The shares trade at ₩4,955, which values PROTIA (KOSDAQ: 303360, 프로티아) at ₩63.8bn, about $45m.

Everything below is measured over the twelve months to 30 June 2026, the latest filed numbers.

• Revenue ₩17.7bn. The June half grew 40%, the year before that 44%

• Operating profit ₩4.7bn, a 26.3% margin

• No borrowings, and ₩13.9bn of cash, 22% of the market value

• 10.7x EV/EBIT on the last twelve months, 8.3x on this year’s estimate

Peter Lynch had a name for this shape: a fast grower. Small, growing 20 to 25% or better, in an industry nobody follows, with the P/E below the growth rate. Here it is about 14 against 40. He also said fast growers are where investors lose the most money when the story breaks.

What PROTIA actually does

Your doctor wants to know what you are allergic to. The old way is the skin prick test: a grid marked on your forearm, a drop of each allergen, a prick through the skin, twenty minutes of waiting to see which spots swell.

PROTIA sells the newer way, and from the patient’s side there is almost nothing to it. A nurse draws blood from a vein, sends the tube off to a laboratory, and you are out of the room in two minutes.

In the lab it looks like this.

The grey plastic sticks are the test. Inside each is a small piece of paper with allergens printed on it as thin lines: one line cat, one dust mite, one peanut. A technician drips in a little of the patient’s blood serum. If you are allergic to cats, your antibodies grab the cat line and stay there. After a few washes the lines you react to go dark, and the machine on the right measures how dark each one went.

Subscribe now

So the test is worth as many allergens as you can fit on that paper. And you cannot use a bigger piece: the strip must slide into the reader and sit in a standard tray, sized by equipment the industry already owns. Everyone is competing inside the same few square centimetres.

Most strips print one column of lines and hold 20 to 60 allergens. PROTIA prints two columns side by side, which you can see in the photo, and since March 2025 its newest kit prints four.

The competition is real, it just sits either side of PROTIA rather than on top of it. Rival line blots, including the one Revvity sells, fit around 50 allergens on a strip. PROTIA fits 176 on the same piece of paper. That is the patent doing its work.

Above them is better technology. An Austrian chip test called ALEX reads close to 300 allergens and beats anything PROTIA makes on breadth, but it costs about $138 a test in reagent alone, German statutory insurance will not pay for it, and NHS laboratories take requests for it from specialists only.

So PROTIA does not make the best allergy test in the world. It makes the densest one that fits the cheap, reimbursed, everyday slot, and that is where the routine volume sits.

The capsule machine

The strips only work in a PROTIA reader, and a PROTIA reader only reads PROTIA strips. A lab that buys the machine is signing up for PROTIA's consumables for the next decade. The reader sells once. The strips sell every week for ten years.

Five kit families run on those same three readers, so every new product walks straight into machines that are already installed and paid for. Kiwoom Securities puts the installed base at nine of Korea’s ten largest screening centres and contract labs. And because a reader has to be sitting on the bench before a single strip can be sold, the instrument line is the best forward indicator this business gives you. It jumped 82% in 2025 and now makes up 12% of sales.

The numbers

Two warnings on that table.

The reported P/E is flattered by tax. In late 2025 the company booked ₩342m of tax credits, which turned the year’s tax charge negative. That is why net profit of ₩5.5bn is higher than operating profit of ₩4.7bn. Screeners show a P/E of 12; on a normal tax rate it is nearer 14. The credits are gone, so operating profit grows about 62% this year while reported net profit grows about 30%.

Profits are not becoming cash. Net income from 2024 to June 2026 adds up to ₩9.4bn. Cash went up ₩0.9bn. Inventory swallowed ₩4.6bn of the difference and property took another ₩3.6bn, and in the June half the business threw off just ₩0.4bn of operating cash against ₩1.8bn a year earlier.

Why the margin moved

PROTIA pays for a set of things that barely move with sales: the lab, the quality team, the regulatory staff, the building. That bill came to ₩5bn when sales were ₩8bn, and today it is ₩6.3bn on sales of ₩17.7bn, so it has grown by a quarter while the top line has more than doubled.

Every ₩100 of sales leaves about ₩60 once the product is made, which means the company has to ship roughly ₩10bn of kits before it has covered that bill at all. Everything above the line is close to pure profit.

In 2023 sales landed only ₩331m above the line and the profit almost disappeared. Three new products started production at once with poor yields, the listing cost money, and Russian demand collapsed. Profit fell 84%.

Over the last twelve months sales landed ₩7.6bn above the line.

Nothing about the company changed. It just moved further from the line. That is the whole story of a margin going from 2% to 26%, and it cuts both ways: a bad year here does not trim the margin, it deletes it.

And the bill is already sized for a much bigger company than this one, so most of the next few years of growth should fall straight through to profit.

Where the growth comes from

Exports were under a fifth of sales in 2021 and are 49% today. Asia alone went from ₩0.8bn to ₩5.4bn, and the distributor network from 54 countries to about 80. Even after all that, global share is still roughly 1%, in a market GMI sizes at $6.1bn today and $13.4bn by 2034.

The surprise in the June half is that home grew faster. Korea was up 56% and exports 26%. Korea finished converting from skin-prick to blood testing years ago, so that growth is share taken from the one real rival, a spun-out LG Chem division called Invitros. It comes concentrated, though: two unnamed customers are 33% of revenue, up from 29% a year ago, and one of them alone supplied a third of the growth in the half.

Two more products ride the same readers. Animal allergy testing grew over 60% in the first half and is 7% of sales, and the canine panel is the only one approved in Korea. Antibiotic susceptibility testing cuts a three-day lab test to under a day, which matters in sepsis.

The single house covering the stock, Kiwoom Securities, expects ₩21.7bn of revenue this year (+44%) and ₩6.02bn of operating profit (+62%), a 28% margin, and points to the European CE IVDR approvals for the 192D and the antibiotic products, due at the end of this year, as the driver for 2027. Kiwoom ran PROTIA’s listing, so this is not independent research, and the first half grew 40%, which means the second half has to speed up to 46%.

In March the chief executive went further and said the company will reach ₩100bn of revenue by 2030. Discount that heavily. In November 2024 he said 2025 would break ₩20bn and it came in at ₩15.1bn. The listing prospectus had promised ₩34.6bn for the same year. The targets are marketing. The delivered numbers, 44% growth in 2025 and 40% in the June half, are the part worth trusting.

The red flags

They raised money for machines and bought property. The 2023 listing raised ₩7bn for production equipment. Three years on, ₩3.6bn of it has gone into buildings and ₩429m into machinery, with ₩2.7bn still sitting in term deposits. Separately the company spent ₩1.79bn on a building it does not use, booked as investment property and rented out for ₩18.9m a year. That is a 1.1% yield. Four of the six board meetings in 2025 had a property purchase on the agenda - none had capacity. One line sums it up. PROTIA owns ₩8.5bn of land and buildings and ₩267m of production machinery. That small amount of machinery turns out ₩17.7bn of kits a year.

The warehouse keeps filling up. Inventory has gone from 235 days of sales in 2023 to 304, then 321, and now 386 days. It grew another ₩2.1bn in the June half alone. Inventory valuation has been the auditor’s one flagged concern for three years running, and it is the reason the profits have not turned into cash. One detail I found: the fire policy covers ₩5.07bn against ₩10.8bn of stock and equipment.

There is an innocent reading of all this: 123 people, a three-person finance team, no investor relations. Sloppiness is likelier than anything worse. It is still what the discount pays for, and with insiders on 42.6% the free float is small enough that it can persist.

The verdict

The moat is real. A patent in six countries, extra allergens that cost almost nothing to add, and readers in nine of Korea’s ten biggest labs that take nothing but PROTIA strips. Add 1% global share in a market heading for $13bn and you have a fast grower with a real edge, at eleven times trailing operating profit and eight times this year’s estimate.

What comes with it is a management team that raised money for production machinery and spent most of it on buildings.

One thing on the other side. The 2023 listing priced at ₩4,500, which was below the strike price of nine separate employee option grants. The people who set those prices thought the company was worth a good deal more.

So: a business with a real edge, run by people who would rather buy buildings than expand the factory, at eleven times operating profit. Even if this year misses badly, the multiple barely moves. That is what a cheap stock looks like. The edge is what you are buying. The management is what the discount pays for.

This is the research I do: Korean and Japanese small caps that almost nobody writes about in English. If this one was worth your time, i-Scream Media and Hyundai Corporation Holdings get the same treatment.

Sources. PROTIA / ProteomeTech (프로티아, KOSDAQ: 303360) filings on DART: annual reports FY2022 to FY2025 and the H1 2026 semi-annual report (12 Aug 2026). All twelve-month figures are FY2025 plus the June 2026 half minus the June 2025 half, from note 26 (revenue by region and cost of sales), note 11 (inventory) and note 12 (property). Use of listing proceeds as at 30 June 2026 from section X of the semi-annual report. Patent and competitor workarounds from the FY2025 annual report, section II. Kiwoom Securities estimates, market share and the CE IVDR timing from the 16 June 2026 note via Infostock Daily and Seoul Economic Daily. Animal segment growth and the 176-allergen count from the company release of 15 July 2026 and Electronic Times, 31 March 2026. CEO revenue targets from Pharm Edaily, 13 November 2024, and Electronic Times, 31 March 2026. Market data as of 27 August 2026, share price ₩4,955, ₩1,414 to the dollar.

How I work. I read the Korean filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

Two companies here share almost the same name, so let me fix the labels first.

Holdco is Hyundai Corporation Holdings (227840). That is the one I am writing about.

Tradeco is Hyundai Corporation (KOSPI: 011760, 현대코퍼레이션), the general trading house. Holdco owns 26.38% of it.

Slightly less than one year of the brand division’s operating profit, for the brand division. And the frozen meat, the mushroom farms, the packaging plants, the property fund and ₩47bn of other investments arrive attached to it, as a gift.

I like the asset for two reasons.

It lasts, and it grows. The licence runs to 2049 and the rent on it was paid in full in 2019. The fees have grown 12% a year since 2020 and the profit on them 15%, and the money comes from independent dealers and manufacturers across the Americas, Europe and Asia, none of whom has anything else to put on the box.

