Golfzon Holdings (KOSDAQ: 121440) – last price ₩7,000 – ~₩300bn (~US$215mm) market cap

Thesis summary: on Friday, the controlling family’s second attempt to take Golfzon Holdings private closed having purchased precisely zero shares — not ‘a few’, not ‘less than hoped’: zero — and the stock promptly fell 10.3% to 7,000 KRW. I think that reaction is bizarre, and added significantly to my position. The bear case in this name was never valuation — the company trades at about a third of its own stated book value and roughly a quarter of a conservatively marked net asset value — it was process risk: that a controller with 85% of the votes would grind the float out at 6,700 KRW, delist the company, and leave holders to pursue an appraisal action through the Korean courts for two or three years.

In my view, though, Friday removed a great deal of that risk. We now know that there is no marginal seller at 6,700, and that the controller has publicly stepped back from the delisting mechanics. We also know that he does not control the process that will crystallise the value here (that is, the Golfzon County sales process); and that he has an 80.3bn KRW loan maturing in June 2027 that must be repaid out of this structure (which was raised to purchase all of Golfzon Holding but now ends up at 85% with a listed stub). In other words, to me it seems the asymmetry improved massively and crucially the key risk – automatic delisting and forced appraisal – looks to have receded almost completely.

The set-up here is increasingly similar to the Hunter Douglas adventure from a few years ago — a controlling family holding ~85-90%, twice failing to buy in its own float at a lowball price, whilst sitting on assets worth multiples of the bid, all whilst pursuing a liquidity event for those same assets. That engagement ended with the Sonnenberg family selling Hunter Douglas at 2.7x its own opening bid, and it was a massive winner for me and my readers. Section 6 sets out why I think the parallel holds — and perhaps where the minority position at Golfzon is actually stronger. Even on the most punitive Korean appraisal maths the shares are worth 1.7-2.3x the current price; on net asset value, three to four times.

Background

Golfzon Holdings is the listed holding company of the Golfzon group. Its assets are a 31.58% interest in Golfzon County (21 golf courses, currently in a live auction run by MBK Partners at a reported ~2 trillion KRW enterprise value), roughly 23% of listed Golfzon Co (215000), a portfolio of investment real estate carried at 316bn KRW against the company’s own disclosed appraisal of 567.5bn KRW, and a collection of smaller operating businesses.

Post latest tender offer, per below, the free float is only about 5.7mm shares, with the controller now owning 85%:

At 7,000 KRW the entire float is worth about 40bn KRW, or some US$29mm.

This whole process begain in March, 2026, when MBK Partners mandated Morgan Stanley and Samjong KPMG to sell Golfzon County; the information memorandum went out in April at an enterprise value of approximately 2 trillion KRW, and preliminary bids closed on 29 May (coverage via thebell and InvestChosun). On 8 April — between the mandate and the bid deadline — Kim Won-il, the founder’s eldest son and a former chief executive, incorporated a shell company called SJ Investment Holdings. On 29 June, six days after preliminary bids closed on the crown-jewel asset, SJ launched a tender offer for the entire float at 6,700 KRW (the tender offer statements are on DART). The stated purpose was delisting. The financing was 25bn KRW of equity and an 80.3bn KRW loan from NH Investment & Securities — 76% debt — maturing June 2027.

The first offer took up 9,742,319 of the 15,485,020 shares sought, leaving SJ at 85.21% of the votes (excluding treasury shares) – a level that does not allow SJ to delist (90%) or squeeze-out (95%) minorities. Much like in the Hunter Douglas situation, I penned a letter to the Board, given their conduct here has been woeful and evidences massive failings of fiduciary duty, and process, even in a market as historically biased against minorities as Korea has been. For example:  the chairman of the ‘independent’ special committee was listed in the offeror’s own tender statement as a special related party of the bidder; a second committee member is disclosed in the company’s own career table as formerly of NH Investment & Securities’ investment banking division; and the committee’s independent valuer, Samjong KPMG, is MBK’s sell-side adviser on the Golfzon County sale — the same firm that advised the committee, on 29 July, that the County process was ‘not at a sufficiently advanced stage to be reflected in the valuation’, on the very day the Korean press reported that process running at ~2 trillion KRW.

More than this, the board supported the offer whilst declaring itself neutral on price and warning that the price ‘does not represent the absolute value of the Company’s current or future enterprise value’ – a fairly extraordinary statement for any supposed fiduciary to make! And for the second tender offer, it voted down its own motion to form a committee at all.

