A company whose largest operational headache is weekend traffic at the entrance ramps of its malls in Hanam and Goyang does not belong in a sentence about the biggest acquisition in Hollywood history. Yet the sentence is not a typo. Shinsegae Property, the Korean developer behind the Starfield chain, is building a KRW 9.5 trillion (approx. USD 6.8 billion at KRW 1,400/USD) tourism complex in Hwaseong, and the theme park at its centre stands on intellectual property licensed from Paramount. On October 6, Paramount closed its acquisition of Warner Bros. with roughly USD 79 billion of net debt on day one.
Consider a man building the house he intends to live in for the rest of his life on land he leases from someone else. One day he hears the landowner has borrowed heavily to buy up the next village. He loses sleep not because the landowner got richer, but because a landowner chased by interest payments may change the terms of his plot, or hand the best of the new land to whoever pays first. The landlord did not change. The landlord’s circumstances did.
In the language of finance, a licensee has become exposed to its licensor’s capital structure, and an equity stake is only one of several tools for handling that exposure. One caveat governs everything below: Shinsegae Property’s participation in Skydance is not a disclosed fact. The desk dissects the structure on the assumption that participation is under review, and asks the question a deal room would ask — not whether the shares are cheap, but what a licensee with KRW 9.5 trillion of land exposure should actually buy.
The Deal
The combined company trades under the Skydance name. Paramount, controlled by the Ellison family and RedBird Capital, completed its acquisition of Warner Bros. Discovery on October 6 and emerged as one U.S. media group holding both studios, their cable networks and two streaming services. What was sold, to whom and at what price is on the public record; what Shinsegae Property might buy is not.
- Control. Every voting share sits with the Ellison family and RedBird. Sovereign wealth funds from Saudi Arabia, Qatar and Abu Dhabi reportedly contributed about USD 24 billion, received non-voting stock and gave up board seats.
- Equity pricing. New shares were priced at the average of the 20 trading days before closing, inside a collar (a pre-agreed floor and ceiling) of USD 12.00 to USD 16.02. The floor applied. Investors received roughly 3.92 billion shares for close to USD 47 billion, taking the share count to just over 5 billion; the Ellison family funded most of its commitment at the same USD 12. Public holders of record before closing — and only they — received free ten-year warrants struck at USD 12.
- Leverage. Pro forma long-term debt of USD 80.3 billion against USD 7.9 billion of cash; net debt of about USD 79 billion, roughly 6.5x adjusted EBITDA (operating cash earnings as the company defines them), against the 4.25x downgrade line rating agencies had cited. Annual interest expense in the same filing: USD 6.4 billion.
- The licensee. Shinsegae Property has injected KRW 341.35 billion (approx. USD 244 million) into the Hwaseong project entity to date, including KRW 30 billion in February and again in September this year. The Paramount licence was signed in 2024; first-phase opening is targeted for 2030, pushed back from 2029; the site is to be developed in phases through 2050. The park occupies 9.3% of the land — about 80,000 pyeong, roughly 26 hectares. The rest is retail, hotels, golf and residential.
Structure Breakdown
Start with what the USD 12 share actually buys. The sovereign funds, with USD 24 billion between them, got neither a vote nor a board seat; a Korean property developer with a fraction of that cheque will not either. The instrument confers exposure to outcomes, not rights over management. That is not a flaw in the paper. It is what the paper is.
The exposure is thin. At USD 10 a share the equity is worth a little over USD 50 billion, around 40% of enterprise value once the debt is added, so a 10% move in enterprise value moves the equity by close to 25%. The desk’s three-year per-share estimates run from the USD 19 range in the bull case to the USD 9 range in the base case and below USD 1 in the bear case. The figures depend on assumptions; their shape does not. An investor entering at USD 12 does not recover principal in the base case, and USD 10 lands near break-even. Buying cheaper does not change the structure.
Look at where the debt service comes from. Of Warner Bros. Discovery’s USD 8.7 billion of adjusted EBITDA in 2025, USD 6.4 billion came from cable networks, and cable earnings fell 21% in a single year — the heaviest load placed on the fastest-melting ice. Management has promised USD 6 billion of cost savings; one analysis puts the figure needed to bring leverage down to 4x at closer to USD 16 billion. The company has also committed to 3x by 2029, a maintained dividend and content spending above USD 30 billion a year. It is hard to see all three kept.