It needs almost no capital. ₩37.8bn of licence fees came in over the last 12 months and ₩26.0bn survived as operating profit, sixty-nine won in every hundred. The licence itself uses none of the ₩96.9bn the group has tied up in inventory, farms and factories, so that profit arrives as cash.

Neither of those is in the price. Value every part fairly and I get ₩62,983 a share. The stock is ₩13,430.

All figures are LTM, meaning the year to June 2026: full-year 2025, less the first half of 2025, plus the first half of 2026. Everything comes from filings on DART.

Holdco owns a quarter of Tradeco, so it must count a quarter of Tradeco’s profit as its own, even though the cash mostly stays at Tradeco. And when it bought more Tradeco shares in the second quarter at roughly half their book value, the rules made it count that gain as profit straight away. No money arrived either time, which is why reported profit is no use here. Everything below uses operating profit.

What the licence actually earns

The brand division and the licence are not the same thing.

Royalty revenue is a fee. A foreign manufacturer sells its own generator or power tool under the HYUNDAI mark and pays Holdco a slice. Holdco buys nothing and ships nothing.

OEM merchandise is an ordinary trading business sitting in the same division. Holdco has generators and air conditioners built for it, buys them and sells them on. That line went from ₩2.1bn to ₩43.0bn in seven years, and it is the only reason the division’s margin looks like it is collapsing. The ₩26.0bn of divisional operating profit includes whatever the OEM line contributes, and the filings do not split the two.

Royalty revenue compounds. OEM merchandise adds revenue and almost no profit.

₩bn. Divisional operating profit tracks the royalty line, not the revenue line.

Subscribe now

How long it runs

About ₩6bn of the ₩37.8bn comes from companies inside the group, at rates written into the filings. The rest comes from outsiders, mostly in the Americas and Europe.

In 2019 the wider Hyundai family sorted out who legally owns the name. Holdco handed the mark to HD Hyundai, the shipbuilder, and took back a thirty-year licence in the same deal, paying all thirty years up front. So it collects royalties and pays no rent, and the licence ends on 31 January 2049. On the balance sheet the mark and the liability against it cancel out, which is why a business earning ₩26bn a year is carried at ₩0.1bn.

After that Holdco may use the mark directly, with no time limit. Direct use means putting the badge on your own products, not renting it out for a fee, so the fee stream has an end date and no renewal terms yet. But direct use is exactly what the OEM line already does, and that line has grown from ₩2.1bn to ₩43.0bn of revenue in seven years. What ends in 2049 is the fee, not the right to use the name.

What else Hyundai Corporation Holdings owns

Frozen meat. The worst business here and the biggest: ₩129bn of revenue in 2025 at a 0.4% operating margin, with market share in imported pork and chicken falling for three years. In the six months to June the inventory went from ₩32.9bn to ₩51.3bn and group operating cash flow turned negative, so every won the licence earned went into a freezer. Book value: ₩51.3bn.

Mushrooms. The one part with a moat. The company says Bulla Mushrooms is the only producer in Australia certified for organic button mushrooms, and that Smithy Mushrooms in Lancashire is one of two British growers that make their own substrate. Between them they supply Coles, Woolworths, Tesco, M&S and Booths, and the operating loss narrowed from ₩616m to ₩53m in the last half year. The farms and the packaging plants share ₩46.1bn of property and equipment between them, and the filings do not say how it splits.

Cardboard boxes. Two plants in Cambodia and one in Indonesia, selling to local brewers and to global shoe and clothing brands. ₩0.9bn of operating profit on ₩20.8bn of revenue in 2025, and the other half of that ₩46.1bn of plant.

The investments. A Seoul property fund, a Capstone trust, a venture vehicle, a stake in a packaging maker and a quarantine joint venture in Cambodia, ₩47bn at book. They paid Holdco about ₩0.8bn of dividends in the first half.

Return on capital at Hyundai Corporation Holdings

Leave out the investments, the cash and the tax assets, and Holdco has ₩96.9bn tied up in its businesses: ₩51.3bn of frozen meat, ₩18.0bn of receivables, ₩46.1bn of farms and packaging plants, ₩1.4bn of goodwill, less ₩32.8bn of payables.

Almost none of it belongs to the licence.

The brand division has out-earned the entire group in all eight periods

₩bn. Each bar is the brand division’s operating profit. The gold portion is what the other businesses took out of it: ₩25.5bn over eight periods, a fifth of what the whole company is worth today.

Return on capital is operating profit divided by operating capital: operating assets less trade payables and other non-interest-bearing operating liabilities. Investments are excluded because their income is reported bellow the operating line and I value them separately.

Sum of the parts: Hyundai Corporation Holdings

What the parts are worth against what the market pays

₩bn. No discount applied to any component.

The other half: Hyundai Corporation

Five years ago Tradeco was a steel broker. Steel made 42% of operating profit and auto parts 6%. Today auto parts are the largest division, steel is shrinking, and operating profit has compounded at 29.4% a year since 2020.

Operating profit by division, and what replaced steel

₩bn. A small “other” division is excluded, so the bars do not foot exactly to the totals. Profit from the Oman and Yemen LNG stakes never reaches operating profit at all: it was ₩31.6bn in 2023 and ₩34.9bn in 2022, reported below the line.

Revenue tripled and the margin on it nearly doubled

Two per cent still sounds thin, and it is. But it is twice what this business made in 2021 on three times the revenue. Tradeco spent the difference consolidating auto parts makers in Korea, China and India and a forklift distributor in Australia.

Return on operating capital is 13.3% excluding cash and 10.4% including it, about half what Holdco earns.

Tradeco earns profits but not cash. Operating cash flow was minus ₩69.9bn in 2023, plus ₩40.5bn in 2024, minus ₩294.7bn in 2025 and minus ₩142.9bn in the first half of 2026. The more it buys and sells, the more money sits with customers who have not paid yet, and that money is borrowed. Net debt is ₩847bn against ₩767bn of equity. The market is not wrong to discount that.

The stock is ₩27,050, so 47% of that. And that is the double discount: Holdco owns a quarter of Tradeco and then trades at a discount to that itself.

What management does with it

The dividend yield is 3.7%. Holdco pays out ₩4.4bn a year and earns ₩24.7bn at the operating line, so most of it stays inside. Holdco raised its dividend twice in its first three years as a listed company, from ₩200 to ₩400 to ₩500, and has not touched it since 2017. Operating profit has more than tripled since 2018. Tradeco moved once in the same nine years, from ₩600 to ₩700 in 2024.

Dividend per share, 2017 to 2025

₩ per share. Both companies have paid a dividend every year since listing, and both keep it small: Holdco is legally allowed to distribute ₩89.1bn.

Korea’s three new corporate laws, and what they mean here

Neither Holdco nor Tradeco has ever published a value-up plan, Korea’s 2024 disclosure in which a company says what it thinks it is worth and how it intends to close the gap. That is the case for the discount, and Korea has spent the past year legislating against it.

Treasury shares must now be cancelled. Holdco holds 3.07% of itself and Tradeco 9.21%. A law effective 6 March 2026 gives both until roughly September 2027 to cancel them or ask shareholders for permission to keep them. Either way, both boards have to raise the subject of returning capital for the first time.

The controlling family’s vote is capped at 3%. When shareholders elect the audit committee, the family and its allies now share a single 3% vote between them instead of 3% each. Only 28.8% of eligible shares voted on that item in 2026, and 47.6% of the company sits outside the control bloc. March 2027 is the first meeting under the new rule.

Small shareholders can now concentrate their votes. Companies with more than ₩2 trillion of assets can no longer opt out of cumulative voting, which lets a minority put every vote behind one board candidate instead of spreading them across all the seats. Holdco is too small to be caught. Tradeco is not.

All three have passed. What I do not know is whether these boards do the minimum or take the hint.

What could break it

Working capital. The frozen meat inventory is the live one. It absorbed every won of first-half operating profit and more. Watch the third quarter.

Capital allocation. The licence generates cash and the board keeps finding somewhere to put it that is not the shareholders. Nothing stops the next ₩50bn going into another business like the meat one.

The control bloc. 49.34% held together by an eleven-year-old voting agreement. Nothing happens if they do not want it to.

Tradeco’s balance sheet. ₩847bn of net debt funding a receivables book that grew 56% in six months. If one large counterparty fails, a 2% margin does not absorb it.

What would make me wrong. Royalty revenue stalling for two years, a write-down on the meat inventory, or the March 2027 meeting passing with nothing changed and the discount still where it is in 2028.

What it comes down to

The best business in this company earned ₩26.0bn in the last 12 months and sits on the balance sheet at ₩0.1bn.

The stock is at 21% of what the parts are worth. It does not have to reach 100% for this to work. It only has to stop being 21%.

Sources. Hyundai Corporation Holdings (현대코퍼레이션홀딩스, KOSPI: 227840) and Hyundai Corporation (현대코퍼레이션, KOSPI: 011760) filings on DART: annual reports for 2020 through 2025, the half-year reports to June 2026, the corporate governance report of 1 June 2026, the shareholders’ meeting documents of February and March 2026 and the large-holding disclosures. Korean Commercial Act amendments of July 2025, August 2025 and February 2026. Market data as of 18 August 2026, at ₩1,416.47 to the dollar.

How I work. I read the Korean filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

Two things make it interesting. It holds more cash than it is worth. And the accumulated deficit that has kept its dividend line blank since 2010 is finally gone.

This is the third in a series. The first two were Korean. I-Scream Media (KOSDAQ: 461300) sells lesson software to 93 percent of Korean primary school teachers. IB Kimyoung (KOSDAQ: 339950) earns a 150 percent return on capital and yields 15 percent. Both are small, both are ignored, and both sell something unglamorous to people who need a qualification. This one is Japanese and quieter than either.

Subscribe now

What KIYO Learning sells

Japan takes professional licences seriously. About 550,000 people sit the bookkeeping exams each year. In 2025, 245,000 sat the real estate broker exam and 18.7 percent passed. Another 43,421 sat the labour consultant exam and only 2,376 got through. These exams gate careers, and people pay for them out of their own pocket.

