ABL Bio (298380) fell 5.46% on 31 August to ₩72,700. The only thing the company filed around that date was a correction to a stock option grant made in 2023. The stock is down 70.4% from its 52-week high of ₩245,500 and sits 19.6% above the low.

If you try to value this company the way you'd value a manufacturer, every ratio you reach for will either be undefined or wrong. That isn't a defect in the company. It's what the accounting for pre-product biotech looks like, and the numbers are worth walking through properly, because most of the mistakes people make with names like this come from reading one line as though it meant what it means elsewhere.

The Revenue Line Is Not Revenue

ABL Bio reported ₩79.3bn of revenue in FY2025. Here is how it arrived:

First quarter, ₩2.2bn. Second quarter, ₩75.7bn. Third quarter, ₩1.4bn. Fourth quarter, zero.

One quarter carried 95.5% of the year. The next one carried nothing at all — not a small number, not a rounding figure, but ₩0.

That's because this line isn't sales. It's the recognition of upfront and milestone payments from partners, and a milestone either triggers in a period or it doesn't. Comparing the fourth quarter of 2025 to the third would tell you nothing. Comparing FY2025's ₩79.3bn to FY2024's ₩33.4bn and FY2023's ₩65.5bn tells you slightly more, but only that deals landed unevenly across three years.

The first half of 2026 is actually the least typical stretch in the data, because both quarters produced something: ₩13.1bn then ₩19.7bn. Two consecutive quarters of recognised revenue is the closest thing to a run rate this company has shown, and even that is probably the amortisation of an upfront rather than anything recurring.

What The Cost Side Says, Which Is More Useful

The expense line behaves normally, which makes it the honest half of the income statement. Back it out of the operating result and you get the cash cost of running this business: about ₩36.5bn in the June quarter, ₩30.4bn in March, ₩29.7bn in the December quarter, ₩23.9bn the quarter before.

Call it ₩30bn a quarter and rising. That is the real number. Revenue is what happens to it in a given three months; the ₩30bn happens regardless.

Cash flow says the same thing more cleanly. In the three quarters where no significant payment arrived — September 2025, December 2025, June 2026 — operating cash flow was negative ₩27.6bn, negative ₩23.6bn and negative ₩23.4bn. Consistent to within a few billion won. In the quarters where a payment landed, March 2026 and June 2025, operating cash flow was positive ₩32.8bn and ₩38.3bn.

So the underlying burn is roughly ₩25bn a quarter, and every so often a partner writes a cheque that covers three or four quarters of it at once.

Runway Is The Ratio That Actually Applies

Cash at 30 June was ₩165.0bn, against ₩186.7bn three months earlier and ₩112.0bn at the December year-end. Divide ₩165.0bn by a ₩25bn quarterly burn and you get about six and a half quarters — into early 2028 — assuming not one further won arrives from a partner.

That is the number to hold on to. Not the price-to-earnings ratio, which doesn't exist, because seven of the last eight quarters produced a loss and the trailing twelve months produced a net loss of ₩79.2bn. Not price-to-book, which is 23.4 times, because book value is ₩173.9bn and the asset being priced — a blood-brain-barrier shuttle platform licensed to large Western drugmakers — appears nowhere on the balance sheet. Research spending is expensed as incurred, so the ₩465.4bn accumulated deficit is, in a sense, the cost of building the thing the market is paying ₩4,070.3bn for.

Look at the arithmetic that way and it's clarifying. The market says ₩465.4bn of expensed research produced roughly ₩3.9tn of value not on the books. That is either a very good return or a very confident assumption, and the runway number tells you how long you have to find out before the company must raise money.

What The Balance Sheet Does Hold

Two things are worth flagging because they're unusual for a biotech this size.

Property, plant and equipment is ₩100.4bn — 36.7% of total assets. It was ₩6.7bn at the end of FY2022 and ₩76.0bn a year later. This company built something physical, and physical things depreciate against an income statement that has no product revenue to absorb them.

Second, current liabilities jumped from ₩51.5bn at 30 September 2025 to ₩123.5bn at 31 December, and stood at ₩99.0bn in June. My reading is that this is deferred revenue from an upfront payment being released as research obligations are performed, which would tie to the pattern in the revenue line. The half-year report's summary lines don't confirm that, so treat it as an inference. If it isn't deferred revenue, it's a much less comfortable number.

Dilution Is Continuous And Should Be Assumed

Financing cash flow was ₩144.6bn in FY2024, ₩32.5bn in FY2025 and ₩36.9bn in the March 2026 quarter alone. Share capital went from ₩23.89bn to ₩27.99bn between the end of FY2022 and June 2026, which at a ₩500 par value means the share count rose from about 47.8m to 56.0m, roughly 17%.

Then there are options. The 31 August filing discloses 1,422,986 options outstanding, 2.54% of shares, all settled in new stock. The tranche it corrects was granted in August 2023 at an exercise price of ₩18,920, when the previous close was ₩18,340. At ₩72,700 those are worth about ₩53,780 apiece to the holder. Ten thousand five hundred of them were cancelled because someone left before vesting, which is the sort of detail that tells you these grants are large enough to matter to the people receiving them.

None of this is objectionable. A company with no product funds itself by selling equity, and it pays people partly in equity. The point is that the share count is a moving figure and any per-share number you compute has a decay rate built into it.

The Case Against Reading It This Way

I've framed this as an accounting exercise, and someone could fairly say the accounting is irrelevant. What decides this stock is whether Sanofi advances the Parkinson's programme, whether GSK and Eli Lilly progress the deals reported last year, and whether clinical data reads out well. The company filed four separate notices between 3 June and 27 July on a single Phase 1 protocol amendment for ABL503, its bispecific immuno-oncology candidate. That's the pipeline talking, and it moves the price far more than the burn rate does.

That's true. But the burn rate determines whether the company reaches those readouts on its own terms or has to raise into weakness at a price 70% below last year's high. Both things matter, and only one of them is knowable today.

What Would Settle It

The third-quarter revenue line, and whether it's another ₩15-20bn or another zero. A third consecutive quarter of recognised revenue would suggest an upfront being amortised on a schedule, which is far more informative than a lump.

Then cash. If it is still above ₩140bn at 30 September, the ₩25bn burn holds and the runway argument stands. If it's materially lower, the cost base is growing faster than the June quarter suggested, and the date at which this company must sell shares moves closer.