Celltrion's biggest buyback won't cancel a share

Between 25 August and 1 September, nine Korean listed companies told the regulator they intend to buy back their own shares. Line the filings up by size and one company towers over the rest — and it is not one of the three that promised to shrink its share count.

The filings, ranked

Source: DART treasury-stock acquisition filings, 2026-08-25 to 2026-09-01. English names are DART’s own registered translation.

Buy, or buy and cancel

A buyback and a cancellation are not the same step. Buying takes shares off the market and can support the price, but the shares still exist — sitting in treasury, available to be sold again or handed out later. Cancellation is the part that is irreversible: it permanently reduces the share count, so every remaining share represents a slightly larger claim on the company. Filed purposes like “price stability” or “shareholder value” describe the first step only; a company has to say the second part out loud for it to count.

Of the seven filings in this window that describe an ordinary, discretionary buyback, only three commit to cancellation in the filing itself: Chinyang Polyurethane, Chinyang Chemical and Ghost Studio — a combined ₩6.1bn. The other four, including the single largest filing on the list, do not.

The size doesn’t predict the commitment

Celltrion’s plan is ₩100.0bn — bigger than the other eight filings put together, several times over. Its stated purpose, like most of the others, is price stability and shareholder value. It does not say the shares will be cancelled.

Put the two facts side by side and the ranking asks its own question: if the largest buyback of the week doesn’t retire a single share, and the ones that do are a fortieth its size, does the size of a buyback tell an outside reader anything about whether it will actually shrink the company’s share count? On this week’s filings, the two numbers move independently.

What we left out, and why

Two filings on the list are not ordinary discretionary buybacks, and we do not count them in the comparison above. EZmedicom’s ₩5.34bn purchase is triggered by dissenting shareholders exercising appraisal rights after a merger — the company is legally required to buy those shares, so “will it cancel them” is a different question than for a voluntary buyback. Justem’s ₩1.22bn filing states its purpose as funding employee compensation — the shares are bought to be distributed, not held or retired, so cancellation was never the plan. Both are real filings and both are in the table; neither belongs in a count of “who promised to cancel and who didn’t.”

This is a record of what nine companies filed with DART in one seven-day window, drawn from the primary disclosures, and it is not investment advice — nothing here says any of these nine is worth owning or avoiding. What it shows is narrow and exact: on this week’s evidence, a bigger buyback is not a more committed one.

The debate

Should regulators or exchanges require a buyback filing to state upfront whether the shares will be cancelled — or is “price stability,” left vague, a legitimate purpose on its own? What do you think? This is a talking point, not a verdict — argue it out below.

Buybacks shrink the share count when cancelled; a rights issue does the opposite — and the same week’s filings show size doesn’t predict how much it dilutes existing holders either. See the three-instrument synthesis for the full pattern.