The old way to prepare is a classroom school. TAC, LEC and Ohara charge around 200,000 yen. KIYO charges 59,400. It has no classrooms, no branches and no sales floor. You get video cut into three minute chapters, a large question bank, and an AI layer that plans your revision and predicts your score. The state runs the exam and does not care where you studied, so the cheap course wins on arithmetic.

Students pay the whole fee up front, and the accounting rules make KIYO spread that money across the months they keep access. Cash in January, revenue all year. That is why the company reports a loss every spring while the bank balance climbs.

The asset play: KIYO trades below its net cash

Peter Lynch sorted stocks into six kinds. The rare ones belong to two kinds at once, and this is one of those. Start with the balance sheet.

At 550 yen a share the company is worth 3,778 million yen. At the end of March 2026 it held 3,606 million of cash against 450 million of debt. So the business itself costs you 622 million.

Now the fair objection. KIYO sits on 2,517 million of tuition it has been paid but not yet earned, and it still owes those students a service. So price the service. Cost of sales is 595 million of outsourced content production, 115 million of studio staff and 94 million of communications. The first two build new courses. Only the last, 1.9 percent of revenue, is what it costs to let a student who already paid press play on a video that already exists. Apply that to the prepaid balance and you get 47 million. Call it 60.

So you pay 682 million for a business that earned 304 million last year and is guided to earn 400 this year. It gets better if you wait.

Guidance implies 330 million of net profit. Add back depreciation, take out capex, assume the prepaid balance adds 100 million instead of last year’s 365, and allow for the leakage in the 2025 accounts. That builds about 385 million of cash. Net cash reaches roughly 4,018 million by December, or 585 yen a share.

You are paying 550. Take the cash off the price and by New Year you are paying nothing for the business.

The fast grower: 304 million of profit on 615 million of assets

A cheap balance sheet on a bad business is a trap, so look at the business.

Revenue went from 2.3 billion in 2021 to 5.0 billion last year and is guided to 5.8 this year. KIYO owns almost nothing to produce it. Fixed assets, software, content and receivables come to about 615 million yen, and on that it earned 304 million of operating profit. A 49 percent return.

It does not even fund those assets. The students do. The 2,517 million of prepaid tuition is larger than everything the company owns, so the customers supply the working capital a year early. KIYO has never asked shareholders for money because it has never needed any. Capital employed is negative, so there is no clean return on capital number to quote, but growth here is close to free and that is rarer than a high margin.

There's one catch that came up in the first quarter report to March 2026, but it doesn't necessarily mean anything bad for the business.

Management is pushing shorter courses because they turn into revenue faster. That drains the prepaid balance. It grew 145 million in the first half of 2025 and 141 million in the first half of 2024. In the first quarter of 2026 it fell by 13 million, in the quarter when the January campaign runs and exam season sign ups arrive.

Revenue rose 14.6 percent and the money coming in the door did not. Some of this year’s growth is faster recognition of cash they already had. The August half year report tells you whether that is a mix shift or a demand problem.

The guidance record

I like companies that do what they said they would do.

KIYO sets full year operating profit guidance every February and reports against it four times a year. The one miss was 2022, when it promised 150 million and delivered a loss of 183 after an advertising push that did not work.

One caveat. The 2024 target was easy, 141 against the 136 they had just made, and they beat it by half. The last two were not easy. They promised 41 percent more, then 32 percent more, and delivered the first. So I take this year’s 400 million seriously.

Here is my favourite number in the file.

In the year to December 2025, advertising went up 2.2 percent. Revenue went up 12.6 percent. The ratio fell from 47.9 percent of sales to 43.5. In the March quarter the operating loss narrowed to 251 million from 313 while sales grew.

That matters because the margin is thin. Gross margin is 83 percent, which sounds like software, but 43.5 yen of every 100 goes to advertising and the operating margin is 6 percent. This is a marketing machine with a good product. The advertising ratio is the only honest scoreboard it has, and one year is not a trend.

KIYO's cash flow, stripped of deferred revenue

In the year to December 2025 the business produced 726 million of operating cash flow and spent 104 million on capex. That looks like 622 million of free cash against a 682 million price.

But 365 million of it is the prepaid pile getting bigger, and the pile only grows while the company does. Strip it out and the business itself produced 257 million. Still a third of the price, every year.

Old losses meant KIYO paid a 3 percent tax rate last year. Guidance shows operating profit up 32 percent and net profit up only 12. That gap is the tax bill turning up, and it widens as the old losses run out.

The founder: Takayoshi Ayabe owns 40 percent

Takayoshi Ayabe founded the company, still runs it, and owns 40.08 percent, worth about 1,514 million yen. The two inside directors shared 40.55 million of pay last year and nobody earned over 100 million. He gets rich the same way you do. Visional, the group behind the recruiter BizReach, took its stake to 6.09 percent in February, and two more recruiters hold another 4.5 between them.

KIYO has never paid a dividend since it was founded in 2010. Japanese law caps payouts at distributable surplus, and the deficit from the loss making years kept that at zero. At December the deficit was 10.2 million yen against 330 million of guided profit, so the 2026 accounts should clear it and the board can then declare one without asking shareholders. For your information, they could have unlocked it years ago by reducing the capital reserve, but they didn't bother.

The annual report calls reinvestment the best return to shareholders, gives no date for a dividend, and says the company is weighing alliances, minority stakes and acquisitions. KIYO has no subsidiaries and has never bought a company.

My worry is not four billion yen sitting in a bank. It is a first acquisition, at that size, by someone who has never made one.

What KIYO Learning is worth: bear, base and bull

Three cases at the December year end.

Base is their own guidance. Revenue 5,800 million, operating profit 400 million. Five times that plus the net cash gives 876 yen, or 59 percent up.

Bull takes revenue to 6,200 million at an 8 percent margin, so 496 million of profit. Eight times gets you 1,175.

The bear case is the reason to own the thing. AI answers eat the search traffic, revenue stalls at 5,300 million and the margin falls to 1.5 percent. The business earns 80 million, which deserves no multiple, so give it none. The prepaid balance is six months of forward revenue, and in that world it unwinds rather than builds. Take 250 million out of it and cash goes backwards by 235 million over the year rather than forwards by 385.

Count the cash and nothing else and you get 495 yen a share. You are paying 550.

When I would sell

Three things would end it.

If the advertising ratio climbs while growth slows, the funnel is breaking. If the June prepaid balance is still stuck under March’s 2,517 million, the mix shift explanation is gone and this is a demand problem. And if they announce a large acquisition outside exam preparation, the asset play is over.

None of this is hidden. The business costs 682 million and is guided to earn 400. The bank account passes the market value in December. The founder owns 40 percent of it and has a chequebook in his hand. It sits in a Japanese filing that almost nobody reads, which is the only reason it is still there.

The half year report lands next Friday. That is when we find out whether the prepaid balance turned back up, and I will go through the numbers again in the chat group where we can discuss it.

Sources. KIYO Learning (KIYOラーニング, TYO: 7353) filings on EDINET: the annual report to December 2025 and the first quarter to March 2026. Course prices from studying.jp (スタディング). Exam results from the administering bodies. Market data as of 6 August 2026.

How I work. I read the Japanese filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don't disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

You would be wrong to. That 1.5% is a year stale; the real dividend is worth over 15% at ₩2,650 in July 2026. A third of the market value sits in cash. The core business earns more than 150% on the capital it uses. And still the stock is ignored, uncovered and cheap, because everyone who glances at it sees the same three things: a company controlled by its parent, a top line that looks like it is shrinking, and profits flattered by accounting. Trap, or gift? The answer is in the filings, so I read them.

The bear case on IB Kimyoung

Start with the bear case, because it is real. MegaStudy Education (KOSDAQ: 215200), the operating arm of the Megastudy group, owns 55.31% and runs the board. It is not the group holding company: that is Megastudy Co., Ltd. (KOSDAQ: 072870), one level up. The parent here runs Korea's largest Suneung cram school business and is run by the founder's younger brother. Group operating profit has swung all over the place: a loss in 2020, a peak, a trough in 2022. A fifth of last year’s profit came from a single one-off. And the company sits on cash it will not return, choosing instead to buy an office building, park money in private funds, and pour capital into a subsidiary that loses money every year. On the surface this is a controlled, erratic, cash-trapping microcap. Now here is what the surface hides.

What IB Kimyoung owns: Korea's university-transfer cram schools

IB Kimyoung runs the cram schools that prepare students for Korea’s university-transfer exam. You spend two years at a lower-ranked school, sit a hard entrance exam, and move up into the third year of a better university: you do not lose the time, you trade up. Its brand has led that market since 1977 and coached roughly 189,000 students through it. It is dominant by a distance, with 21 academies where rivals run two or three, the most-used textbooks and the most exam-takers in the country, and a share last measured near 80%. It owns the whole chain, from its own research lab to its own bookstore.

The one thing to test is durability, and it cuts both ways. This is a one-shot business: a student passes, transfers, and never comes back, so the company refills its classrooms from scratch every year, and demand rides on an exam system the government can redraw. The moat is real. It is a moat around a gate someone else controls.

IB Kimyoung's core earns 152% on the capital it uses

Now the numbers the screen buries. Split operating profit by segment and one part carries the company: the core earns about ₩20bn, the beauty arm ₩0.5bn, the IT arm loses ₩1.6bn.

It earns that on almost no capital. Students pay months before the first class, so the business runs on their money, not its own; that prepaid float reached ₩27bn by the first quarter of 2026. It leases its classrooms and owns little else. Gross margin has held above 70% for years, and cash beats profit: operating cash flow was ₩32.8bn in 2025 against ₩20.1bn of net income, and free cash flow has averaged about 120% of earnings over three years.

Measured against all its capital, return on capital is about 25%, because most of that capital is idle. Strip the surplus cash and it is 50%; count only the ₩16bn the core actually uses, and it is about 152%. Three numbers for the same business: the screener quotes the first, the last is the one that describes the economics.

And there is the problem hidden inside the strength. A business this profitable and this light cannot spend its own cash. The core needs ₩16bn and throws off ₩20bn a year, so the surplus has nowhere to go inside the business, and that is precisely how it ends up in an office building and a stock portfolio. A 150% return you cannot reinvest is a cash machine, not a compounder, and that distinction decides how you value the whole thing.