The opposing case deserves its paragraph. This is one of the largest film and television libraries in existence; the two streaming services together exceed 200 million subscribers; Paramount beat second-quarter expectations and raised guidance; and the Ellison family, personal guarantee attached, came in at the same price. There are reasons to own the shares. They amount to a bet that the integration succeeds — a different problem from the one Shinsegae has to solve in Hwaseong.
Here the real centre of gravity shows itself. A company that must find USD 6.4 billion of interest every year looks for cash it can raise without printing more stock, and licensing characters to theme parks and experiential venues is one of the few channels that produces upfront payments without the licensor’s own capital. The company itself disclosed that its equity investors bring strategic and commercial opportunities, and a Warner Bros. park already stands in Abu Dhabi. The shareholder register and the licensing contract can be discussed at the same table. What Shinsegae should be buying is not the stock but the right to sit at that table.
That table’s map is more crowded than it looks. Harry Potter is placed with Universal’s park in Japan and a studio tour in Tokyo; in China it sits inside Universal Beijing, with a Shanghai facility reportedly in the works. Batman and Superman are tied to Six Flags in North America, and Warner is reported to have discussed new ride licences with operators including Universal late last year. Which characters are unencumbered in Korea cannot be confirmed from public sources. The first act of this deal is therefore not a wire transfer but a written schedule of what is actually available in Korea and the surrounding region. If the schedule is empty, the transaction ends there.
Shinsegae’s own ledger explains what those rights are worth. The park takes less than a tenth of the land; it is a signboard that lifts the land value and rents of the other 90% more than it is a profit centre, and a change in the grade of the characters on that signboard moves the value of the entire complex, not ride revenue. Shinsegae’s leverage is not a threat to switch franchises — the licence is signed, this year’s money is already going into design, and walking away is expensive even before groundbreaking. Its real levers are two: cash it can hand over first to a counterparty chased by interest, and the remaining phases of a site that runs to 2050. Both lighten as the design hardens and the counterparty’s finances settle. The Gulf capital on the same register is both opportunity and variable: Shinsegae opened 49% of Starfield Goyang to an outside fund and may want the same in Hwaseong, but buying stock in the market seats no one at a table set by Ellison and RedBird, and the Gulf holders may prefer the next Warner facility built at home.
The Governing Logic
Run the arithmetic the equity has to survive. If adjusted EBITDA stays near USD 12 billion — the desk’s pre-synergy working estimate, of which more than half already goes to interest — and the market pays 6.5x, enterprise value is USD 78 billion. Set remaining net debt at USD 75 billion and the whole equity is worth USD 3 billion: under a dollar a share across 5 billion shares. The figures are assumptions. The fact that shareholders stand last in line behind USD 6.4 billion of annual interest is not. An asset that can lose half is handled one way; an asset that can lose nearly everything is handled another.
So the first design principle is sequence. Obtain the written schedule of purchasable rights before any money moves; fix the terms of those rights; only then decide the form the money takes. Start from equity and the rights become an afterthought; start from rights and the equity does. Keep the sequence and the deal ends at zero cost the moment the counterparty fails to produce a schedule. The cheapest downside device is not a clause; it is the option not to invest.
Allocation comes second. Assume a USD 200 million envelope — approx. KRW 280 billion at KRW 1,400/USD, about what Shinsegae Property has already put into the Hwaseong entity. Split it three ways: half as prepaid royalties and territorial rights; 30% into a joint venture for experiential businesses in Korea and the region; equity only if the counterparty demands it as the price of the rights, and never more than 20%. Direct exposure to the share price falls to USD 40 million. That is not the maximum loss. The other two compartments trade share-price risk for the risk that the counterparty performs its contract.
