News Corp Class B NWS

Thesis

News Corporation Class B (NWS) is the voting share class of a Murdoch family-controlled media conglomerate spun out of 21st Century Fox in 2013. The asset mix is Dow Jones (Wall Street Journal, Barron's, Dow Jones Risk & Compliance, Factiva), Digital Real Estate Services (REA Group ASX listing plus Realtor.com via Move), Book Publishing (HarperCollins), Subscription Video (Foxtel, recently sold to DAZN), and News Media (The Sun, The Times, New York Post, The Australian).

The Buffett-Munger lens is unsparing here. The composite scorecard score is 64/100. Trailing P/E is 41.88 against a 10-year average of 51.04 - both extreme for a media holdco whose 10-year average ROIC is 0.0% and whose maintenance-capex spread is wide enough that the scorer flagged the IV range. Reverse DCF implies the market is pricing in 14.7% perpetual growth, which is roughly four times the secular growth rate of any sub-business inside the wrapper. Owner earnings TTM are $0.32 billion against a price of $30.41 and an enterprise value commensurate with that.

The IV range from the deterministic scorer is $6.38 (low) to $8.08 (high). Today's price of $30.41 is 4.77x even the high end. Even crediting a holdco discount unwind, a Dow Jones spinoff catalyst, or an REA stake mark-to-market, you cannot bridge to today's quote without giving Murdoch credit he has not earned in thirteen public years. The 0.17% net share count change over a decade tells you buybacks are not the story either. At $30.41 against an $8 IV ceiling, this is a Sell at the index level. A constructive buy zone would require a Dow Jones standalone print and a price closer to $12-$14.

Moat

News Corp is not a single business; it is at least five businesses bolted together. Moat analysis must be done segment by segment.

Pricing power. Dow Jones (WSJ, Barron's, Factiva, Risk & Compliance) has genuine pricing power. The WSJ runs ~4 million digital subscribers at price points that have moved from $19/mo to $39+/mo without meaningful churn shock; B2B Risk & Compliance and Factiva sell into bank compliance and legal-information workflows at prices a customer cannot bargain on the margin. HarperCollins has limited pricing power because Amazon controls the channel. News Media (Sun, Times, Post, Australian) has almost none - readers are price-takers but advertisers are not, and the secular decline continues. Foxtel sale to DAZN removes a chronically pressured pay-TV asset.

Switching costs. Dow Jones B2B (Factiva, Risk & Compliance, OPIS energy data, Dow Jones Newswires) is the strongest claim in the entire holdco. These products integrate into compliance workflows, screen for sanctions and PEPs, and feed terminals; a Tier-1 bank does not rip Factiva out without budgeted multi-quarter migration. Realtor.com has no switching cost - agents post listings everywhere. WSJ subscribers churn modestly but the bundle (WSJ + Barron's + MarketWatch) raises stickiness.

Network effects. Realtor.com is the #2 US listings portal behind Zillow and is sub-scale; the network effect runs against it, not for it. REA Group in Australia is the dominant portal - that is a real two-sided network and a genuine moat, but Class B holders own it through a 61% stake, not directly. The WSJ has reader/advertiser circularity that Buffett described as 'survival of the fattest' [3], but the internet broke that circularity in financial news long ago.

Intangible assets. WSJ brand, Barron's brand, HarperCollins backlist (Tolkien estate, Agatha Christie, C.S. Lewis Narnia rights), The Times of London masthead. These are real but Buffett warned starting in 1991 [1] that media intangibles have shifted from 'franchise' to 'business' economics: 'gone are the days of bullet-proof franchises and cornucopian economics.' He doubled down in 2006 [4]: 'When an industry's underlying economics are crumbling, talented management may slow the rate of decline. Eventually, though, eroding fundamentals will overwhelm managerial brilliance.' That is the precise diagnosis for NWS's News Media segment.

Cost advantages. None of consequence. Print is a high-fixed-cost business in structural decline. Digital distribution is a level playing field, not a moat - Bloomberg, Reuters, NYT, FT all compete head-on with WSJ in financial news, and OpenAI/Anthropic ChatGPT/Claude integrations are now an existential pressure on the per-article-economics of all of them. The recent OpenAI content-licensing deal monetizes some of this but at a fraction of the per-reader value being lost.

