Warner Bros Discovery Inc WBD

Thesis

Warner Bros Discovery is the 2022 marriage of WarnerMedia (HBO, Warner Bros studios, CNN, Turner) and Discovery (Discovery, HGTV, Food Network, TLC). The combined company carries roughly $35B+ of net debt against a melting linear-TV cash cow and a streaming business (HBO Max / Max) that is only just inflecting to GAAP profitability while the cable bundle that subsidizes the whole company unwinds at 6-10% per year.

The scorecard tells the story without ambiguity. 10-year average ROIC is 2.99% — below any plausible cost of capital. Owner earnings TTM are -$4.57B. Share count is up 53.91% over ten years because the Discovery/WarnerMedia merger was paid for in stock. The model's intrinsic-value range is -$82.45 (low), -$63.79 (base), -$44.13 (high). All three numbers are negative, meaning the deterministic DCF says the equity is worth less than zero before any margin of safety. EV/FCF of 21.28 looks reasonable on the surface, but FCF conversion of 2.08x and the scorer's flag that maintenance capex is uncertain (>50% spread) and that NOPAT is declining tell you the FCF print is flattered by working-capital releases and content-amortization shenanigans, not durable cash.

There is no price at which a Buffett-Munger investor should own this on the thesis presented. The composite score of 60 is generous and is propped up by the optical FCF number; the underlying business does not earn its cost of capital and the moat (premium IP, HBO brand) is being eroded faster than the debt is being paid down. A serious investor's question is not 'when do I buy' but 'is the equity a zero in a refinancing window.' Pass.

Moat

Warner Bros Discovery sits on genuinely valuable intangible assets — Harry Potter, DC, HBO, Warner Bros studio library, CNN brand, Discovery/HGTV/Food Network unscripted libraries — but a moat is not the same as an asset. Buffett's 1991 letter [4] is the single most important reference point here: he wrote that 'newspaper, television, and magazine properties have begun to resemble businesses more than franchises,' and that 'the economic strength of once-mighty media enterprises continues to erode as retailing patterns change and advertising and entertainment choices proliferate.' Buffett wrote that in 1991. Thirty-five years later WBD is the case study.

1. Pricing power — NARROW to NONE. HBO Max can raise prices a few dollars a year, but it is one of seven SVOD services Americans rotate through, and churn is high. Linear networks (TNT, TBS, CNN, Discovery, HGTV) are losing affiliate fees as cord-cutting drops US pay-TV households by 6-10% annually. The 1985 Cap Cities/ABC excerpt [5] — Buffett's old standard for a great media franchise with pricing power — is the inverse of what WBD looks like today.

2. Switching costs — NONE for consumers, MODEST for distributors. Damodaran [3] notes Microsoft's switching-cost moat. Streaming has the opposite shape: one click, no contract, no installed base. The studio side has switching costs in talent relationships and franchise IP (a director who built a Batman trilogy doesn't easily move it), but those are negotiated rents, not moat.

3. Network effects — WEAK. Streaming has weak content-flywheel network effects (more subs fund more content funds more subs) but Netflix has decisively won that war at scale. WBD is a sub-scale follower spending into a leader.

4. Intangibles — REAL but commoditizing. The libraries are genuine. Damodaran [2] writes that 'the managers of a firm who take over a valuable brand name and then dissipate its value, will reduce the values of the firm substantially.' That is exactly the AT&T-WarnerMedia-then-Discovery saga: HBO under AT&T was managed for sub growth at any cost, then Zaslav has been cutting and writing off content (Batgirl, Coyote vs. Acme, the CNN+ implosion) to manage debt. The brand survives; the franchise economics do not.

5. Cost advantages — NONE. Content is a scale game and Netflix, Disney, Amazon, and Apple all spend more or have deeper balance sheets. WBD's content spend has been cut to service debt, which compounds the disadvantage.