Downside Protection
The prepaid royalty. Royalties are a cost Hwaseong pays anyway once the park opens; the logic is to pay them early in exchange for wider character scope and a longer term — cash for the licensor without new shares, and for Shinsegae a defined-use outlay indifferent to the share price. Size it carefully. The opening target is 2030 and the royalty rate is undisclosed, so prepaying several years means lending to the counterparty’s credit for that long. Nothing is handed over in one piece: the prepayment sits in third-party escrow and is released against design approval, groundbreaking and delivery of the character assets.
Survival of the licence. The acquisition financing holds a first lien (first claim on collateral) over substantially all assets of the combined company, Paramount and Warner included. When the company stumbles, lenders reach the assets first, and an unsecured licensee that has prepaid stands at the back of the queue. Three sentences belong in the contract: the licence survives the licensor’s insolvency proceedings; it follows the IP if the IP is sold; its terms hold through a change of control. A 2019 U.S. Supreme Court decision on trademark licences was favourable to licensees, but how far it reaches a Hwaseong contract requires separate legal review, and it does not guarantee return of the prepayment.
Clean title. Take a representation that the characters on offer are not already committed to another operator in the territory, backed by a schedule of existing contracts, with an indemnity for the full consideration plus redesign costs if it proves false. Given how finely Harry Potter and Batman are already carved by country and operator, this is the core of the agreement, not boilerplate. A right that cannot be delivered is expensive at any price.
The joint venture. At parent level even sovereign funds got no board seat. A separate entity holding only the experiential business in Korea and the region is a different proposition: Skydance contributes character usage rights, Shinsegae contributes land, development and operations, and Shinsegae holds the majority with consent rights over budget, the scope of character use, additional borrowing and share transfers. Control comes not from a small stake in a large company but from a large stake in a small one. The weakness is obvious — the JV’s core asset is a licence from the parent — and is addressed by a clause vesting the usage rights in the JV for the remaining term, at no additional consideration, if the parent defaults.
Two opening positions. First, a put option (the right to sell at a pre-agreed price) on the JV stake after a set number of years at principal plus a minimum return, written against an Ellison family entity rather than the company, because the moment the put matters is the moment the company may be unable to pay. Second, a set-off under which equity mark-to-market losses beyond a threshold are deducted from Hwaseong royalties. Candidly, the odds of either being accepted are low: a family has no reason to lend personal credit to a small transaction, and lenders will not welcome a clause that trims royalty income. Their use is not in being accepted but in being conceded, in exchange for the survival clauses and the width of the territory.
The equity, smallest of all. The allotment is finished and the warrants went to pre-closing holders; anything bought now is secondary stock from an existing investor or common shares in the market, with a base-case three-year value in the USD 9 range. Take equity only when it is demanded as the price of rights, and book it as a cost of the transaction rather than a source of return. Currency partly hedges itself: a weaker won raises the dollar royalty burden but also lifts the won value of dollar assets.
The limits are worth stating plainly. This structure removes no risk. It swaps share-price risk for contract risk, and the counterparty to the contract is still a company under a mountain of debt. The exchange is still worth making because the share price is a variable Shinsegae cannot touch, while the contract is a document Shinsegae can write. An architect converts uncontrollable risk into controllable risk — and the time to write that document runs out as the design hardens and the balance sheet stabilises. The leverage is usually heaviest in the few quarters right after a deal closes.
What Must Be True
Proposition 1: the adjusted numbers approximate cash. The 6.5x figure is computed on adjusted EBITDA, which excludes restructuring and transaction costs; Paramount booked USD 153 million of deal-related costs in the second quarter alone. The pro forma statements filed at closing show combined 2025 revenue of USD 66.1 billion against an operating loss of USD 2.6 billion and a net loss of USD 6.1 billion — distorted by acquisition accounting and one-offs, but a reminder of the distance between the word “adjusted” and the number in the ledger. The same lens applies to the licensee: Shinsegae Property’s 2025 operating profit rose sharply, but the increase appears to rest on one-off items — a stake sale, changes in consolidation scope, valuation gains on financial assets — rather than recurring cash. The 30 million annual visitor forecast is the company’s expectation, raised from 19 million in 2019; the project has litigated with K-water over a construction deadline; the opening has already slipped a year. USD 200 million is not surplus; it competes with the capital Hwaseong itself requires. Falsifier: a first full post-closing year in which interest cover deteriorates on reported rather than adjusted cash flow would put the bear case at the centre.
