What We Said in February
"Classic auction theory predicts this outcome: when one bidder has a structural cost advantage and the other has higher willingness to pay but worse financing, the advantaged bidder wins at a price set by the disadvantaged bidder's maximum."
We predicted Netflix would win the auction using its matching rights, with Paramount's escalation serving only to set the final price between $33-35/share. We identified the DOJ as the most underpriced variable but assumed Netflix would negotiate conditions and close. We called the Polymarket-Kalshi spread correctly — Polymarket's signal was right.
Scorecard:
| Prediction | Result |
|---|---|
| Netflix wins auction at price set by Paramount's ceiling | WRONG — Netflix walked, Paramount won at $31 |
| DOJ is the most underpriced variable | RIGHT — Section 2 monopolization probe forced Netflix out |
| Auction benefits Warner shareholders regardless | RIGHT — $31/share, 147% premium, $2.8B breakup fee |
| Polymarket more accurate than Kalshi | RIGHT — Polymarket had Paramount at 43%, Kalshi had Netflix at 68% |
What Broke and What Held
The auction theory was sound but incomplete. In a clean auction, the bidder with the better balance sheet and matching rights wins every time. Netflix had both. What the model missed: the DOJ turned it from an auction into a regulatory gauntlet. Netflix didn't lose on price. It decided the gauntlet wasn't worth running.
The tell was Netflix's stock. It surged 8.8% the morning they walked away. Investors weren't disappointed — they were relieved. Netflix called Warner a "nice to have, not a must have." That's language you use when you've already decided. They didn't use the four-day matching window. They walked immediately.
The DOJ gave Netflix cover for a retreat it already wanted to make. The deal economics were marginal even before Paramount bid them up. $72 billion for content Netflix could produce in-house over four years at $20 billion per year. The DOJ investigation — Section 2 monopolization, scrutiny of Netflix's leverage over creators, the ousting of the antitrust chief mid-review — made the regulatory timeline uncertain and the political risk real. But the underlying math was the real driver.
The smartest move in the whole saga was walking away. Netflix's stock price said so. The market said so. Every prior acquirer of Warner Bros. has destroyed value. Netflix figured that out in time. Paramount hasn't.
The Reference Class That Should Terrify Paramount
Every time someone has bought a major media company with heavy debt during a content spending war, the debt has won.
Four acquisitions. Four value destructions. Zero exceptions.
The mechanism is the same every time. The acquirer takes on massive debt, which demands cost discipline. But the content market demands spending to compete. Netflix spends $20 billion a year. YouTube is free and infinite. Amazon and Apple subsidize streaming from other businesses and don't need it to be profitable. A debt-loaded acquirer needs to cut costs from day one to service the interest — exactly when it needs to spend more to keep subscribers from leaving for platforms that can outspend it.
Debt demands cuts. Competition demands investment. The company can't do both. Debt wins. Content starves. Subscribers leave. Write-down follows.
AT&T ran this exact playbook at almost the exact same enterprise value ($108.7 billion then, $110 billion now) and wrote down $40 billion. Paramount's leverage is worse — 7x EBITDA versus AT&T's 2.9x. The content war is more intense. The competitor set is richer. And the EBITDA that underpins the multiple is anchored to a linear TV ecosystem that's unraveling faster than the deal can close.
The Structural Trap
Here's the math that keeps Paramount's CFO awake.
At $87 billion in debt and a blended interest rate of roughly 5-6%, Paramount-WBD would owe $4.5-5.2 billion per year in interest alone. That's before any principal repayment, before content spending, before operating the business.
Netflix spends $20 billion a year on content. To stay competitive with Netflix, Amazon, Apple, and YouTube, Paramount-WBD would need to spend at least $15-17 billion. Add $5 billion in interest. Add operating costs. The combined company would need to generate $25+ billion in annual EBITDA just to tread water — and the current combined EBITDA is roughly $12-14 billion.
The gap between what the debt requires and what the business generates is the trap. Every dollar that goes to debt service is a dollar that doesn't go to content. Every dollar not spent on content is a subscriber that leaves for Netflix or YouTube. Every subscriber that leaves reduces the EBITDA that services the debt. It's a reinforcing loop that ends in a write-down or a restructuring.
The RJR Nabisco parallel: KKR bought RJR for $25B in 1988 using extreme leverage during a market where competitors were in a spending war (tobacco marketing). The debt prevented RJR from competing on marketing spend. Market share eroded. The investment was a disaster. Paramount is running the same play in a different industry — and the competitors are richer.