Why IB Kimyoung's revenue only looks like it is shrinking

The shrinking is an illusion too. The core grew for years, revenue from ₩56bn in 2021 to ₩84bn in 2024, profit from ₩12.7bn to ₩20bn, before slipping 3% in 2025 on the sale of a small division to the parent.

The lumpy history is the same story. The 2020 loss was a ₩3.97bn non-cash charge from listing through a shell, plus COVID; operating profit that year was fine. The 2022 trough was the loss-making academies the company had just bought, one taken over by converting unpaid debt into equity, dragging a ₩12.7bn core to ₩3.08bn at the group line. Consolidation hid the core; it was never the problem.

And it is turning up. The first quarter is always the seasonal loss, and that loss has narrowed every year to a first-ever profit in early 2026, on revenue up 6.3%.

The market behind it grows, not shrinks: the transfer quota is up 27% since 2020, a hard-won medical-school expansion lands in 2027, and an admissions reform follows in 2028. More students drop out every year, and dropouts are what feed transfers.

Where IB Kimyoung's ₩130bn of assets actually sit

So the bear case had one true point, the cash, and this is where it goes.

Of ₩130bn in assets, ₩44bn is cash and deposits, ₩10bn an office building, ₩9.4bn private real-estate funds, ₩5bn loans, and ₩3.5bn a portfolio of listed shares. The property came from the parent and another group company; ₩4bn of the loans went to a fund and ₩0.5bn to the company’s own managers. And the worst of it: the IT subsidiary had a ₩14bn hole in its equity, and in January 2026 the company put ₩14.5bn into it, the injection almost exactly the size of the hole. That is a rescue, not an investment. The same instinct shows in the share count, where 1.58 million treasury shares sit uncancelled since listing while 1.23 million new options were handed out below today’s price, diluting you for no reason the treasury could not have covered.

The dividend: ₩400 a year, near a 15% yield

Against all that, the payout has turned generous. The 2025 dividend was ₩300, ten times the year before; from July 2026 the company pays ₩100 a quarter, which if held is ₩400 a year, near a 15% yield.

It is covered. ₩400 costs about ₩17bn against ₩20bn of free cash flow, so earnings pay for it and the ₩42bn cash pile stays put: the yield and the cushion are not the same money. For now the payout is dressed as a tax-free return of capital, which lasts about a year; after that the same dividend simply becomes taxable. And notice the aligned incentive: the parent wants cash, and the cleanest way it gets cash is by making this company pay it, which every minority holder collects at the same rate.

What IB Kimyoung is worth

Value it in parts. The core earns ₩20bn of operating profit, ₩16bn after tax; at a conservative 8 to 10 times, that alone is ₩130bn to ₩160bn, above the entire ₩115bn market cap, and even at six times it roughly equals the whole company’s price. Add ₩42bn of net cash and, at a haircut, ₩25bn for the rest, and a sober sum runs past ₩200bn. Or more simply: strip out the cash and the non-core and you pay about ₩48bn for a business earning ₩20bn, under 2.5 times EV/EBIT.

The catch keeps it cheap. This is a controlled company, and the record, an office bought from the parent, cash poured into the losers, a one-off gain from selling a division to the parent at a price no outsider could test, says the controller allocates capital for itself, not for you. Good management allocates like an owner; this one allocates like a parent with other priorities.

But the case does not need the controller to reform, which is what most people miss. You are paid three ways, and only the last depends on him: the core alone, under 2.5 times EV/EBIT, a dominant franchise on a distressed multiple; a roughly 15% yield, paid from earnings, that pays you to wait; and, for free on top, ₩42bn of cash and non-core assets a buyback or a Korea Value-Up push could one day surface. If the third never comes, the first two still stand.

So it is not a compounder and not a clean buy. It is a dominant little business, cheap, paying you well to wait, wrapped in a shell whose owner has not yet worked for you. Watch three things: whether the quarterly dividend holds, whether the non-core buying stops, and whether the treasury is ever cancelled. Any one would say the parent has begun to care about the price.

Until then, the numbers say a fine business is on sale, and the narrative, controlled, shrinking, propped up, is what keeps it cheap. Reading the filings is how you tell them apart.

This is the second in a series on listed education businesses in Asia. The first was i-Scream Media (KOSDAQ: 461300), the third KIYO Learning (TYO: 7353).

Sources. IB Kimyoung (아이비김영, KOSDAQ: 339950) filings on DART: the annual report to December 2025 and the quarterly report to March 2026. Shareholding and related-party data from the same filings. University transfer quotas and admissions reform from the Ministry of Education. Market data as of 21 July 2026, at a share price of ₩2,650.

How I work. I read the Korean filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

Almost every primary-school teacher in South Korea opens its software on their screen every day. It sells textbooks that schools re-order year after year. It keeps 31 of every 100 won of sales as operating profit, has grown that profit by roughly a quarter a year for three years, earns triple-digit returns on the capital it actually uses, and pays a dividend yielding close to 10%. And you can buy the whole thing for about four times FY2025 earnings.

Do not confuse it with i-Scream Edu (KOSDAQ: 289010, 아이스크림에듀), a separate listed company that was spun out of this one in 2013 and sells home-study products.

At ₩15,860 on 11 July 2026 the entire company is worth about ₩200bn, and it's sitting on roughly ₩182bn of cash and securities, about 91% of its own market value. Strip that out and the market is pricing the actual operating business, which earns ₩62bn of operating profit a year, at only around ₩8bn, a few million dollars. On paper, that's not a valuation; it's a rounding error. So the obvious question is: what's wrong with it? The answer isn't the business, it's excellent. It's what the owners do with the cash. Let's take both apart.

Subscribe now

What i-Scream Media actually does

The heart of i-Scream Media is a piece of software called iScream S (아이스크림S), the daily toolkit a Korean elementary teacher uses to run their classroom: lesson slides, worksheets, videos, quizzes, a library of more than 6.5 million teaching materials built over twenty years. Teachers pay nothing for it. That’s the clever bit: it isn’t where the money is, it’s the front door.

Because about 93% of the country’s primary teachers already work inside this free tool every day, iScream has an enormous head start at the thing that does make money — textbooks. In 2022 it began publishing certified primary textbooks under Korea’s new curriculum and immediately won the highest adoption of the thirteen licensed publishers, around 40–58% on its maths, science and social-studies books.

Why is that such a good business? Four reasons, in plain terms:

• The state pays, but teachers choose. Schools pick the publisher; the government funds it. iScream just has to win the teacher — and it already owns the teacher’s daily routine.

• It repeats. A new class of pupils arrives every year, so schools re-order the same textbooks annually. Not a one-off sale — a recurring, predictable stream.

• It’s sticky. Once a teacher has built a whole year around a textbook and its matching materials in iScream S, switching means relearning everything. Most don’t bother.

• It’s almost impossible to copy. A rival would have to rebuild twenty years and 6.5 million pieces of content — then prise teachers off a platform they already open daily.

Put those together and you get a genuine moat — a durable edge a competitor can’t cheaply buy or copy. The loop is the moat: a free platform everyone uses, feeding a paid textbook business that recurs every year and is a pain to leave.

Two smaller businesses sit alongside. iScream Mall, an online shop for classroom supplies, owns 90%+ of education e-commerce and brings in about a quarter of revenue at healthy margins — though its profit has stopped growing. And teacher training, where iScream is the #1 private provider, is small (~6% of sales) and actually loses a little money right now as it expands into in-person courses. Neither moves the needle. The textbook engine is the story.

The numbers: 24% revenue growth at rising margins

Revenue and operating profit compounded at ~24% and ~26% a year — and margins rose while they did it.

Revenue climbed from ₩104bn to ₩196bn over three years, about 24% a year; operating profit grew even faster, ₩31bn to ₩62bn. The tell of a quality business is right there: the operating margin went up as the company grew, from 29% to 31.5%. Once the content library exists, each extra textbook costs almost nothing to make, so profit outruns sales.

And there’s one number that really seals it: how much profit iScream squeezes out of the capital it uses. On the standard measure most stock screeners use — which counts the giant cash pile as “capital” — its return on capital is around 30%, already excellent. But that understates it badly. The operating business barely needs any capital to run, so the return on the money genuinely tied up in it is far higher — well over 50%, and north of 100% on a strict reading. In plain English: it makes a lot of money from very little. That is the single clearest fingerprint of a high-quality business — and it’s exactly what you’re not being asked to pay for.

(One caveat: reported net profit is bumpier than operating profit, because the company runs an investment portfolio whose swings flow straight through the bottom line. For the clean read on the business, watch operating profit.)

How i-Scream Media became a textbook company

Publishing went from half the business to two-thirds — and is now essentially all of the growth.

Three years ago revenue was split fairly evenly between textbooks and the shop. Today publishing is two-thirds of the company and effectively all of the growth: more grades covered each year (3–6 now complete), rising adoption, and a textbook price that’s climbed from ₩8,755 in 2021 to ₩11,287 in 2025.

The upside is that publishing is the highest-margin, stickiest, best-protected leg. The trade-off is that iScream now leans on a regulated, government-funded revenue stream — so education policy can move the numbers, for better or worse.

Why you can't judge i-Scream Media on one quarter

Textbooks ship only in Q2 and Q4, so Q1 and Q3 look tiny — and Q1 is a planned seasonal loss every year.

Look only at iScream’s first quarter and you’d think it was tiny or collapsing. It’s neither — it’s the calendar. Textbooks are delivered in just two quarters, Q2 and Q4, so Q1 and Q3 are small by design and the company runs a seasonal loss in Q1 every year. Normal, not a warning.

Q4 2025 alone did ₩102bn, over half the year. Q1 2026 came in at ₩15.3bn, up ~15% on the year before, with the seasonal loss slightly smaller than a year earlier. Judge this company on full years, never on one quarter.

Free cash flow: ₩48bn on ₩196bn of sales

Free cash flow hit ₩48bn (about 25% of sales) in FY2025; the 2024 dip was a one-off content build.