Proposition 2: the approvals hold. The U.S. Department of Justice cleared the deal unusually quickly, followed by reports of staff-level concerns. Several state attorneys general sued in July, and the transaction closed only after a court approved a settlement whose operating constraints still need to be read. The European Union required exit from an overseas distribution joint venture within 13 months. Senators raised foreign-influence objections to Gulf capital and a news network under one roof. Approval obtained through political access becomes a liability when the politics turn. Falsifier: Tencent’s withdrawal from a USD 1 billion commitment is the precedent — any sign that an Asian investor on this register draws the same scrutiny narrows Shinsegae’s room before a clause is negotiated.
Proposition 3: Korea is open on the rights map. Reports conflict on when Six Flags’ Batman and Superman rights end — 2053 in one account, late-2020s renewals in another — and a reversion clause is reported to return the rights to Warner if another studio attempts to buy them. The map may be redrawn after the merger, and the pen is held by the voting holders, with an Abu Dhabi shareholder that already hosts a Warner park beside them. Add integration: large-scale layoffs are announced, the organisation will look inward for some time, and a changed contact or a delayed character approval is foreseeable; a deemed-approval clause — silence past a fixed deadline counts as consent — is the only defence. Falsifier: the written schedule itself. If it is thin, the Korean vacancy Shinsegae is counting on does not exist.
Let these meet an external shock. Rates rise again, the advertising market cools, cord-cutting accelerates; adjusted EBITDA is pinned near USD 12 billion while USD 6.4 billion of interest leaves regardless. The company hunts for cash without issuing shares, and overseas character rights become the easiest asset to sell — to whoever has the most money. The day those rights go to another operator, a Shinsegae holding only shares takes the equity loss and a competing facility at once; a Shinsegae that has already tied the rights into a contract finds the same event an opening to widen its terms. The chain does not stop at the share price. A licensor judged unstable raises the return demanded by Hwaseong’s lenders and co-investors, and the valuation loss consolidates into E-Mart’s books. A crack that starts at USD 200 million spreads to the funding cost of a KRW 9.5 trillion project.
The Architect’s Takeaway
Strip the scale away and one sentence remains: whoever builds a business on someone else’s asset should not try to become that someone’s owner, but should buy rights at the moment cash is most precious to them. A seller whose revenue rides one platform’s algorithm, a five-person development shop living on one client’s orders, a franchisee whose storefront carries one head office’s trademark — all stand where Shinsegae stands, burying real capital in the ground while renting the signboard. The disease has the same name at every size: the tenant’s disease.
If you were building your own deal room, the first task is a dependency map with three numbers: the share of your income one counterparty represents, the remaining term of your contract with them, and the time and cost of moving elsewhere if they change terms or disappear. Most people can recite their revenue and have never written these three down. The second is to read the counterparty’s calendar. Every counterparty has a window when cash or results become scarce — a sales team closing a quarter, a start-up that needs signed contracts before a funding round — and in that window it concedes terms it never would otherwise. You cannot read the counterparty’s books, but you can read its calendar. This is not about squeezing the cornered: Shinsegae’s lever was never a threat to leave but the ability to hand over first what the other side needs now. Delivering early, becoming the first case, committing volume ahead of time cost no cash, and they convert a relationship into a contract instead of spending it on a threat.
Third, take rights, not price. In a favourable window most people cut the fee; this deal teaches the opposite. Trade the discount for exclusivity in a region or sector, the rate increase for term and auto-renewal — a discount reverts at the next negotiation, a right in the contract survives the counterparty’s recovery. Fourth, sequence: the written schedule of what is actually available comes before any money or promise, and empty-looking seats are rarely empty, as Harry Potter’s in Japan and China were not. Fifth, the survival clause, the sentence most often missing from small contracts — this term holds if the contact changes, the company is acquired or the policy is revised — plus the deadline past which silence is consent, because to a small operator a large organisation’s silence is the most expensive cost there is.