The Contrarian Case: AI Changes the Math
Galloway identified the one variable that could break the pattern: the Ellisons don't care about the traditional Hollywood cost structure because they believe AI will replace it.
If AI cuts content production costs by 40-60% within three years — and Oracle's infrastructure is the engine — then the debt math changes fundamentally. A show that costs $15 million to produce today might cost $6 million. A film that costs $200 million might cost $80 million. The content spending war becomes a technology war, and Paramount-WBD with Oracle backing has a different kind of advantage than AT&T or AOL ever had.
This is the Ellison thesis: Warner Bros.' library isn't just content — it's training data. HBO's scripts, DC's visual language, decades of film grammar. Feed it to AI models running on Oracle Cloud, and you build a content factory that produces at a fraction of the cost while competitors are still paying human-scale prices.
It's a real thesis. It might even be right eventually. But "eventually" is the problem. The debt payments start immediately. AI content production at scale is 2-4 years away from replacing human production at HBO quality. The interest payments don't wait for the technology to mature.
Every leveraged acquirer in the reference class had a thesis for why this time was different. AOL had the internet. AT&T had distribution synergies. Disney had theme parks and merchandise. The thesis was always plausible. The debt was always impatient.
The Sports Rights Cliff
Even if AI cuts scripted content costs in half, it doesn't touch the fastest-growing line item on the balance sheet: live sports. You can't AI-generate an NFL game. You can't cut the cost of sports rights with a language model. And the rights holders know exactly who has money and who doesn't.
The combined Paramount-WBD has a real sports portfolio. CBS brings the NFL, March Madness, Champions League, PGA, and the new UFC deal. WBD brings the NHL and its half of March Madness. But the NBA — TNT's crown jewel — is already gone, lost to ESPN, NBC, and Amazon. And every remaining property comes with a renewal clock.
The NFL change-of-control clause is the loaded gun. The Paramount acquisition gives the NFL the right to break the CBS deal as early as 2027 — next year. Even without breaking it, the NFL is already in active negotiations demanding over $1 billion more per year. The league knows Paramount is desperate to keep the rights and leveraged to the gills. They will extract maximum value.
Here's the trap: CBS without the NFL is a dramatically less valuable network. TNT without the NHL is a cable channel with no sports anchor. The leagues know this. The sports rights don't just cost more — they create the leverage for leagues to extract even more, because Paramount can't afford to lose them. The debt makes Paramount the weakest bidder and the most desperate buyer at every negotiating table from 2027 forward.
The competitors at those tables:
AI cuts scripted content costs. Sports costs are going in the opposite direction, faster, with no technological solution. The NFL's next deal will cost more. The Champions League renewal in 2030 will cost more. Every renewal cycle is a bidding war against companies with 10 to 100 times the financial flexibility. The AI thesis saves money on one side of the ledger while sports bleeds it faster on the other.
Why Are They Doing This?
If the business case is this weak, the question becomes: what's the real motivation?
Three forces stacked on top of each other, and only one is a business case:
The legacy play. David Ellison dropped out of USC film school, funded a box office bomb with his dad's money, built Skydance into a real but inconsistent studio — five of his six biggest hits are Tom Cruise movies — and is now assembling the largest media empire since the Golden Age of Hollywood. Bloomberg reported he "used political ties and deep pockets" to win. He's 43. This is the arc of a billionaire's son who wants to be a mogul. When you have Oracle-level family wealth, the acquisition doesn't need to generate returns. It needs to generate status.
The political alignment. Trump "strongly suggested" he favored a Paramount deal because of the Ellisons' politics. David assured Trump administration officials that "sweeping changes to CNN" would follow. The DOJ ousted its own antitrust chief during the Netflix review, then didn't block Paramount. Larry Ellison has cozied up to Trump. The regulatory path was greased by relationships that Netflix didn't have.
The AI thesis. Oracle infrastructure plus Warner's content library as training data plus AI production cost reduction. This is the only version of the deal that makes financial sense, and it requires AI to mature faster than the debt compounds — while simultaneously ignoring that sports costs, the biggest line item, are immune to AI.
The deal makes sense if you're not trying to make money on the deal. It makes sense if you're a billionaire's son building a legacy, if political access is part of the return, and if the AI thesis is the story you tell investors while the real motivation is something else entirely.
Twenty Days to the Vote
The shareholder vote is April 23. The deal offers $31/share — a 147% premium. The stock trades at $27.22. The $3.78 gap is merger arbitrage pricing the risk that something goes wrong.