Competitor stress test ($10B + 5 years): Apple TV+ has spent roughly that and produced exactly one cultural phenomenon (Ted Lasso) plus a string of prestige misses. The lesson is that $10B does not buy you HBO's brand. So the IP itself is moat-like. But the moat that matters for the equity is not 'can someone else recreate HBO' — it is 'can WBD earn a return on capital above its cost of capital with HBO inside it.' The 2.99% 10-year ROIC says no. The intangibles are a moat around assets whose owners cannot capitalize on them under the current capital structure.

Erosion risk: the linear cable bundle — the cash cow that funds everything else — is in secular decline at 6-10% per year. Streaming gross margins are structurally lower than linear network margins ever were, even at scale. The studio business is hit-driven (high variance) and theatrical attendance is structurally below 2019. Every leg of the stool is shorter than it was a decade ago.

Moat verdict: NARROW (and narrowing). The IP is real; the franchise economics around it are not.

Management & Capital Allocation

David Zaslav inherited the WarnerMedia assets via the April 2022 Reverse Morris Trust spinoff/merger and has spent three-plus years doing two things: cutting costs and paying down debt. Both were necessary. Neither has created shareholder value, and the price tag — in dilution, write-offs, and franchise damage — has been substantial.

The five capital-allocation choices, scored:

1. Reinvestment — WEAK. ROIC 10-year average is 2.99%. ROIIC is 'not meaningful' per the scorer because NOPAT declined. By Buffett's basic test, the business should be returning every dollar of marginal capital to shareholders, not reinvesting. Instead WBD is reinvesting heavily into Max (technology, content, international launches) at a return that the market has so far refused to capitalize.

2. Acquisitions — DESTRUCTIVE. The defining capital-allocation event of the last decade is the WarnerMedia merger itself, which more than doubled share count (10-year share count change: +53.91%) and saddled the combined company with ~$55B+ of debt at close. Discovery shareholders bought a melting linear-TV asset and a sub-scale streaming asset at the top of the streaming-at-any-cost cycle. The 2025 announcement to split the company in two (Streaming & Studios vs. Global Networks) is a tacit admission that the 2022 deal logic was wrong. Splitting back apart three years after merging is the textbook signal of a destroyed deal.

3. Debt — DOMINANT. Net leverage was ~5x at close and has been worked down toward ~3.5-4x via aggressive deleveraging, asset sales, and FCF redirection. Management deserves credit for honoring the deleveraging commitment in a rising-rate environment. But this is debt management, not value creation — every dollar of FCF that goes to debt paydown is a dollar that can't go to shareholders or to widening the moat.

4. Buybacks — NONE. Cannot afford them. With negative owner earnings (-$4.57B TTM) and debt-paydown priority, there is no buyback program. Average P/IV when buying is moot. (Note: heavy stock-based compensation continues, modestly diluting the share count even without M&A.)

5. Dividends — NONE since the merger. Suspended.

Communication quality: B-minus. Zaslav is candid about the linear decline and deleveraging math, but the company has had multiple strategic resets — Max name (renamed from HBO Max, then arguably reverting), CNN+ killed at launch, content write-offs for tax purposes, and now a full corporate split — that suggest the strategy is reactive rather than designed. Compensation has drawn shareholder ire (failed Say-on-Pay vote in 2023) and the $50M+/year package while shareholders absorbed 70%+ drawdowns is a poor look.

The Buffett 1991 letter [4] frame is again instructive: 'The fact is that newspaper, television, and magazine properties have begun to resemble businesses more than franchises in their economic behavior.' Zaslav is, fairly, running this as a business — squeezing costs, paying down debt, monetizing IP — not protecting a franchise. A franchise operator would have invested through the cycle to widen HBO's moat. A business operator harvests. The choice is rational for a CEO inheriting too much debt, but it tells you the asset is not what the bulls hope.

Capital allocator: C. Honoring debt obligations is table stakes; he has done that. But the merger that created the problem was the biggest capital-allocation decision of the era for these assets and it was destructive. Splitting the company back apart confirms it.