$10B / 5-year stress test. If a $10B competitor entered tomorrow: (a) WSJ - hard to dislodge in US business news but Bloomberg and FT already do this with deeper pockets; (b) Realtor.com - $10B could fund several years of CompasMatch-style price war that erodes Move's share further; (c) HarperCollins - publishing is fragmented and Penguin Random House already dwarfs HC; (d) REA - effectively unassailable in AU short of a sovereign-backed entry; (e) News Media tabloids - already losing to TikTok and YouTube without any competitor needing to spend $10B.

Erosion risk. High in News Media, moderate in HarperCollins, low in Dow Jones B2B, asymmetric in Realtor.com (LLM answer-engines may eat top-of-funnel traffic), low in REA Group. Buffett 1991 [1]: 'The intrinsic value of this investment has declined materially because of the secular transformation that the industry is experiencing.' That sentence is older than most analysts on this stock and it applies more, not less, to NWS today.

The holdco aggregates one wide-moat asset (Dow Jones B2B), one regionally wide-moat asset (REA Group, only 61% owned), one narrow-moat consumer brand (WSJ direct subs), and three structurally challenged businesses (HarperCollins, News Media, what's left of Foxtel-related obligations). The conglomerate wrapper itself has negative moat properties: it traps capital, blocks the focused operator who could maximize each piece, and lets the family extract control rents through dual-class voting.

Moat verdict: NARROW (held back from WIDE by the wrapper).

Management & Capital Allocation

The capital allocator under examination is the Murdoch family - Lachlan as Executive Chairman/CEO, Robert Thomson as CEO of News Corp - operating through dual-class voting where NWS Class B carries the votes that NWSA Class A does not. The succession trust litigation in Nevada in late 2024 made explicit what was already true: this is a family business whose strategic decisions are made to preserve family control across generations, not to maximize per-share intrinsic value of either share class.

Let me grade the five capital-allocation choices.

Reinvest in the business. Mixed. Investment in Dow Jones professional information (Risk & Compliance growth ~10%+, OPIS energy data, factiva) is the best capital allocation in the company and has compounded high-single-digit revenue. Investment in Realtor.com has been less rewarded - share has been lost to Zillow over the decade. News Media reinvestment is largely defensive maintenance.

Acquire. History is mixed. The 2007 Dow Jones acquisition (under the predecessor entity) at ~$5B was a strategic masterstroke and is the foundation of today's best segment. Move/Realtor.com acquisition in 2014 has been a structural disappointment relative to the Zillow comparable. Foxtel was a multi-decade money pit finally being exited via DAZN. OPIS and Investor's Business Daily were good tuck-ins. Net: lumpy, neither destructive nor accretive on aggregate.

Debt. Conservative. Net-debt-to-EBITDA of -0.26 - the company is net cash. The newly-amended March 2026 credit agreement is $1.5B unsecured ($1B revolver, $500M Term A) with an AOI net leverage covenant of 3.5x. That is conservative and gives the company optionality. Buffett 1991 [1] specifically praised Cap Cities and Washington Post for being 'essentially free of debt' in a deteriorating media industry; News Corp's balance sheet would meet that bar. This is the single brightest spot in the management report card.

Buybacks. Negligible. Share count change over 10 years is -0.17% - effectively zero. For a stock trading at 4.77x estimated IV (i.e., 79% premium to the high IV), refusing to repurchase shares aggressively is correct on price and timing. But it also means there has been no meaningful return of capital via shrinking the float, and there has been no meaningful dilution either - the shares simply sit. A buyback program at the right price would help; today's price is not it.

Dividends. A token ~1.3% yield. The company does not return material capital. Cash builds.

Communication quality. Robert Thomson is articulate and the proxy materials are above industry average. Segment disclosure is reasonable post-Foxtel sale. The Murdoch family communication is largely opaque - succession plans, the future of the Class A vs. Class B structure, and any potential separation of Dow Jones (a recurring activist ask, including from Starboard Value in 2023) are addressed only when the family chooses to address them.