What could go wrong:
DOJ blocks or conditions the deal. The antitrust division is issuing subpoenas, won't fast-track, and the California AG is running a parallel review. Paramount cleared the Hart-Scott-Rodino waiting period, but the DOJ can still sue to block. The same DOJ that killed Netflix's deal hasn't approved Paramount's.
Shareholders vote no. The $31 premium is compelling, but some institutional holders may prefer WBD's standalone value plus optionality over cash-and-exit. The proxy fight hasn't started yet.
Financing stress. Larry Ellison personally guaranteed $40.4 billion. Bank commitments from BofA, Citi, and Apollo total $54 billion in debt. In the current rate environment — with oil at $126 and inflation reigniting — credit conditions are tightening. If bank commitments waver, the deal structure cracks regardless of shareholder approval.
The Hitch
In February, we asked who wins the auction. The answer — Paramount — turned out to be the wrong question.
The right question: is winning the auction the same as winning?
The reference class says no. Four of four leveraged media acquisitions during content wars destroyed value. The structural trap — debt demands cuts, competition demands spending — hasn't changed in 25 years. The competitors are richer than ever. The content war is more intense than ever. And the EBITDA supporting the deal is built on linear TV revenue that declines every quarter.
But the sports rights calendar is where the thesis dies. The AI contrarian case has a real logic — cut scripted content costs with technology. Fine. But the NFL is coming for an extra billion dollars next year. The Champions League renewal is in 2030. The NHL is up in 2028. Every one of those negotiations is against Apple, Amazon, Google, and Netflix — companies with 10 to 100 times Paramount's financial flexibility. AI doesn't make sports cheaper. It doesn't give you leverage at a rights negotiation table. And the rights are what make the cable bundle worth anything at all.
Netflix walking away was the real signal. Its stock surged 8.8% the day it declined to match. The market was telling everyone what Netflix already knew: the winner of a Warner Bros. bidding war has always been the loser.
Paramount didn't outbid Netflix. Netflix let them win.
The deeper question is whether this deal was ever about business returns at all. David Ellison is 43, the son of the fifth-richest person in the world, assembling the largest media empire in a generation with political relationships that smoothed the regulatory path. When Oracle-level wealth backs the equity, the deal doesn't need to generate returns. It needs to generate something else — a legacy, political influence, a seat at a table that money alone can't buy.
The reference class covers acquirers who were trying to make money. If the Ellisons aren't trying to make money — if this is a political-media platform play funded by Oracle wealth — then the reference class doesn't apply. The debt still applies. The sports rights still apply. But the willingness to absorb losses for non-financial returns changes the calculus for how long they can sustain it.
That's what makes this different from AOL or AT&T. Those acquirers destroyed value by accident. The Ellisons might be willing to destroy value on purpose — because the returns they're seeking aren't on the income statement.
The Bets
What We're Watching Now
-
April 23 shareholder vote outcome → Approval likely given 147% premium, but watch the margin. A narrow vote signals institutional skepticism.
-
DOJ action before or after close → The same DOJ that killed Netflix's deal hasn't approved Paramount's. Subpoenas issued March 27. If they file suit before close, the deal enters limbo.
-
Bank commitment letters — any conditions or pullbacks → $54B in debt commitments. In the current rate environment, watch for conditions being quietly added.
-
First content budget announcement post-close → The single most important data point. If content spending is flat or up, the Ellison thesis is real. If it's cut 15%+, AT&T pattern is repeating.
-
Netflix subscriber growth in Q2-Q3 2026 → Netflix walked away and kept $72B. If subscriber growth accelerates without Warner's library, it proves the walkaway was the right call.
Portfolio implication: The $3.78 gap between WBD's stock price ($27.22) and the deal price ($31) is merger arb pricing ~12% failure risk. If you believe the deal closes, that's a 14% return in 6 months. If you believe the reference class, short Paramount post-close.
Chapter 2: We asked who wins the auction. The answer was the wrong question. The reference class says winning a Warner Bros. bidding war is losing — four of four leveraged media acquisitions during content wars destroyed value. The AI thesis cuts scripted costs but sports rights are going in the opposite direction — the NFL can break the CBS deal by 2027, and every renewal is against bidders with 10-100x the financial flexibility. Netflix let Paramount win. Its stock surged 8.8% on the walkaway. The deal only makes sense if you're not trying to make money on the deal.