Look at KT&G Corporation's (KRX:033780) retained earnings across one quarter. At 31 December 2025: ₩1,723,397m. At 31 March 2026: ₩8,276,355m.
The company earned ₩378,240m in that quarter. Total equity went from ₩9,336,169m to ₩9,327,855m — down slightly. Share capital was unchanged at ₩954,959m.
So roughly ₩6.2tn appeared in retained earnings without being earned, and without total equity changing. Money moved from one shelf inside equity to another.
Only one mechanism produces that pattern: a capital reserve reduction. Under Korea's Commercial Act a company can, with shareholder approval, reduce its legal capital reserve and transfer the amount into retained earnings. Nothing about the company's assets, liabilities or net worth changes. What changes is how much of the equity is legally available to distribute.
KT&G's annual general meeting was held on 26 March 2026, five days before the quarter closed. The timing fits exactly, and I would read the AGM resolutions to confirm it. The consolidated summary does not label the movement, so this is an inference from the roll-forward rather than a quotation. But equity does not behave this way for any other reason.
Understanding why it matters requires knowing something odd about KT&G's balance sheet. This is a company with ₩9.3tn of equity that was carrying only ₩1.7tn of retained earnings. The rest sat in capital surplus and other reserves, largely a legacy of its privatization from the state tobacco and ginseng monopoly. Korean law caps distributions at distributable profit, which is anchored on retained earnings. So a company generating ₩1.1tn of annual net income was operating with a distributable base barely larger than a year and a half of it.
Converting ₩6.2tn multiplies that base roughly fivefold. It does not create a single won of value. It removes a legal ceiling.
What followed reads as a sequence rather than a series of separate announcements.
On 16 April, three weeks after the AGM, the board resolved to cancel 10,866,189 common shares out of 114,676,645 issued — 9.5% of the company — with the amount stated as ₩1,851,598,605,600, computed at the prior session's close of ₩170,400. Six outside directors attended and none were absent. The shares were retired on 23 April.
The filing carries a revealing note: the company decided to cancel its entire treasury holding in light of the revised Commercial Act and shareholder value considerations, which is why the quantity differs from what the 2025 annual report's treasury share section had planned. KT&G cancelled more than it had told the market it would, and it said so.
Then on 6 August the board declared a quarterly dividend of ₩2,000 per share against ₩1,400 a year earlier, a 43% increase. The total came to ₩207,620,912,000 on 103,810,456 shares — exactly the post-cancellation count. Record date 21 August, payment 7 September, stated yield 1.1%.
Annualize ₩2,000 a quarter across 103.8m shares and the run rate is ₩830.5bn a year. KT&G paid ₩603.8bn in dividends in FY2025. That is a 37% increase in cash distributions, on a share count 9.5% smaller.
Put the numbers together and the scale of what was unlocked becomes clear.
Before the conversion, distributable reserves of ₩1.7tn would have covered roughly two years of the old dividend and nothing else. After it, a base near ₩8.3tn covers the current ₩830bn annual dividend for a decade, or funds several more cancellations the size of April's ₩1.85tn, or some combination.
Management has said it will announce a new medium- to long-term shareholder return policy in the fourth quarter of 2026, with a stronger emphasis on dividends alongside continued buybacks, and that further buyback and cancellation activity is planned for the second half with timing and scale depending on the share price. That announcement is now the most consequential thing on KT&G's calendar, and the March reserve conversion is what makes an ambitious version of it possible.
Working back from the stated 1.1% quarterly yield, the shares sat near ₩182,000 in early August, implying a market value around ₩18.9tn. Against trailing four-quarter net income of about ₩1,441.2bn that is roughly 13 times earnings, with the annualized dividend alone worth about 4.4%.
The reserve conversion is a legal manoeuvre, not an operating improvement, and it should not be confused with one. KT&G's operating income was ₩1,473.2bn in FY2020 and ₩1,343.7bn in FY2025. Six years on, the company earns less from operations than it did before the pandemic, even after a genuine recovery from the FY2023 trough of ₩1,167.3bn.
Gross margin has also drifted. Cost of revenue was 49.4% of sales in FY2022 and 52.2% in FY2025, which reflects leaf tobacco costs and a growing share of lower-margin supply business. It recovered to 49.3% in the second quarter of 2026, which is encouraging, but the multi-year direction has been the wrong one.
Returning capital more aggressively raises per-share figures on a business that is not growing much. That is a legitimate strategy for a mature cash generator and it is precisely what tobacco investors want. It is not the same as the business getting better, and a company that has just handed itself ₩6.2tn of distribution headroom will find the temptation to substitute one for the other quite strong.
The other side is straightforward. KT&G was carrying a legal constraint that served nobody — capital surplus locked up from a privatization two decades ago, preventing a profitable company from returning money it plainly did not need. Removing it required a shareholder vote, which it got. Cancelling 9.5% of the shares outright rather than parking them in treasury is the cleaner version of a buyback, and doing it as the first company to act under the revised Commercial Act is leadership rather than opportunism.
The foreign ownership register supports that reading. Capital Research, BlackRock and First Eagle all filed substantial shareholding reports in the months around these actions. This is a shareholder base that has pushed for exactly this, and got it.
The fourth-quarter return policy announcement. Two numbers: the annual payout ratio committed to, and the size of any new buyback authorization. With ₩8.3tn of distributable reserves and roughly ₩1.4tn of annual net income, a policy that commits to returning 70% or more would be genuinely aggressive; one that lands near 60% would suggest the reserve conversion was about optionality rather than intent.
The second thing is where the buyback shares come from. Treasury holdings went from 12.0% to zero in April, so every future cancellation now requires open-market purchases first. That makes the next programme visible in daily trading data in a way the last one was not.
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.