SJ relaunched on 10 August at the same 6,700 KRW for the remaining 5,742,701 shares. The offer closed on 2 September having acquired no shares at all — an outcome which, so far as I can tell, is without precedent in the modern history of Korean tender offers. Throughout the offer period the stock closed above the offer price every single day. And then, on 4 September, SJ’s comment on the failure (reported second-hand in a newspaper report) was thus: “현재로서는 추가 공개매수나 포괄적 주식 교환 등 상장폐지와 관련된 후속 절차를 검토하고 있지 않다” — “as of now, we are not reviewing follow-on procedures related to delisting, such as an additional tender offer or a comprehensive share exchange.”

I will make three observations on that statement. First, it is a press comment, not a regulatory filing – that is to say I believe it is not legally binding on SJ (even though it includes the proviso ‘for now’ anyway. Second, ‘as of now’ is an explicit temporal carve-out, and ‘are not reviewing’ describes an internal process rather than making a commitment — the weakest form of denial available. Third, and most usefully, the denial is specific: it names the two routes that require a public filing, and says nothing about on-market purchases, off-market blocks from fewer than ten counterparties, capital reduction, or a merger.

Remember also that because there is no cooling off period in Korea, there is no limitation on SJ launching another tender in the near-term (or even buying on market or off, right now). Given the status quo is not only humiliating but also quite possibly the worst outcome for SJ – having drawn down short-term debt to fund his MBO, only to be unable to delist and fully consolidate at 100% – and with the price having been anchored in the first two tenders at such an unbelievably low level, I find it very difficult to fathom how even if he attempts to go to a comprehensive share exchange, there isn’t (at the very least) one further tender at a big premium to last, to try to get this done cleanly. I will unpack a few of the time-sensitive reasons for this belief, further below.

What the business actually earns

On an earnings basis Golfzon is hardly exceptional (explaining a good portion of the discount to book historically) but it should be noted it is a consistent dividend payer and had (before the tenders) been retiring some stock too:

In H1 2026 revenue was up 0.2% year on year, operating profit down 4.0%, and net income down 43.7% — the top line has stopped falling after two years of ~9% declines, operating profit is broadly flat, and the damage is below the operating line. Frankly, the company is comfortably profitable but the operating business is unremarkable — call it 28-35bn KRW of operating profit and 17-24bn of net income on a run-rate basis — and it has a record of returning capital, paying a steadily rising dividend (167 KRW in FY2023, 250 KRW in FY2024 and FY2025 — a ~3.6% yield at 7,000) and repurchasing shares in each of the last three years. So you are paid at least something to wait. But nobody should own this stock for the P&L, and one balance-sheet point matters: contrary to the usual assumption about Korean holdcos, this is not a net cash story — 77.0bn of cash and short-term investments sits against 172.1bn of borrowings, or net debt of 95.1bn KRW, against total assets of 1,092.5bn.

What is Golfzon worth?

Terton Capital, the US fund that has campaigned publicly against the offer — its open letters of 22 July and 18 August are worth reading in full, and the campaign has drawn international press coverage — marks the company as follows:

The Golfzon County figure is 2 trillion of enterprise value less 727.8bn of County net debt, times 31.58% (Golfzon’s equity stake); the property appraisal is the company’s own audited disclosure, so the NAV arithmetic is fine. But I personally would take the 31k KRW number with a bit of a grain of salt, for a few reasons: both mark-ups are pre-tax (a realised gain leaks roughly 26.4% before it reaches equity — call it 3,100 KRW per share); Terton marks up two assets whilst ignoring a third carried above market (the ~23% of Golfzon Co sits at 98.8bn under the equity method against 50.7bn at market — consistency costs 1,245 KRW per share); 2 trillion for the County is an ask, not a bid — this is MBK’s fourth exit attempt, and the uplift vanishes at about 1.35 trillion. Still, even if you adjust for the taxes and the Golfzon Co stake to market, you still get a range of 21k-31k KRW:

We also need to remember that Korean courts and Korean transaction practice value a company on a blend prescribed in the FSC’s Enforcement Rules on the Issuance and Public Disclosure of Securities — the basis used to price mergers and share exchanges:

Intrinsic value = (asset value × 1 + earnings value × 1.5) ÷ 2.5

Note the weighting: earnings are weighted 1.5x against assets. For an asset-heavy holding company whose operating businesses earn very little relative to what they own, that formula is structurally punitive. Trailing net income of 17.2bn and FY2025 of 24.0bn give a capitalised earnings value of roughly 4,500-6,200 KRW per share. In other words, the bid was priced entirely off earnings and the depressed market price, with zero weight to the assets. That is not an accident, and it is the quantitative proof of what the board’s own opinion admits when it concedes the price ‘does not represent the absolute value’ of the company. If we apply the statutory blend to a properly marked asset value instead we get 14k-16k KRW in intrinsic value even under the punitive Korean valuation methodology:

This is about 2.0-2.2x Friday’s close. Calibrating what a court would actually award in an appraisal proceeding is obviously difficult, but experts I have consulted with suggest 12,000-16,000 KRW: well below full net asset value, and still more than double the market price.

Hunter Douglas redux?

Regular followers will recall that Raper Capital engaged publicly with Hunter Douglas’ independent committee through three open letters in the spring of 2021 (the initial letter, second and third letters all remain on the Engagement page of this site), and the position ultimately resolved as one of the great winners in my history. Bergson, the Sonnenberg family’s vehicle, already held ~84% when it set out in April 2021 to buy in the remaining 16.1% free float at €64 a share. Minorities initially refused such an insulting bid; In June the family raised to €82 — and captured just 3.8% of the shares. That took Bergson to 93.5% of total capital but only 87.66% of the ordinary shares: enough to compel a squeeze-out under Curaçao law, but short of the 95% Euronext Amsterdam required for delisting. The family held the legal right to force minorities out, but at the time declined to use it, and suggested they would maintain the listing ‘for the time being.’

The process criticisms will sound familiar. The Dutch investors’ association found that the independent committee had ‘failed to take its role sufficiently seriously’ in endorsing the first bid; that no peer-comparison valuation was performed; that the fairness opinion came from a bank that had played the same role defending a thin bid for the same company in 2008; and that the valuation reasoning was circular — attributing the discount to governance problems rather than recognising that the discount was investors rationally pricing the risk of precisely what the family was attempting.

Then, five months after telling minorities that €82 was full value — the Sonnenberg family agreed to sell 75% of Hunter Douglas to 3G Capital at €175 per share, retaining 25%, with minorities offered the same €175. That is 2.7x the family’s opening bid and 2.1x its ‘final’ raised offer. Value that is invisible whilst the controller is buying becomes perfectly visible when he is selling. And when you stack the HDG case sidfe by side with Golfzon, the parallels are almost eerie:

Note the one material difference is that Bergson had the squeeze-out right and chose not to exercise it. SJ does not have it at all — 85.21% against a threshold requiring 3.56m more shares, or 62% of a float that has just tendered nothing. Hunter Douglas minorities were legally squeezable, sat still, and were paid 2.7x. We are not squeezable, and the asset that determines our value is being auctioned by a third party on a timetable the controller cannot influence. Were it not for my unfamiliarity with Korean markets, on a neutral view I would posit this is probably therefore a better position than the post-tender HDG set-up.

Friday’s disclosures made this materially safer

The market has misread every element of what it learned on Friday:

• There are zero sellers at 6,700: Zero, out of 5.74m shares sought. Everyone remaining on this register has now refused 6,700 at least once, and many bought above it. That is about as clean a demonstration of a hard price floor in the underlying holder base as one ever gets in a public market.
• The delisting risk — the actual risk in this name — has receded. The controller has publicly stated he is not pursuing the delisting mechanics. I do not take that at face value — the language is deliberately provisional, as noted above. But the practical position is this: he needs 3,561,002 more shares to reach the 95% squeeze-out threshold, which is 62% of a float that just tendered nothing, and any comprehensive share exchange requires a public filing into a regulator that has spent 2026 issuing correction orders on precisely these structures. I cannot see this company being delisted without at least one further tender offer at a materially higher price. That converts the downside from ‘trapped in an appraisal action’ to ‘wait, and get paid to wait’.
• The controller does not control the catalyst. To me this is the most persuasive reason to add more capital here post failed tender (and also the biggest risk to monitor going forward). The Golfzon County sale is run by MBK Partners through its own SPC, with Morgan Stanley and Samjong KPMG mandated by MBK, and MBK holds the drag-along right. Golfzon Holdings is a 31.58% passenger. The family cannot slow-walk, pause or cancel the process that crystallises the asset value — which was the entire reason to attempt this buyout now, ahead of that crystallisation. He tried to get in front of it and failed. The value event proceeds on someone else’s timetable, and as long as that GC sales proceeds concludes – the big risk now – I believe a rebid/retender/etc is inevitable at Golfzon.
• There is a hard funding clock. SJ financed 76% of its bid with an 80.3bn KRW loan from NH maturing June 2027 — nine months away. That debt sits above a holding company with 14-17bn KRW of parent net cash and a business generating negative free cash flow. It gets repaid from a refinancing, an equity injection, or a distribution out of Golfzon Holdings. A County sale at anything near the marketed value delivers roughly 400bn KRW of proceeds against a 195.8bn carrying value — a ~204bn pre-tax gain, ~155bn after tax, about 4,000 KRW per share of distributable capacity. A large special dividend is tax-efficient for a Korean corporate shareholder, requires no minority consent, and repays the NH loan several times over. Thus, Even if he never buys another share, the most likely path to servicing his own debt suggests a massive special dividend at the Golfzon level.
• Even on punitive Korean appraisal maths it is more than a double. 13,900-15,700 on the statutory blend; 12,000-16,000 as a realistic court range; 21,500-31,100 on net asset value. Against 7,000 the most punitive of those figures is still 1.7x, the midpoint is a double, and NAV is three to four times the market. And that is the appraisal outcome — the downside scenario, the one that only matters if he forces us into court.
• This is an unprecedented situation in Korea, and the regulator and the market should now be watching it far more closely. Let me be plain about how unusual this is. A controlling shareholder launched a delisting tender against a live auction of the crown-jewel asset; staffed the ‘independent’ committee with his own related party; used a valuer who was simultaneously the seller-side adviser on that very asset; was designated for unfaithful disclosure over an undisclosed share pledge mid-offer; declined to convene any committee at all for the second attempt; and then failed to buy a single share. Any one of those would be remarkable; the full set, so far as I can tell, has simply never happened before in the Korean market. And it has happened at exactly the wrong moment for the controller: after an amended Commercial Act that runs directors’ duties to shareholders directly, after a Ministry of Justice guideline written for this precise fact pattern, in a year when the FSS has repeatedly forced delisting sponsors to correct and raise their terms. Golfzon is now the reference case for controlling-shareholder conduct in Korea — foreign funds are campaigning on it publicly, the international press is covering it, and every future filing this controller makes will land on a regulator’s desk with this history attached. Against that backdrop, and with a loan maturing in June, I think the controller is far more likely to return quickly at something in the region of 11,000 KRW — some 57% above Friday’s close — than to sit and play out the string for years whilst the political environment gets steadily worse for him. At 11,000 the entire float costs him 63bn KRW — less than he already spent on the first tender — to capture assets I mark at 21,500-31,100 per share. That is an easy decision for a rational buyer, and he has already demonstrated he wants it badly.

Risks

Despite the attractiveness of the investment, there are a number of risks to keep front and centre. The Golfzon County sale could fail again, or it could clear materially below 2 trillion (not killing the thesis in the latter case but certainly hurting it). Alternately MBK’s preferred shares (in GC) turn out to carry a preference that takes the first slice of proceeds, in which case our 31.58% is worth less than pro rata. Of course, the logic behind this remaining a value trap is much weaker now, in my view, but that risk remains: the controller could do nothing at all for two or three years, collectsthe dividend (a 3.6% yield at 7,000, for what it is worth), and lets the float rot in an illiquid stock. There is also the risk that he accumulates quietly through on-market and small-lot off-market purchases, reaches 95% without ever making another public bid, and squeezes us out at a court-determined price that lands nearer 12,000 than 16,000 (though in the current political environment and given the underlying illiquidity I am less worried about this). Crucially, that even that adverse path is 1.7-2.3x from here. Nevertheless, given all these risks, and my newness to Korea, this is still a reasonably modest position for me.

Conclusion

The tender failed at zero; the floor is proven; the delisting mechanics are publicly shelved; the catalyst is in a third party’s hands; and the controller’s own loan matures in June 2027. The punitive-case appraisal math give 1.7-2.3x; the statutory blend on properly marked assets gives 2.0-2.2x; net asset value gives 3-4x; and the most likely near-term resolution — a return to the well at something like 11,000 KRW — is a 57% up-move that would still be a bargain for the man paying it. Buy

Disclosure: Long Golfzon (121440.KQ)