Free cash flow — the cash left after running and investing in the business — was ₩48bn in FY2025, about 25% of sales. That’s strong. The dip in 2024 wasn’t a problem; it was a deliberate ₩22bn+ splurge on building an AI textbook product. Once that normalised, cash flow snapped back. iScream’s “capex” is really content spending, and it comes in lumps: a heavy build year flatters nothing, a normal year converts beautifully. FY2025 was normal.

The elephant on i-Scream Media's balance sheet

IPO cash was rotated into a securities portfolio worth ₩94bn — nearly as large as the entire operating business.

Now the part that complicates an otherwise easy story. The balance sheet is rock-solid — but look at what’s on it. iScream holds about ₩88bn of cash and ₩94bn of securities — together ₩182bn, roughly 91% of the whole ₩200bn company. The operating business that throws off all those lovely margins is a small slice of the assets; the rest is a war chest.

The trouble is that it isn’t clear that pile is really yours. It’s new, and it’s growing fast: the company raised money in its August 2024 listing and, instead of returning it or reinvesting it, poured it into an investment portfolio that has swelled from ₩33bn to ₩94bn in eighteen months. And some of those holdings are venture funds tied to the founding family — so a big slice of “your” cash is being run as an in-house fund, partly into related parties.

A note for the screener crowd. A typical data screener will show “cash & short-term investments” of about ₩111bn, not the ₩182bn here. The gap is that most of the securities (~₩70bn+) sit in long-term investments, which screens don’t count as cash. Read the actual filing and the pile — and the question hanging over it — is bigger than the screen lets on.

Korea's shrinking cohorts, and why i-Scream still has runway

Here’s the pushback you’ll hear: Korea is running out of children. It’s true. The total school-age population is about 5.07 million today and is set to fall to roughly 4.25 million by 2029 — a ~3% annual decline — and the incoming first-grade cohort dropped below 300,000 for the first time in 2026. Fewer pupils eventually means fewer elementary textbooks.

But that misses how much room iScream still has to grow. It has grown straight through the demographic decline — publishing revenue rose 45% last year while pupil numbers fell — on three engines: it keeps winning share (40–58% adoption and climbing), it keeps raising prices (the per-book price is up ~29% in four years), and Korean families keep spending more per child (national private-education spend hit a record ₩29tn in 2024, up four years running).

And the bigger point: iScream is only just leaving its home turf. Management plans to take the publishing muscle it built in primary school and move, step by step, into the middle-school textbook market — a whole new set of grades and subjects to win. It’s pushing teacher training into secondary schools (where each teacher covers a single subject, so appetite for content is higher) and into the early-childhood and daycare market, and it’s taking its AI teaching tools abroad — to Vietnam, India and Japan. None of that is proven yet. But it means the addressable market is getting bigger, not smaller — and at this price you’re paying for none of it.

The catch: control and policy

So why does a business this good trade this cheap? Two reasons, and both are real.

Control. The founding Park family controls roughly 61% of the company, partly through the listed parent SigongTech (KOSDAQ: 020710) and partly through direct personal stakes held by chairman Park Ki-seok and his son. That makes bankruptcy unthinkable, but it also means no outside investor can force that ₩180bn back to shareholders. Cash comes back only when the family chooses to send it.

Policy. iScream’s revenue rides on government decisions. The clearest reminder came in August 2025, when the government demoted the AI textbook iScream had spent heavily on from an official “textbook” to mere “educational material” — erasing the upside it had invested behind and forcing about ₩13bn of write-offs. When the state changes its mind, this business feels it.

But the catch is being actively chipped away. Under Korea’s national “Value-Up” programme, iScream now commits to paying out 40% of profit as dividends. The dividend more than doubled last year, to ₩1,554 a share — a yield of nearly 10% — and it uses only about 40% of free cash flow, so it’s well covered with room to grow. Add buybacks and total cash returned already tops 10% of the market value a year. The cash has started moving toward shareholders. The only question is how far it goes.

Putting a price on i-Scream Media

• Price: ₩15,860

• Market value (ex-treasury): ~₩200bn (~$150m)

• Price-to-earnings (FY2025): ~4.1×

• Price-to-book: ~1.0×

• Return on capital (reported / operating): ~30% / >50%

• Dividend yield: ~9.8%

• Total shareholder yield (dividends + buybacks): >10%

• Enterprise value: ~₩8bn — a few $m

• FY2025 revenue growth: +29%

• Operating margin: 31.5%

• Free cash flow / sales: ~25%

• Cash + securities: ~₩182bn (≈91% of market value)

• Founding family control: ~61%

For the record, 13,302,433 shares are in issue, of which 598,969 sit in treasury (pledged against a exchangeable bond), so the free-float market value is about ₩200bn.

The summary is almost too simple: four times earnings for a business growing ~25% a year at 31% margins, with a near-10% dividend while you wait — and the ₩62bn-a-year operating company effectively thrown in for an enterprise value of barely ₩8bn. Numbers like that usually mean a business is dying. This one is thriving.

The verdict

iScream Media is two things at once. Underneath is a wonderful business — capital-light, high-return, sticky, growing, cash-generative, and rarely this cheap. Around it is the reason it stays cheap: an owner who controls the company outright, a revenue base exposed to the government’s pen, and a demographic tide that will eventually turn.

So the cheapness isn’t a mistake — it’s the price of the catch. Whether it’s a genuine bargain turns on one question: does the cash ever really reach outside shareholders? The Value-Up dividend already says “some, yes,” and broader Korean governance reform is a slow but real tailwind. If the cash starts flowing, the re-rating could be violent. If it doesn’t, you’re still collecting a near-10% yield on an excellent business you don’t quite control — far from the worst place to sit.

Numbers, not narrative: the business is as good as the headline figures suggest, and the stock is as cheap as it looks. The only real question is who the cash belongs to.

This is the first in a series on listed education businesses in Asia. The second is IB Kimyoung (KOSDAQ: 339950), the third KIYO Learning (TYO: 7353).

Sources. i-Scream Media (아이스크림미디어, KOSDAQ: 461300) filings on DART, FY2022 to Q1 2026. School-age population projections and private-education spending from Statistics Korea. Textbook adoption and curriculum data from the Ministry of Education. Market data as of 13 July 2026.

How I work. I read the Korean filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

Value investors are trained to flinch at one setup in particular: the cyclical at the top of its cycle, throwing off record profits, wearing a single-digit P/E like a trap. Buy peak earnings at a multiple that only looks cheap until the denominator caves, and you lose money slowly. Dongsung FineTec looks like that company — and the stock, down ~46% from its ₩34,300 high to ₩18,460, is being priced like it.

I think the market has the right instinct and the wrong object. The cyclicality is real. But the thing that usually makes a peak-cycle short dangerous is sitting in the filings in plain sight: ₩2.08 trillion of order backlog, nearly three times last year’s revenue, for a product installed on a construction schedule the company neither controls nor can accelerate. You aren’t guessing at the next three years of revenue. You’re reading them off signed contracts.

Here’s the case from the numbers — then the reasons it might be wrong, because the red flags are not decorative.

The franchise: a toll booth at −163°C

Dongsung makes the insulation that holds liquefied natural gas at minus 163 degrees inside the cargo holds of LNG carriers. That’s the franchise. About 96% of revenue is one thing — polyurethane cryogenic insulation (reinforced foam, panels, membrane sheet) sold to shipyards. A small refrigerant-trading arm is the other 4%: a rounding error with some optionality bolted on.

This isn’t a commodity. The membrane containment systems (GTT’s Mark-III and NO96) are safety-critical on a thirty-year ship carrying a cryogenic, flammable cargo, so a supplier has to be qualified by the licensor and the yard — and nobody re-qualifies a vendor mid-programme to save a few percent. In Korea that’s produced a clean duopoly: Dongsung and Hankuk Carbon, with total global cold-insulation capacity under ~90 ships a year. Dongsung localised the technology first and runs ~34 ships of capacity, at roughly ₩25bn of content per large carrier.

So, plainly: Dongsung is a toll booth on the LNG construction bridge. Real barriers — qualification, switching costs, safety lock-in. No pricing power over its customers (three Korean shipbuilders that are its order book) or over GTT (which keeps the IP rent). Good economics while traffic is heavy; no leverage over whoever sets the toll. That tension is the entire bull-and-bear axis.

Dongsung FineTec's operating leverage: a 7× jump in earnings on 1.7× revenue

Three years ago this was a low-margin parts maker: ₩434bn of revenue, ₩8bn of profit. FY2025 closed at ₩742bn of revenue (+24%), ₩72.7bn of operating profit, and ₩56.2bn of net income (+43%) — roughly a 7× jump in earnings on a 1.7× rise in revenue, operating margin up from 3.5% to 9.8%, ROE through 25%. The latest quarter shows where that leverage comes from, in concentrated form.

Q1 2026: a price story in a volume costume

Q1 revenue was ₩169bn — down 1.9% year on year. Yet gross margin went from 13.9% to 18.8%, operating profit rose 58%, and net profit rose 90% to ₩17.1bn. You don’t get that from working harder. You get it because what you sell is worth more and what you buy to make it costs less.

The filing’s price tables draw it cleanly. The average selling price of cryogenic insulation — quoted per cubic metre of carrier — went from ₩941,000 (FY2024) to ₩1,008,000 (FY2025) to ₩1,248,000 in Q1 2026, up a third in fifteen months. Over the same window the per-kilo cost of membrane raw material fell, from ₩2,524 to ₩2,382. Prices up, costs down.

Two mechanics underneath, and they decide how durable this is. First, the revenue lag is a feature: insulation is fitted late in the build, so Dongsung books revenue three to four quarters after a yard cuts steel. The ships being insulated now were ordered in 2023–24, at higher prices than the 2021–22 vintage that once depressed margins — so the richest contracts are still flowing through, and management can see its own average selling price two years out. Second, the cost tailwind is luck, and should be treated as luck: ~60% of cost is raw material, mostly MDI and steel membrane, helped by a weak won on dollar-priced sales. Both mean-revert. Underwrite ~9–10% normalised operating margins, not Q1’s 11.7%.

