB74 to someone larger, and that is plainly the plan.
The bear case isn't that the pipeline is bad. It's that this is a company whose revenue arrives at unpredictable intervals, which has just committed to pay at least ₩54.1bn it didn't plan to pay, funded by a convertible it had to re-price twice, at eight times book.
Q3 revenue in November, and whether the Ono milestone plus anything else pulls it back above the ₩35.9bn of Q1. Two quarters in a row near ₩5bn would say the licensing engine has stalled rather than paused.
Then any out-licensing announcement on LNCB74. LigaChem bought the right to sell this to a third party, and the deal only makes sense if it finds one. The FY2026 annual report in March is also where the R&D line finally appears explicitly rather than as a gap in the arithmetic.
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.