The dual-class issue. NWS (Class B, voting) and NWSA (Class A, non-voting) typically trade within a few percent of each other. Damodaran's voting-control work [4] notes that two share classes with identical economics often trade similarly when control is not contestable - and Murdoch Family Trust holding precludes contest. The voting premium for NWS is therefore largely theoretical absent a control event.

The Starboard Dow Jones spin question. Activists have repeatedly pointed out that a standalone Dow Jones could trade at multiples (>20x EBITDA, comparable to other professional information assets) that imply News Corp's other segments are valued at less than zero in the holdco. Management has consistently declined to separate. This is the single largest indictment of capital allocation: refusal to do the structural action that would close the price-to-IV gap, because doing so would dilute family control.

ROIC and ROIIC. 10-year average ROIC is 0.0% - this is not a number you see at a high-quality compounder. The scorer notes 'NOPAT declined; ROIIC not meaningful.' That is the smoking gun: the wrapper has consumed the value created inside Dow Jones.

Capital allocator: C (would be B+ at the segment level for Dow Jones; would be D for the holdco wrapper; netting to C).

Industry Structure

News Corp spans four distinct industries; Five Forces analysis is segment-weighted.

Threat of new entrants. High in news/media (substack, AI-assisted content creation, podcast networks, TikTok-native publishers, ChatGPT/Claude/Perplexity answer engines that disintermediate publishers entirely). Low in financial professional information (Bloomberg, Refinitiv, FactSet have entrenched terminal economics; Risk & Compliance has regulatory tailwinds and integration depth). Moderate in real-estate portals (Zillow already established; CoStar via Homes.com is throwing money at the category). Moderate in book publishing (self-publishing via Amazon KDP has eaten the lower tier of trade publishing; the Big Five remain).

Bargaining power of suppliers. Suppliers in media are talent (writers, editors, on-air) and content rights. Talent has gained leverage with Substack-style direct alternatives. AI training-data licensing has emerged as a new revenue line for content-rich publishers (NWS signed an OpenAI deal in 2024) but the long-term economics depend on whether AI search reduces clickthrough by more than license fees compensate. Net: suppliers gaining power.

Bargaining power of buyers. In B2C news/digital subs: high - one cancellation away from churn. In B2B (Factiva, Risk & Compliance, Dow Jones Newswires): low - integrated workflows, multi-year contracts, compliance mandates. In real estate: agents have power because they have multiple portals to choose from. In book retail: Amazon dominates and sets terms. Mixed but trending unfavorable in B2C, favorable in B2B.

Threat of substitutes. Severe. Buffett wrote in 2006 [4]: 'if cable and satellite broadcasting, as well as the internet, had come along first, newspapers as we know them probably would never have existed.' The 2025 update to that line is: if generative AI search had come along first, even WSJ-style aggregator portals would never have existed. Substitutes for general news are essentially infinite (social, streaming, AI). Substitutes for B2B financial information are narrower (Bloomberg, Reuters, FactSet). Substitutes for real-estate portals exist but are differentiated. Substitutes for trade books are alternative entertainment (streaming, gaming).

Rivalry. Intense in news (NYT, FT, Bloomberg, WaPo, Axios, Politico, plus AI). Intense in real estate (Zillow + CoStar). Moderate in B2B financial information. Moderate in book publishing (5 majors).

Value pool location and trajectory. The value pool in the broader media stack has migrated away from publishers and toward platforms (Google, Meta, Amazon, Apple, TikTok) and toward AI inference (OpenAI, Anthropic). Within publishing, the value pool has consolidated into a few brands with subscriber economics (NYT being the case study; WSJ being a credible second). In real estate digital, value has consolidated to Zillow with Realtor.com a structural #2. In professional information, the value pool has been held by Bloomberg/Refinitiv/FactSet with Dow Jones a focused niche player.

Aggregate verdict. The industry economics are average-to-poor for most of the holdco and good for the Dow Jones B2B segment specifically. Buffett's 1991 framing [1] - 'media properties have begun to resemble businesses more than franchises' - is now thirty-four years old and has only intensified. The B2B information segment is the exception within the exception.

Industry Verdict: Average (skewed by one Good segment and three Poor segments).

Inversion (Bear Case)

I am now playing the short-seller. I am not hedging. I am not defending a balanced view. The objective is to articulate why this stock is genuinely a Sell from $30 today.