One concrete scene. A five-person developer earns 70% of its revenue from a client racing a launch ahead of a large funding round. Demanding a rate increase here is buying the stock: even if granted, it reverts at the next contract. Instead, commit to a shorter schedule and take three years of priority on maintenance work, a clause that the contract survives the client’s acquisition, and the right to resell the jointly built module into other industries. That last right is the developer’s joint venture. No money changed hands; only the kind and order of the asks changed.
Allocate your 100 units of time and money the same way. Half goes to converting what you must do anyway — deliverables, fees, volumes — into rights. Thirty goes to a small table where you hold the majority: not 0.1% of a large company, but operating control of one project built with it. Twenty, and no more, goes to pure bets. The 80 is not safe money; it carries the risk that the counterparty keeps its word, and the difference is that you can manage that risk with a document you wrote. The same logic binds anyone running a co-investment circle: you hold the land, the deals that fill it are borrowed, so ask the originator not for a lower fee but for first look at a defined category, paid for with the participation your members pre-committed. And the deadline is the same. The day every product is built to one platform, or the shop is fitted in one brand’s interior, there is less to offer and nowhere to retreat. Negotiate the rights before the investment sets. Build first and negotiate later, and you remain a tenant; negotiate first and build later, and you are the architect.
Syndicate’s Note
The detonator in this transaction is not inside the shares. It sits on the line that connects the shares to Hwaseong. A USD 12 non-voting share solves nothing Shinsegae needs solved; what does is a schedule of rights obtained before the design sets, and a contractual sentence that keeps those rights alive through the licensor’s distress. If the schedule is empty, the best structure is the one that never enters. The desk’s position is accordingly to reject an equity-only proposal at any price and to reopen the file only when the rights schedule and the survival clauses are in hand.
House Verdict: WATCH. Falsifier: if Shinsegae Property’s participation is disclosed as a straight share purchase with no accompanying written rights package for Korea and the region, the premise that rights rather than shares are what is for sale has been rejected by the one party able to test it; if the disclosure pairs a modest equity line with a territorial rights or joint-venture agreement, the read is confirmed. The verdict classifies structural learning value; it is not an investment judgment.
On the pattern ledger this is a single case built on an undisclosed premise, so it enters the declassification queue rather than the library, provisionally as The Tenant’s Prepayment: a licensee that cannot buy its licensor prepays the royalties it owes anyway to purchase scope, term and survival, takes control only at the level of a regional vehicle, and accepts equity solely as the price of the rights. It is the mirror image of The Tenant’s Buyout, in which the largest licensee buys the licensor outright and pays with the royalties it stops paying. One tenant takes the building. The other, who cannot, writes a lease the landlord’s lenders cannot tear up.
FAQ
What are non-voting shares, and why do they matter in the Skydance deal?
Non-voting shares carry economic exposure to a company’s results but no vote on its management. In Skydance, every voting share is held by the Ellison family and RedBird, while sovereign wealth investors who reportedly contributed about USD 24 billion received non-voting stock and no board seats. A smaller investor can expect exposure to outcomes, not influence over how the company licenses its characters.
Why would Shinsegae Property consider investing in Skydance at all?
Because the theme park at the centre of its KRW 9.5 trillion Hwaseong complex is built on characters licensed from Paramount, which now sits inside a company carrying about USD 79 billion of net debt and USD 6.4 billion of annual interest. The licensee is exposed to the licensor’s balance sheet. The desk’s reading is that this exposure is better handled by prepaying royalties for wider rights and survival clauses than by buying the shares — and participation itself remains an assumption, not a disclosed fact.
What is a licence survival clause, and why does a licensee need one when the licensor is heavily leveraged?
A survival clause keeps a licence in force through events that would otherwise disrupt it: the licensor’s insolvency, a sale of the underlying IP to a third party, or a change of control. When acquisition lenders hold a first lien over the licensor’s assets, an unsecured licensee that has prepaid stands behind them in any restructuring. The clause does not guarantee the money back, but it keeps the right to build alive — which, for a project with KRW 9.5 trillion of land exposure, is the asset that matters.
This article is for informational purposes only and does not constitute investment advice, an offer, or a solicitation. Figures are based on public reports and may be estimates.
For more structural insights and deep-dive video breakdowns, visit Structure Syndicate on YouTube.