What's already locked in: a ₩2.08trn backlog

This is where Dongsung stops behaving like a normal cyclical. At 31 March 2026 the cryogenic-insulation backlog stood at ₩2,076bn — about 2.8× FY2025 revenue, three years of contracted work for HD Hyundai, Samsung Heavy, and Hanwha Ocean. The balance sheet agrees: inventory built to ₩178bn, and ₩112bn of customer advances sit in current liabilities against that book.

And the macro filling it from here isn't a hope. Deliveries peak in 2026–2028, exactly Dongsung's window, because it gets paid when ships are built, not ordered. New ordering did pause in 2025, the cliff the bears point at, but Qatar's fleet renewal is live and US liquefaction capacity keeps reaching final investment decision. Those refill the years after the current backlog.

Valuing Dongsung FineTec: earnings up, multiple down

At ₩18,460 on June 22 2026, the market cap is ~₩554bn on 30.0m shares. Against that:

• ~8.6× trailing earnings, on ~₩2,140 of last-twelve-month EPS (about 10× on FY2025’s reported ₩1,879).

• P/B ~2.1× on a 25% ROE — a touch over two times book for a business earning a quarter on it.

• Net cash ~₩90bn — ~₩79bn cash plus ~₩11bn of short-term instruments against zero bank debt (the ₩7bn of short-term borrowings was repaid in the quarter). Strip it out and EV is ~₩460bn, putting EV/EBIT near 6×.

• A ~1.8% dividend on a modest ~25% payout — room to grow, not a thesis.

Now set that against the company’s own history: 12–13× through 2023–24, north of 18× at the last peak, a trough-cycle average around 16–17×. It trades below its own cyclical floor while earnings still rise and three years of revenue are contracted. The de-rating has run ahead of any deterioration in the actual numbers. That gap is the opportunity.

The bear case on Dongsung FineTec

A cheap multiple on a peak cyclical is a trap until proven otherwise. The bear case, as strongly as I can put it:

One product, one cycle. ~96% of revenue is LNG-carrier insulation; there’s no second leg. The backlog that looks like a fortress empties on a schedule, and the real question isn’t 2026–28 earnings (largely locked) but normalised earnings power after it delivers — which depends on the 2026–27 ordering wave actually arriving.

Price-taker to three customers. HD Hyundai, Samsung Heavy, and Hanwha Ocean are the order book, they dwarf Dongsung, and they granted no price increases in 2021–22. Today’s pricing is a tight market, not structural leverage.

GTT keeps the crown jewels. Dongsung is a qualified manufacturer, not the IP owner — which caps how much of the chain it can ever capture.

China is the long shadow. Chinese yards are taking LNG share and don’t buy Korean insulation; Korea’s 2025 share recovery owed more to US port-fee actions stalling Chinese orders than to any moat of Dongsung’s.

The balance sheet flatters both ways. That ₩90bn of net cash is partly the flip side of ₩112bn in customer advances; as the book delivers and isn’t fully replaced, advances unwind and the cash normalises down. And the parent, Dongsung Chemical, owns ~39% — capital allocation will be made with its interests in view.

The optionality on the other side — DNV approval-in-principle for ammonia-fuelled tanks, liquid-hydrogen storage patents, refrigerant recycling into Korea’s 2027 mandate, a possible Alaska LNG tailwind — is real and free in the price, but immaterial today. Buy this for the LNG cycle and the balance sheet; treat the rest as a call option you didn’t pay for.

The verdict

Strip it to the frame. A niche manufacturer with a narrow-but-real moat, 25% ROE, no net debt, and three years of contracted revenue — at under 9× earnings, below its own cyclical-trough multiple, because the market has decided the cycle is over. The bears say it is over and you’re buying flattered, single-product earnings from a price-taker. The bulls say the 2026–28 delivery peak and the next ordering wave stretch the runway far enough that today’s multiple re-rates as visibility proves out.

Both can be true in sequence — a couple of very good, largely banked years, then a genuine question mark. That makes Dongsung less a forever-holding than a mispriced cyclical with an unusually long fuse: the rare case where the order book lets you price the next three years instead of guessing them. Size it as a cyclical, demand your margin of safety against normalised — not peak — earnings, and decide now what you’ll do when the backlog stops growing. Because the day that headline first prints lower, the multiple will move long before the revenue does.

Numbers, not narrative. The numbers are good. The narrative — peak, over, done — is what the price is selling you. Whether it’s right is the only question that matters, and the backlog buys you three years to answer it.

I have covered several other overlooked Korean and Japanese micro caps, including i-Scream Media (KOSDAQ: 461300), Total Soft Bank (KOSDAQ: 045340) and KIYO Learning (TYO: 7353).

Sources. Dongsung FineTec (동성화인텍, KOSDAQ: 033500) filings on DART: the FY2025 audited statements filed 19 March 2026 and the Q1 2026 report filed 14 May 2026, including the segment, selling-price and raw-material cost tables. Market data as of 22 June 2026, at a share price of ₩18,460. Won conversions at roughly ₩1,500 to the dollar.

How I work. I read the Korean filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

You wouldn’t know that from the number most screeners show you. Greens closes its books at the end of June, so the first three quarters run from July 2025 to March 2026. Through those nine months revenue is up 11.5% and operating profit is up nearly 33%. By every measure of how the actual business is doing, things are going extremely well. And yet reported net income is down 2.3%.

That gap is the whole story. It’s a moment where accounting tells you something that has almost nothing to do with the business — and reading past the top line pays off.

Who Greens Co. actually is

Greens was founded in the 1950s and is still effectively controlled by its founding family: the top two holders (Shinryoku Co. at ~18% and TM Co. at ~16%) plus president Yuya Muraki personally (~5%) form a commanding block. This is a regionally rooted, owner-operated business that happens to be listed.

What lifts it above the average regional hotelier is one contract. Greens has been the exclusive master franchisee for Choice Hotels International (NYSE: CHH) in Japan since 2003, the American giant behind Comfort Inn, Comfort Suites, Quality Inn and the Ascend Hotel Collection. In Japan, the whole Comfort ecosystem runs through Greens.

The portfolio splits cleanly into two:

• Choice Brand (チョイスブランド) — limited-service, accommodation-focused hotels, mostly Comfort Hotels clustered near major railway stations and built for efficiency. This is the engine: roughly ¥24.8bn of ¥28.5bn in first-half revenue, about 87% of the business.

• Original Brand, marketed as Greens Hotels (グリーンズホテルズ) — full-service regional hotels with banquet halls, restaurants and wedding facilities, concentrated in Mie and the wider Tokai region, under names like Hotel Econo and Green Hotel. Smaller and lower-margin (a banquet takes far more labour than a check-in desk), but it’s the sticky, community-driven local anchor.

As of the December 2025 interim, the group ran about 120 hotels and ~16,900 rooms.

Greens Co.'s operating numbers: occupancy and ADR both rising

Strip out the accounting noise and what’s left is a textbook recovery firing on the two cylinders that matter in hospitality: occupancy and rate.

Across the most recent quarters, group occupancy and average daily rate (ADR) beat the prior year in every single month — and crucially, they’re doing both at once. They aren’t trading price for volume; they’re getting more of each. For the first half, group occupancy hit 82.5% (+2.1 points) with ADR at ¥10,724 (+7.0%).

That ADR line is the one I’d stare at. Rising rates alongside rising occupancy is the signature of real pricing power — demand outrunning supply — not a discount-driven fill.

By brand for the first half: Choice Brand grew revenue +13.1% (occupancy 83.2%, ADR ¥11,203); Original Brand +10.6% (occupancy 78.6%, ADR ¥7,919), with its operating profit up nearly 75% on the back of better rates. And they kept investing into the strength, opening two new-build hotels: Comfort Hotel Mito (Ibaraki, November 2025) and Comfort Hotel ERA Sapporo Kitaguchi (Hokkaido, December 2025).

Why Greens Co.'s net income is falling while operating profit rises

Taxes, specifically the end of a tax holiday.

Like many Japanese hotel companies, Greens went through a brutal pandemic stretch and built up loss carryforwards: past losses that shelter future profits from corporate income tax. For the last few years it has been profitable while paying almost nothing in tax, because those old losses absorbed the bill. In the prior-year nine-month period the tax line was negligible.

This year the carryforwards ran out, and the tax line jumped by well over a billion yen. That single swing is the entire reason a ~33% rise in operating profit becomes a 2.3% decline in net income. It’s the same mechanism behind the full-year guidance, which still calls for net income down 31.6% even as revenue grows to ¥53.2bn.

Be clear about what this isn’t: not margin compression, not weak demand, not a cost blowout. It’s the business graduating to a normal tax rate — painful optically, neutral-to-healthy in substance. A company that pays full tax is usually one that has finished cleaning up its past. (And notice the year-on-year drag is already easing: net income was down 8% at the half, down only 2.3% by the nine-month mark.)

The detail that jumps out: nine-month profit above the full-year forecast

Here’s the kicker. Net income for the first nine months already came in at ¥4.02bn — comfortably above the company’s ¥3.6bn full-year forecast. In other words, on the official numbers the firm is guiding to a small net loss in the final quarter.

There are a couple of innocent explanations — interim tax is estimated using a full-year effective rate, so the heavier tax true-up can land late in the year, and management tends to be conservative. But either way, it’s a flashing signal: this is a business whose underlying earnings are running well ahead of its own published expectations. Watch for a guidance upgrade.

Greens Co.'s balance sheet and the DBJ preferred shares

The clean-up shows elsewhere too. Total assets have grown to ¥37.3bn (from ¥29.4bn a year earlier) as the new hotels come online, while the equity ratio improved to 36.8% from 34.6%. First-half operating cash flow was a healthy ¥5.5bn; the company is repaying long-term debt and still raising its dividend, from ¥35 to a forecast ¥40 per share.

One artifact of the COVID era remains on the cap table: a small slug of Class A preferred shares held by the DBJ Food & Lodging Support Fund, the state-backed Development Bank of Japan’s rescue vehicle. They carry a 4.0% preferred dividend, and under the agreement DBJ generally can’t force redemption until June 30, 2028. A modest overhang — and a reminder of how close the sector came to the edge — but increasingly one the company looks set to simply grow out of.