Section 1 - The single event that kills this. The single event is generative AI achieving the inflection where users get their news, financial summaries, and real-estate research from an LLM-based answer engine rather than from publisher destinations. We are inside that transition right now. WSJ's traffic from Google Search is already declining as AI Overviews intercept queries. Realtor.com's value depends on consumers landing on Realtor.com to research listings; if Claude or ChatGPT or Google's AI mode synthesizes the equivalent answer from listing-feed data, the portal monetization model collapses on traffic, not on listings. The 2024 OpenAI content licensing deal generates one-time-flavored cash but does not replace the per-impression advertising and subscription economics being eroded. The 'event' is not a single day; it is the realization across 2025-2027 that engagement on publisher properties is in absolute decline at a rate that subscription gains cannot offset.

Section 2 - Why the moat is narrower than bulls think. Bulls point to WSJ's 4M+ digital subscribers and Dow Jones Risk & Compliance growth as proof of moat. Counter: WSJ's subscriber base is a fraction of NYT's, and NYT is the more focused operator with the better technology stack and the deeper bench in product. WSJ's subscription growth has plateaued; price increases are doing the heavy lifting and there is a ceiling on that. Risk & Compliance is real but small relative to the Bloomberg/Refinitiv/FactSet incumbents and competes at the lower end of the workflow stack. Realtor.com bulls point to Move's market position as a stable #2; counter: stable #2 in a category that Zillow has been relentlessly extending away from while CoStar bombs in money via Homes.com is not a moat - it is a sandwich. HarperCollins bulls point to backlist value (Tolkien, Lewis, Christie); counter: rights are licensable, not exclusive, and AI-generated companion content is on the horizon. The composite picture is not a wide moat; it is one decent professional-information business inside a company priced as if the whole thing were Bloomberg.

Section 3 - Why management is worse than it appears. Three reasons. First, the 10-year ROIC of 0.0% is not a measurement glitch - it is the literal output of a decade of capital allocation. NOPAT declined over the period per the scorer notes. A company that earns its cost of capital sometimes is not a compounder; a company that does not earn it on average over a full decade is a value destroyer with a brand. Second, share count change of -0.17% over a decade in a stock that has frequently traded below the company's own internally-implied IV is a refusal to use the most powerful capital-allocation lever available. Third, the refusal to spin Dow Jones - despite repeated activist pressure - tells you the family's objective function is control preservation, not per-share value maximization. The Nevada succession trust trial of 2024 made this textually explicit in court filings. Management is not bad in the operating sense; they are misaligned in the structural sense, which is worse for a long-term minority owner because no operating improvement can fix it.

Section 4 - What bulls are extrapolating that won't hold. Bulls extrapolate three things that I do not believe will hold. (a) That Dow Jones professional information growth scales linearly into a Bloomberg-like multiple if separated. The TAM for compliance and corporate intelligence is large but Bloomberg's moat (terminal hardware/software bundle, chat ubiquity, sales force depth) is not what Dow Jones has. A spun Dow Jones probably trades at 15-18x EBITDA, not 25-30x. (b) That Realtor.com benefits from cyclical recovery in US housing transaction volumes. It will, but the share trajectory vs. Zillow has been negative for a decade and AI-mediated home search may compress the entire portal category before the cycle helps. (c) That the Foxtel divestiture and Move improvements 'streamline' the company into something coherent. The wrapper is the problem. Streamlining the wrapper does not unwrap it.

Section 5 - Valuation trap (multiple compression / regime change). P/E TTM is 41.88 against a 10-year average of 51.04 - a metric the scorer cites verbatim. EV/FCF is 96.29. Reverse DCF implied growth is 14.7% in perpetuity. None of those numbers are defensible for a holdco whose underlying revenue grows low single digits and whose ROIC is zero. The IV range from the deterministic scorer is $6.38 to $8.08. Today's price is 4.77x even the high end. The setup is exactly the pattern Buffett described in 1991 [1]: 'expectations for media have moved toward the bob-around model. And, as our simplified example illustrates, valuations must change dramatically when expectations are revised.' That valuation change is the trap. As AI-search disintermediation becomes undeniable across 2026-2027, the market repricing will not be gentle. A move from 96x EV/FCF to a more defensible 18-25x EV/FCF is a 70-80% drawdown in price holding owner-earnings flat. If owner-earnings also compress (which the scorer flagged as the trajectory), the drawdown is larger.