Industry Structure

Porter's Five Forces on the US/global media-and-entertainment complex in 2026:

1. Rivalry among existing competitors — INTENSE. Netflix is the structural winner of streaming (300M+ subs, dominant content spend, profitable). Disney+, Amazon Prime Video, Apple TV+, Paramount+, Peacock, and Max are all chasing share with overlapping content libraries. Linear networks (CNN, TNT, TBS, Discovery, HGTV) compete with each other and with YouTube, TikTok, and free ad-supported (FAST) channels. Content cost inflation has run double-digit for years; pricing power has been weak.

2. Threat of new entrants — MIXED. New streaming entrants are unlikely (the field is saturated and profitable scale requires $15B+ of annual content spend). But adjacent entrants — YouTube/Alphabet for ad-supported video, TikTok/ByteDance for short-form attention, Roku/Amazon for the FAST/CTV ad layer — have already entered and are taking share of viewership. Apple and Amazon entered streaming with effectively unlimited balance sheets.

3. Bargaining power of suppliers — HIGH. Top-tier showrunners, talent, sports rights holders (NBA, NFL, college football), and franchise IP holders extract most of the economic rent. The $76B NBA rights deal that WBD lost to Amazon/NBC/Disney in 2024 is the prototype: the league captured most of the value, the network owners competed margin away, and WBD lost a 40-year cornerstone of TNT's bundle economics.

4. Bargaining power of buyers — RISING. Distributors (cable MSOs, virtual MVPDs) and now streaming consumers have far more leverage than in the franchise era of the 1980s-2000s. Cable subscribers can cut the cord. Streaming subscribers can pause subscriptions monthly. Advertisers can move dollars to performance channels (Google, Meta, Amazon) that prove ROI. The era of guaranteed affiliate-fee escalators is over.

5. Threat of substitutes — EXISTENTIAL. YouTube alone now accounts for ~10%+ of total US TV viewing time. TikTok captures the youngest cohort. Gaming (Roblox, Fortnite, console gaming) is the dominant entertainment for the under-25 cohort. Sports betting, social media, and AI-generated content are all substitutes for scripted/unscripted entertainment. The Buffett 1991 letter [4] called the early stages of this — 'advertising and entertainment choices proliferate.' That proliferation has accelerated, not slowed.

Value-pool location and trajectory: the value pool that funded WBD's entire existence — the pay-TV bundle, where ~100M US households paid $100+/month and networks shared the affiliate fees — has been collapsing for a decade and is in terminal decline. The new value pools are (a) Netflix-scale streaming, where two or three winners will earn good returns; (b) ad-supported CTV inventory, where Roku/YouTube/Amazon are winning the ad-tech layer; (c) sports rights, where the leagues are extracting the rent. WBD is not the structural winner in any of these pools.

Industry Verdict: Poor. Media-and-entertainment as a whole is a Buffett 'too hard' industry — it has been since 1991 by his own writing. The specific subset WBD operates in (linear cable networks plus second-tier streaming) is structurally worse than the average of the industry.

Inversion (Bear Case)

I am a short-seller. My fund is short WBD and I have spent six months in the 10-K, the trade press, the Nielsen data, and the bond complex. Here is what I see.

1. The single event that kills this. A linear-network goodwill impairment in the next 12-24 months that triggers a debt-covenant renegotiation. WBD took a ~$9B impairment on the TV networks reporting unit in Q2 2024. Goodwill on the linear segment is still substantial. As affiliate revenue continues to decline 6-10% annually and as the planned 2026 split forces fair-value carving of assets, another impairment is highly probable. By itself, an accounting impairment doesn't kill the equity. But it interacts with two things: (a) the indenture covenants on the ~$35B+ of long-dated debt, and (b) credit-rating sensitivity (WBD has been on negative watch repeatedly). One downgrade to non-investment-grade triggers margin calls in index funds, forced selling by investment-grade-only mandates, and a step-up in coupon costs on covenant-protected paper. The equity is the residual claim on a balance sheet where ~$35B of debt sits in front of you. At an EV/FCF of 21.28 with FCF flattered by working capital, a 10-15% FCF decline plus a refinancing-spread blowout collapses the equity-to-EV ratio fast.