The cloud on the horizon: China

No Japanese tourism story is complete without the inbound question. Calendar 2025 was a record 42.68 million foreign visitors. But the back half of the calendar year wobbled: nationwide accommodation nights fell year-on-year in both November and December 2025, driven by a sharp drop in Chinese travelers amid worsening Japan–China relations.

Here Greens has an unusually comfortable seat. Management puts Chinese guests at only about 4% of its total — far below the inbound market overall, where Chinese visitors have historically been close to a fifth of arrivals. (Outside analysts peg China-linked revenue a touch higher, around 6–8%, and note management has been shifting marketing spend toward North American and European source markets.) Just as important, roughly half of Greens’ demand comes from domestic business travelers — the steady, off-peak, recession-resistant base that doesn’t care about diplomatic spats and keeps the station-side Comfort Hotels ticking over regardless of inbound leisure swings.

That mix — a domestic, business-traveler floor under an inbound-leisure upside — is arguably the most attractive thing about the model, and exactly what should cushion the China softness likely to surface in the coming prints.

What I’m watching

1. Does the ADR line keep rising? Pricing power is the thesis. The day occupancy gains start coming only from discounting is the day this gets less interesting.

2. A guidance upgrade. Nine-month profit is already above the full-year forecast. Either Q4 carries a real tax true-up, or the guide gets raised.

3. How hard China actually bites. The ~4% self-reported exposure gets tested in the upcoming numbers; a small, manageable dent confirms the domestic-floor story.

4. The 2028 preferred-share clock. An early redemption of the DBJ preferreds would be a clean signal that the COVID chapter is formally closed.

Greens does one thing — running mid-market hotels near train stations — and is doing it well into a strong demand cycle. The bottom line is temporarily misleading. The business underneath it is not.

I have covered several other overlooked Japanese and Korean micro caps, including KIYO Learning (TYO: 7353), i-Scream Media (KOSDAQ: 461300) and Total Soft Bank (KOSDAQ: 045340).

Sources. Greens Co. (株式会社グリーンズ, TSE & Nagoya: 6547) filings on EDINET and the company’s results releases for the first quarter, interim and third quarter of the year to June 2026. Inbound visitor and accommodation data from the Japan National Tourism Organization and the Japan Tourism Agency. Market data as of 5 June 2026.

How I work. I read the Korean and Japanese filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

That reflex is the whole opportunity. Strip the label off and read the statements, and you find a debt-free, cash-generative business that has compounded operating profit at roughly 45% a year for five years, earns a ~48% return on invested capital, and trades at about 14× earnings with a 4%-plus dividend. So one of two things is true. Either the market is right that this is a fragile pyramid about to topple — or it’s mispricing a genuinely good business because of the company it keeps. This is my attempt to settle that with numbers, not adjectives.

What Zinzino actually is

Zinzino is a Gothenburg-based, Nasdaq First North–listed direct seller founded in 2007. Its hook is genuinely differentiated for the category: test-based, personalized nutrition. You take a home blood test — the BalanceTest — which measures your omega-6 : omega-3 ratio; the company sells you the protocol to fix it, then re-tests to prove it worked. That test → sell → re-test loop is the moat management keeps pointing at, because it turns a commodity (fish oil) into a measurable, subscription-friendly outcome.

Three mechanics matter for an investor:

• Recurring revenue in disguise. Most sales run through six-month subscriptions that auto-renew until cancelled — recurring revenue dressed up as network marketing.

• Asset-light and global. Products ship from one web shop to 100-plus markets. Zinzino owns its Norwegian R&D and production units (Faun Pharma, BioActive Foods); the sales force is independent distributors it pays no salary.

• Growth by acquisition, paid in stock. It has spent recent years buying distributor networks — Valentus, Zurvita, Truvy, Bodé Pro, Sanki, ItWorks — and running them through its own product engine. The deals are funded with cash and shares, not debt, so it has scaled to billions in revenue while staying net cash.

If you want the canonical case that a well-run direct seller can be a wonderful business, remember Buffett bought The Pampered Chef in 2002. The model isn’t the problem. Execution is.

The quarter Zinzino's profitability inflected: Q1 2026

The Q1 2026 report (released 22 May 2026) is where operating leverage stops being a thesis and becomes arithmetic. Revenue grew 27% — but every line beneath it grew faster.

The margin stack tells the same story from the other side. Gross margin jumped 6.1 points to 37.0%, the EBITDA margin 4.5 points to 15.4%, and the net margin to 11.3%.

The balance sheet followed the income statement. Cash rose to SEK 889M, the equity ratio to 50.4% (from 34.6%), and book value per share to SEK 31. Net profit cleared SEK 100M for the first time, and operating cash flow went from a rounding error (SEK 21M) to SEK 126M — roughly six times higher. The CEO called it a “record quarter”; for once the press-release adjective survives contact with the numbers.

When the bottom line grows three times faster than the top, you’re watching operating leverage, not spin.

The catch you have to respect: May was soft

#NumbersNotNarrative cuts both ways, so here’s the other side of the ledger. The preliminary May sales report (3 June 2026) showed group revenue up just 14% to SEK 316M, pulling January–May growth down to +23%. After +27% in Q1, that’s a visible deceleration.

Three things to hold in your head at once:

• FX is masking the real number. May’s 14% is reported in SEK, and the krona has strengthened. In Q1 the gap between reported (+27%) and local-currency (+37%) growth was a full ten points. May’s local-currency figure was almost certainly well above 14% — Zinzino just doesn’t publish it monthly.

• The comparables are stiffening. Q2 2025 was already a SEK 794M quarter; the easy year-ago layups are gone.

• It still beats the plan. Management’s 2026–2028 target is ≥20% sales growth at a >11% EBITDA margin. The thesis never needed 50% growth — it needs ~20% at expanding margins. That is exactly what’s on the board.

Zinzino's growth is becoming a geography story

Break the year-to-date number into regions and the single most important picture in this whole post appears: a two-speed business.

Carrying the load: North America +58% — the engine room, now home to ItWorks, Sanki, Truvy, Bodé Pro and Zurvita; the region jumped from 15% to 21% of group revenue. South America +379% off a tiny base (Peru launched in February as the continent’s first official market, Colombia next). Central Europe +39% — the most reassuring line, because it’s broad-based organic strength, not a bolt-on.

Going backwards: Eastern Europe −12%, Asia-Pacific and the Baltics −1%, the Nordics flat. India cratered inside APAC; Turkey — where a large distributor organization walked in 2025 — gutted an otherwise strong Southern & Western Europe.

The honest bear read: a growing share of the headline is bought, not grown. The bull read: that is the strategy — buy distributor networks cheaply, run them through the test-based engine, harvest the margin — and Q1’s margin expansion while integrating ItWorks is the receipt that the playbook converts. Both are true. Which one dominates over the next four quarters is the real debate.

Acquisitions, and the dilution bill

Funding deals with stock keeps the balance sheet pristine and quietly dilutes you. In Q1 alone, ItWorks (26 Jan) cost SEK 291M entirely in new shares (1.84M Class B at SEK 145.62), contributed SEK 69M of revenue in about two months, and is guided to add >USD 45M in 2026; Sanki (effective 1 Jan) cost SEK 75M, mostly stock, and was the key that opened South America; Bodé Pro added another ~70,000 shares.

Share count has drifted up to roughly 38.9M, and warrant holders keep nudging it higher. If every remaining warrant were exercised, that’s ~1.79M new Class B shares — about 4.6% dilution — struck between SEK 56 and SEK 271.60. The good news for per-share math: the big SEK 271.60 programs running to 2030 are deeply out of the money at ~SEK 135, so the most dilutive tranches only bite if the stock roughly doubles. When you build owner earnings, charge yourself the economic cost of that gap — but note it shrinks as the quote falls.

My own back-of-envelope owner earnings — normalizing out 2025’s working-capital tailwind, maintenance capex, warrant dilution and maintenance M&A — lands near SEK 366M. Against ~20% growth, that’s a PEG comfortably under 1. Cheap, for this quality.

Valuing Zinzino: priced like it's about to roll over

At ~SEK 135 on 5 June 2026, rolling Q1 into the trailing twelve months puts net profit near SEK 373M (~SEK 9.6 of EPS), so you’re paying roughly 14× earnings for a business with a five-year ~45% operating-profit CAGR, ~48% ROIC, SEK 889M of net cash, a ~4.4% dividend (just gone ex), and a 20%-plus near-term growth runway.

The stock has been brutalized. It ran to SEK 293 in mid-2025 before short-selling pressure and a broker note flagging “some moderation of growth” knocked it into the low 100s; it bounced toward the mid-140s and now changes hands around SEK 135. The sell-side consensus target sits around SEK 215–222 — though it was recently trimmed from ~SEK 234, which is precisely the “moderation” the bears are pressing. Do with that what you will.

Insiders send a mixed-but-leaning-constructive signal. In mid-April 2026 the CEO, Dag Bergheim Pettersen, 56 and fourteen years at the helm, exercised 100,000 options at a ~SEK 63 strike (about SEK 6.3M of his own cash) and kept the shares, holding roughly 1.2–1.3M directly. Set against that, company insiders as a group have been net sellers by around SEK 37M over the past year, and the Chairman trimmed earlier in the spring. The man who knows the order book best is adding; the broader insider tape is not a one-way buy. Weigh both.

The bear case on Zinzino

• Growth normalization is real. May’s +14% SEK print is Exhibit A, and the flat-to-negative mature regions back it up.

• MLM structural risk. Distributor-led models can unwind fast when momentum reverses (see Turkey). Regulatory scrutiny of direct selling is a permanent overhang.

• Acquisition dependence. If the deal pipeline slows or integrations stumble, the organic base may not be enough to hit 20%.

• Dilution and FX. Stock-funded M&A and a strengthening krona both eat reported per-share growth.

None of that is dismissible. But weigh it against a quarter in which net income crossed SEK 100M, cash flow rose roughly sixfold, and margins expanded ~450bps while integrating an acquisition. That is not a business breaking. That is a business proving its model scales.

What I'm watching into Q2 (reports 25 August 2026)

• Local-currency growth — does the FX-adjusted number stay above 20%?

• Organic vs. acquired — any stabilization in the Nordics, Baltics, Eastern Europe?

