APR Co., Ltd. (KRX:278470) earned ₩289.7bn in FY2025. Average shareholders' equity across that year was roughly ₩384.7bn. That is a return on equity above 75%.
Run the same calculation on the first half of 2026 — net income of ₩258.8bn, annualized against average equity of about ₩546.9bn — and the figure is closer to 95%. Return on assets in FY2025 was above 43%.
Companies do not earn those returns for long. Either the business has a genuine barrier around it, or capital arrives and competes the returns away. Which of those describes APR is the only question that matters for a stock trading north of 30 times earnings.
The reason the returns are so high is that there is very little capital in the denominator.
Property, plant and equipment was ₩69.9bn at the end of June 2026, against total assets of ₩1,102.1bn. Intangible assets were ₩7.3bn. So under 7% of the balance sheet is productive fixed asset, and effectively none of it is acquired goodwill.
Capital spending tells the same story. Purchases of property, plant and equipment were ₩15.4bn in FY2025 against revenue of ₩1,527.3bn — about 1% of sales. FY2024 was heavier at ₩45.0bn, which looks like a one-off build, and the first half of 2026 ran ₩16.8bn.
Set that against the companies APR is often bracketed with. A cosmetics manufacturer with its own plants carries tens of billions in equipment per production line. A consumer electronics maker carries more. APR generates over ₩1.5tn of annual revenue on ₩70bn of plant, which means it does not own most of what makes its products.
The balance sheet is instead working capital: ₩369.9bn of inventory, ₩191.6bn of receivables, ₩94.5bn of cash. It is a trading and marketing business with a manufacturing supply chain attached, not a manufacturer.
Where the money actually goes is visible in the gap between two margins. Gross margin in the second quarter was 79.2%. Operating margin was 24.8%. The ₩417.2bn in between, roughly 54% of revenue, buys advertising, creator programmes, marketplace commissions, logistics and the people who run them.
That is the moat, such as it is. Nothing in APR's cost structure stops a competitor from contracting the same Korean manufacturers, formulating a similar serum and listing it on the same marketplace. What stops them is that APR has learned how to acquire customers at a cost that leaves 25 points of operating margin, at scale, in markets where the customer acquisition auction is contested by every consumer brand on earth.
Marketing capability is a real capability. It is also the most imitable kind, and it depreciates. A brand that owns a shelf position at a US club retailer has something durable; a brand that owns an advertising formula has something that works until the platform changes its algorithm or the cost per acquisition rises.
The trend in that ratio is therefore the most important number APR reports and the one nobody leads with. Operating expenses ran 63.4% of revenue in FY2022, 58.2% in FY2024 and 52.7% in FY2025. In the second quarter of 2026 they were 54.4% — the first meaningful step back up.
One quarter is not a trend. It is the quarter to watch repeated.
APR listed in Seoul in 2024 and paid no dividend in FY2023 or FY2024. In FY2025 it paid ₩134.4bn. On 16 July 2026 the board declared an interim dividend of ₩2,500 a share, ₩93.6bn in total on 37,438,155 eligible shares, with payment on 31 August.
The company has also been buying back stock — FY2025 financing outflows of ₩181.7bn against ₩134.4bn of dividends leaves roughly ₩47bn unaccounted for, and the share count implies treasury holdings of around 1.5m shares against 38.9m issued.
For a business compounding revenue at 134%, distributing cash is a choice worth interrogating. The generous reading is that the model genuinely needs no capital, so returning it is correct. The less generous reading is that a company whose first-half operating cash flow covered only a third of its net income is distributing profit it has not yet collected.
Both can be true. What it definitely signals is a management team that does not see reinvestment opportunities requiring the money — which for a business with 1% capital intensity is entirely consistent.
Take the other side. APR's original proposition was not a cream. It was a home skincare device, sold once, that creates a recurring reason to buy the consumables designed to work with it. If that mechanism holds, the customer relationship is not rented from an advertising platform. It sits inside a piece of hardware the customer already owns.
That is a genuinely different competitive position from a brand that lives on marketplace search results, and it would justify returns well above a normal consumer company's for a long time. Fidelity Management and Research has filed substantial shareholding reports twice in three months, which suggests at least one large foreign investor has done the work and concluded something along these lines.
The financial evidence is also better than the sceptic's case allows. A company that improved its operating expense ratio by ten percentage points between FY2022 and FY2025 while growing revenue nearly fourfold was not buying that growth at deteriorating prices. It was getting better at it.
And 79% gross margin is not a marketing artefact. It reflects genuine pricing power on a product people choose to repurchase.
Three quarters of the operating expense ratio. If it settles back below 52%, the acquisition machine is still improving with scale and the returns are defensible. Two more quarters above 54% would say the cost of growth is rising, and at these multiples that repricing happens quickly.
Second, whatever the company discloses about device attachment — how many device owners buy consumables, and how often. That is the number that separates a razor-and-blades business from a cosmetics brand with a gadget. It is not in the summary financials and it should be in an investor presentation.
Third, the FY2026 dividend total. If it lands near ₩200bn against a year of roughly ₩500bn in net income, APR is running a 40% payout while growing over 100%. That is either extraordinary confidence in the model, or the wrong use of capital while working capital is absorbing everything the business earns.
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.