2. Why the moat is narrower than bulls think. The bull case is 'HBO + Warner Bros library + DC + Harry Potter is irreplaceable IP.' True, and irrelevant. The question for the equity is not whether the IP exists but whether WBD's capital structure lets it monetize the IP at a return above its cost of capital. The 10-year ROIC of 2.99% answers that decisively: it does not. The Damodaran excerpt [2] is exactly on point — managers can dissipate brand value, and the AT&T-then-WBD chapter has done meaningful damage. Batgirl shelved for a tax write-off. Coyote vs. Acme shelved. CNN+ killed in 30 days. The Snyder-cut/non-Snyder-cut DC reboots. The HBO Max → Max → arguably-reverting brand thrash. Each individual decision is defensible; the cumulative signal to talent and audiences is that this is not a franchise being protected, it is a balance sheet being managed. The Buffett 1985 Cap Cities reference [5] was about a media franchise where management was a multiplier on the assets. WBD is the inverse case.

3. Why management is worse than it appears. Zaslav's debt-paydown narrative is the only thing keeping the equity bid. Look harder. (a) The 2026 split announcement is an admission that the 2022 merger was wrong — and the split itself loads transaction costs, dis-synergies, and probable debt-covenant complexity onto a company that cannot afford either. (b) Compensation: Zaslav has earned ~$200M+ since the merger while equity holders are down 60-70%. The 2023 Say-on-Pay vote failed. The board is not adversarial. (c) Strategy churn: as listed above, multiple high-profile strategic reversals in three years. (d) The bull case relies on Max reaching streaming profitability. It has — barely — but at content-spend levels far below Netflix and Disney+, which is not a recipe for retention or growth. Cutting your way to streaming profitability is exactly the playbook newspapers ran in 2008-2018; it ended in irrelevance.

4. What bulls are extrapolating that won't hold. Three extrapolations: (a) 'Linear declines but stabilizes around a smaller base.' No — affiliate-fee economics have negative operating leverage; once you lose 30-40% of subs, the price-per-sub increases the affiliate negotiates can't keep up, and the cost structure (sports rights, talent contracts, transmission) is largely fixed. The endgame for second-tier linear networks (TBS, TNT without NBA, truTV) is not stabilization, it is bundle exclusion. (b) 'Streaming will inflect and rerate.' Streaming has inflected to GAAP profitability at very modest absolute dollars (~$1B annualized), at content-spend levels the bull case can't easily defend. To grow the streaming business you need to spend more; spending more delays the deleveraging that supports the equity. The optimization is unstable. (c) 'Studio business has option value via a hit cycle.' The studio business has averaged ~$0 EBIT over five years; theatrical is structurally smaller post-COVID; the option is real but priced.

5. Valuation trap — multiple compression / regime change. EV/FCF of 21.28 is not cheap on any historical media multiple — broadcast and cable historically traded 8-12x — and the 'cheap' narrative depends on a bull-case FCF rebound that requires both linear stabilization (it won't) and streaming acceleration without spend increase (it won't). The reverse-DCF the scorer ran returned a negative intrinsic value across all three scenarios (-$82.45 / -$63.79 / -$44.13). That is not a valuation gap, it is a structural mismatch: the operations as currently configured do not support the debt, and the equity is the variance term. P/E TTM is N/A because earnings are negative. Owner earnings TTM are -$4.57B. The composite score of 60 is generous and is the highest the model can charitably produce given this picture.

If I am right, the stock could be worth $5-10 within 24-36 months. The path: another impairment, a downgrade or refinancing scare, a forced asset sale at a discount, and a re-rating of the residual equity to a distressed multiple of a smaller, simpler post-split entity. There is also a non-trivial path to $0 if the split goes badly and the linear sub-entity files.

Lollapalooza Bias Check

Several biases are active in me as I write this and I want to name them so the reader (and I) can discount accordingly.

Recency bias and confirmation bias, in tandem. WBD has been a public-market disaster since the April 2022 merger. The stock is down ~60-70% from the deal close. Every quarter has produced bad news on subscribers, leverage, or strategy. When I read the 10-K I am pattern-matching to a 'this is a bad business' story I have already half-formed, and I am giving disproportionate weight to data points that confirm it. A genuinely fair analyst would weight 2026 streaming-profitability inflection more heavily than I am.