• ItWorks synergies — Q1 baked in none; Q2 is when the goodwill is supposed to start earning its keep.

• Margin durability — that 37% gross margin rode a weak dollar; how much holds if the greenback turns?

The market is underwriting "an MLM about to roll over." The financials describe a high-ROIC, cash-generative compounder at ~14× with a 4%-plus yield. Those two stories can't both be true for long. My money, literally, is on the numbers.

Sources. Zinzino AB (STO: ZZ B) Q1 2026 interim report published 22 May 2026 and the preliminary May 2026 sales report published 3 June 2026. Insider transactions from the Swedish FSA (Finansinspektionen) insider register. Market data as of 5 June 2026, at a share price of around SEK 135. Five-year metrics and the owner-earnings estimate are my own.

How I work. I read the Korean and Japanese filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.

If you’ve ever watched a giant container ship get unloaded, those towering cranes plucking forty-ton boxes off the deck like Lego bricks, there’s a decent chance the software choreographing that ballet was written in Busan, South Korea.

Despite the name, Total Soft Bank has nothing to do with the Japanese telecom-and-investing giant SoftBank. It’s a 37-year-old maritime-software house founded in 1988 and listed on the KOSDAQ since 2002, and it has spent its entire life solving one unglamorous, fiendishly hard problem: how do you run a port?

This is my first post here, and I want to use it to introduce a business I find genuinely fascinating — a tiny, profitable, cash-rich company sitting at the centre of an industry most investors never think about. Let’s get into it.

Before I start, credit where it's due: I'm not the first to write up TSB. Michael Fritzell (Asian Century Stocks) covered it in a deep-dive and Ryan Albert (@ryan_a_albert) in an X thread — both worth a read. I'd taken my position before I read either, and the model and analysis below are my own.

What Total Soft Bank actually does: CATOS and terminal operating systems

Ports are deceptively complex. A single large terminal handles thousands of containers a day, each one needing to be planned, tracked, stacked, moved, and loaded onto the right ship in the right order — without the vessel tipping over, the yard gridlocking, or a crane sitting idle. Get it wrong and you burn money by the hour. Get it right and you squeeze more throughput out of the same concrete.

TSB sells the brain that runs all of this. Its flagship product, CATOS (Computer Automated Terminal Operating System, 케이토스), is what the industry calls a Terminal Operating System — the central nervous system of a container terminal. It plans where ships berth, how the yard is laid out, how cranes and trucks are scheduled, and monitors the whole operation in real time. CATOS is installed at more than 110 container terminals across 20-plus countries.

Around that core, TSB has built a whole stack:

• CASP / Cloud CASP — stowage planning for ships (how to physically load the boxes safely and efficiently). Used by a large share of the world’s biggest shipping lines.

• TSB Supercargo — an onboard “loading computer” that calculates a vessel’s stability and structural strength.

• MOST — an operating system for multi-purpose terminals (bulk, liquid, odd-shaped cargo).

• PLUS — a port community system that links shippers, agents, carriers and authorities.

• Simulators & training — crane, welding and painting simulators, plus a 3D “digital twin” platform delivered to a terminal in Vietnam in 2025.

The strategic logic is what they call vertical integration: TSB can supply software for the ship, the shipping line, the terminal, and the wider port community — a one-stop maritime IT shop. In an industry that increasingly wants integrated platforms rather than bolted-together point solutions, that’s a real advantage.

Why Total Soft Bank has a moat

Port operating software is about as sticky as software gets. Three things stand out:

1. Brutal barriers to entry. Terminals are critical infrastructure. Operators are conservative and risk-averse — nobody wants to be the person who installed the buggy software that gridlocked the national port. So they buy from a tiny club of vendors with long track records. TSB estimates only ~20 companies worldwide truly compete at the global level.

2. Long contracts, long relationships. These are multi-year deployments, often with maintenance tails that run for years afterward. Once your TOS is embedded, ripping it out is a nightmare. That makes revenue durable and recurring.

3. An oligopoly with structural tailwinds. Ships keep getting bigger, ports keep automating, labour keeps getting scarcer and more expensive, and global trade keeps getting more volatile (US–China tensions, supply-chain reshuffling). All of that pushes operators toward smarter, more automated systems — exactly what TSB sells. The company is also leaning into a subscription/SaaS model (cloud CATOS, cloud CASP), which over time should mean more predictable, recurring revenue.

This is a classic “picks and shovels” story: TSB doesn’t own ports or ships, it just sells the indispensable software that makes them run — and gets paid whether trade is booming or merely chugging along.

Total Soft Bank Q1 2026 results: revenue up 39%

On to the part that made me want to write this. TSB reported its Q1 2026 results (the three months ending 31 March 2026), and they were excellent.

• Revenue: ₩7.38bn, up ~39% year-on-year.

• Operating profit: ₩1.86bn, up ~30%, for an operating margin north of 25%.

• Net profit: ₩2.57bn, up a remarkable ~92% — nearly doubling.

Roughly 80% of revenue came from CATOS, underlining how central the flagship product is. It’s worth noting the net-profit surge was flattered by foreign-exchange gains — TSB earns the large majority of its revenue abroad (in USD, EUR and others), so a weaker won inflates reported profit. Strip out the FX noise and the underlying operating story is still very healthy, just less explosive than that +92% headline suggests. Always read the won-denominated profit of an exporter with one eye on the currency.

For context, this isn’t a one-off. Full-year FY2025 revenue was about ₩27.5bn with operating profit of ₩9.4bn — so the business has been compounding nicely, not just printing a single lucky quarter.

The order backlog: ₩38bn at Q1 2026

For a project-and-contract business like this, the single most useful forward-looking number isn’t quarterly revenue — it’s the order backlog (수주잔고), the work already signed but not yet recognised as revenue. It’s the closest thing to a crystal ball.

The trend is the story. Backlog roughly doubled between early 2024 and early 2025 as a wave of new contracts landed, and even after working some of it off, TSB ended Q1 2026 with ~₩38bn of backlog — back near its all-time highs. CATOS makes up the bulk of it, and notably the maintenance slice jumped to its highest level in the dataset (~₩8.9bn), which is exactly the kind of sticky, recurring revenue you want to see growing.

Put differently: TSB is sitting on a backlog worth well over a year of revenue. The recent contract announcements behind this include terminal-system deals in Abu Dhabi, Malaysia (Penang and Johor), Kenya, Spain (Valencia) and Taiwan — a genuinely global order book.

The balance sheet: ₩34bn of cash against a ₩57bn market cap

Here’s the part that really caught my eye. TSB is not just profitable — it is drowning in cash and carries essentially no debt.

At the end of Q1 2026 the company held roughly ₩16bn in cash plus another ~₩18bn in short-term financial instruments — call it ~₩34bn in cash and near-cash. Against that, interest-bearing debt is a rounding error (a single small lease). The debt-to-equity ratio is under 50%, and the current ratio is a fortress-like ~390%.

Now compare that to the market value. As of early June 2026, Total Soft Bank's market capitalisation sits around ₩55–59bn. So cash and investments alone are close to 60% of the entire market cap. Back out the net cash and you’re paying a strikingly small amount for a profitable, growing, moaty software business.

The company is also returning capital: it paid a dividend for FY2025 (₩100/share) and has been buying back its own stock through a treasury-share trust in the first half of 2026. For a small Korean company, that shareholder-friendly behaviour is worth noting, especially against the backdrop of Korea's broader Value-Up push to get cash-hoarding companies to reward shareholders.

The risks in Total Soft Bank

No write-up is honest without the risks, and there are real ones:

• It’s tiny and illiquid. This is a micro-cap on the KOSDAQ. Annual revenue is only around $18–20m. Position sizing and liquidity matter.

• Lumpy, project-based revenue. Big contracts land unevenly. Any single quarter can swing hard, and government-tied port budgets can slip.

• FX is a double-edged sword. The won’s weakness juiced this quarter’s profit. It can just as easily reverse and create a nasty-looking headline in the other direction.

• Customer concentration. A handful of large terminal operators and shipping lines drive a big chunk of the order book — in Q1 2026, two customers alone were ~30% of revenue.

• Founder/owner-led, old-guard governance. The founder-CEO was born in 1941 and still runs the company; the largest-shareholder group holds ~31%. That’s stability, but also key-person and succession risk, and the kind of governance overhang that has historically kept Korean small-caps cheap. (At the March 2026 AGM, a proposed executive-compensation rule was actually voted down — a small sign the shareholder base is engaged.)

The bottom line on Total Soft Bank

Total Soft Bank is the kind of company that almost never makes headlines: small, profitable, boring-in-a-good-way, sitting on a mountain of cash, and quietly embedded in the plumbing of global trade. It owns a defensible niche in an oligopolistic, high-barrier industry with secular tailwinds (automation, bigger ships, digitalisation), it’s growing its backlog, and it trades at a valuation that barely seems to credit the operating business at all once you account for the cash.

That combination — durable moat, net-cash balance sheet, growing order book, micro-cap obscurity — is exactly the sort of thing that tends to get overlooked precisely because it’s quiet.

I’ll be following the next couple of quarters closely, especially how much of that fat backlog converts to revenue and whether the SaaS transition starts showing up as more recurring, less lumpy income. If it does, the market may eventually have to start paying attention.

Since writing this I have covered several other overlooked Asian micro caps, including for example i-Scream Media (KOSDAQ: 461300) or KIYO Learning (TYO: 7353).

Sources. Total Soft Bank (토탈소프트뱅크, KOSDAQ: 045340) filings on DART: the Q1 2026 quarterly report (38th term) and the FY2025 annual report. Contract announcements and backlog data from the company's DART disclosures. Market data as of 5 June 2026. Won conversions at roughly ₩1,500 to the dollar.

How I work. I read the Korean filings myself. AI helps me with translation and with checking the arithmetic. Every judgement, model and conclusion here is mine.

Disclosure. Assume I own shares in every company I write about, and read accordingly. I don’t disclose individual positions or sizes.

Disclaimer. This is analysis for discussion, not investment advice, and not a recommendation to buy or sell anything. I’m not your adviser. The figures come from primary filings but I make mistakes, so check them yourself before you act. Micro caps carry real liquidity and volatility risk, and you can lose your money.