Authority bias — Buffett 1991 [4]. I leaned heavily on a 35-year-old Buffett letter calling media a structurally weakening industry. Buffett has been right directionally — but he was also wrong about Netflix, IBM, and airlines on different timelines. Citing him does not prove the case; it just makes me feel safer making it. I should hold the argument on its own merits.

Anchoring bias — the scorer's negative IV range. The deterministic model says intrinsic value is between -$82 and -$44. That is an enormous anchor against the bull case. But I should remember that the model assumes maintenance capex is high and stable, content-amortization charges are real economic costs, and that mid-teens base CAGR was clamped to 14%. If Max reaches Netflix-like operating leverage at scale (10-year time frame), those assumptions are too pessimistic. The model does not capture optionality.

Social proof / herd. The short side of WBD is consensus. Every value-leaning fundraising deck I have seen for two years either avoids it entirely or explains a creative way to be short. When a trade is consensus short, the asymmetric risk is to the upside on any positive surprise (a rights deal, an unexpected hit, a credible buyer for a piece of the business). I am underweighting that asymmetry.

Deprival super-reaction — the merger thesis. I am, emotionally, frustrated with the 2022 merger. That frustration is doing analytic work it shouldn't. Whether WBD is a sensible investment at $26.97 today is a different question from whether the merger should have happened. I should evaluate the equity from the price I see, not from the 2021 deal price.

Incentive bias on management. Zaslav has made ~$200M+ during a period of large shareholder losses. That fact is doing more work in my analysis than it should. Many CEOs of distressed companies are paid well for a difficult job; that does not by itself make them bad capital allocators. The capital-allocation record on the merger is what condemns them, not the comp.

Net effect: my biases all point in the bear direction. I have tried to size the bull case fairly in the 10-year section. The position guidance below reflects an attempt to discount the bear case for the biases I just named.

10-Year Outlook

Will WBD in 2036 be the same fundamental business it is today? Almost certainly not — and that is the problem. The 10-year test asks whether the business shape is durable. WBD is mid-transformation: a planned 2026 corporate split into Streaming & Studios versus Global Networks; a streaming product (Max) still finding its operating model; a linear-networks business in secular decline; ongoing portfolio reshaping (asset sales, possible studio JV's, possible further M&A on either side of the split).

Will the customer base be larger? On streaming, plausibly yes — Max should grow from ~100M to perhaps 150-180M global subs with international rollouts. On linear, definitively no — US pay-TV households decline 6-10% per year and there is no path to stabilization at a smaller, profitable base.

Will profit per customer be higher? Streaming ARPU is structurally below linear-bundle ARPU; the math of the transition is permanently lower revenue per viewer-hour. Operating margins on streaming at scale (Netflix-comparable) are 20-25%; legacy linear network margins were 35-45%. The mix shift is dilutive to per-customer economics for the foreseeable future.

Will the moat be wider? Almost certainly narrower. The IP library is a fixed asset; the competitive landscape (Netflix, Disney, Amazon, Apple, YouTube, TikTok) is more intense; talent costs are higher; bundle economics are gone.

Single biggest threat: the interaction of debt service and revenue decline. If linear-network revenue falls faster than streaming and studio EBITDA grow, FCF declines, debt-paydown stalls, refinancing spreads blow out, and the equity becomes the residual claim on a deteriorating balance sheet. The split is intended to ring-fence this, but a split also forces fair-value testing of assets and may catalyze the impairment / downgrade scenario.

This is exactly the kind of business where you cannot say with high confidence what the 2036 income statement looks like. Step 4 of the methodology — Munger's circle-of-competence test — flags this: 'does it require predicting tech adoption curves / regulatory outcomes / consumer fads.' WBD requires predicting cord-cutting trajectory, streaming-market structure, AI's impact on content production, and consumer entertainment attention allocation. All four are uncertain.

CONFIDENCE: low