If I am right, the stock could be worth $7-$10 within 3-5 years.

Lollapalooza Bias Check

Active biases in me as the analyst right now.

Anchoring. I am anchored to the deterministic scorer's IV of $6.38-$8.08. That number is the output of one model with stated maintenance-capex uncertainty and a clamped CAGR (the scorer 'clamped from -6.8% to -5.0%'). It is the best estimate I have but it is not God. If I un-clamp the CAGR or assign a separation catalyst probability, the IV widens upward materially. I should hold the IV as a central tendency, not as a hard ceiling.

Authority bias. I am leaning hard on Buffett's 1991, 2006, and 2012 letters because they explicitly diagnose newspaper-industry decline. Buffett was right. But Buffett also bought 28 newspapers in 2012 [2] because 'their economics make sense' at the right price. The lesson is not 'avoid all newspaper-adjacent equities forever.' The lesson is 'price determines whether bad-economics businesses are buys or sells.' At $7 NWS this becomes a credible deep-value play; at $30 it is a Sell. I am at risk of using Buffett's authority to skip the price discrimination he himself practiced.

Confirmation bias. Every data point I surfaced (ROIC 0%, P/E 41.88, EV/FCF 96.29, dual-class control structure, refusal to spin Dow Jones) lines up against the bull case. I deliberately did not surface counter-evidence with the same energy: the OpenAI licensing deal, Risk & Compliance growth rates, REA Group's AU dominance, the Foxtel divestiture as cleanup. A balanced researcher would weight those harder.

Recency bias. I am giving heavy weight to the AI-disintermediation thesis because it is the loudest current narrative in publishing. AI may not eat publisher economics on the timeline I implied. The transition could take 5-10 years rather than 2-4. NYT continues to grow subscribers in this environment; WSJ may too.

Authority bias (control structure). I am pattern-matching the Murdoch dual-class structure to the worst archetype (family control destroys minority value). The pattern often holds (see Comcast, Fox, NYT itself), but family control sometimes preserves long-term thinking that quarterly-driven public markets erode. Berkshire is the canonical example. Murdoch is not Buffett, but the pattern-match cuts only one way and I am cutting it that way.

Deprival super-reaction. None active - I do not own this stock and have nothing to lose by recommending against it.

Incentive bias. A Compounder framework rewards finding clean wide-moat compounders at attractive prices. NWS is structurally not that. The framework's incentive is to land on 'Avoid/Sell' for messy holdcos, which is the conclusion I reached. That alignment between framework incentive and conclusion is itself something to flag.

10-Year Outlook

Will News Corp look like the same business in 2036? Substantially yes for Dow Jones (still selling business news and B2B information), substantially yes for HarperCollins (still publishing books), probably no for Realtor.com (the portal model may be subsumed by AI agents acting on listing feeds), and probably yes-but-shrunken for News Media (Sun, Times, NY Post, Australian still publishing but with smaller economic footprint).

Will the customer base be larger? For Dow Jones B2B - yes, regulatory complexity continues to drive Risk & Compliance demand. For WSJ digital subscribers - flat to modestly up if pricing power holds; the subscriber TAM in English-language financial news is finite. For Realtor.com - probably smaller in user-engagement terms even if listing inventory is stable, because AI may answer the search before the user lands on the portal. For HarperCollins - flat in unit terms, with AI/audio shifting the mix. For News Media - smaller.

Will profit per customer be higher? For Dow Jones B2B - yes, with mid-single-digit pricing. For WSJ - yes via price. For Realtor.com - uncertain, depends on agent ARPU which is contested with Zillow. For HarperCollins - flat. For News Media - down.

Will the moat be wider? Dow Jones B2B - moderately wider as compliance-data integration deepens. WSJ - flat to narrower as AI compresses news consumption. Realtor.com - narrower as Zillow extends. HarperCollins - flat. News Media - narrower.

The single biggest threat is generative AI restructuring how end-users access news, financial information, and real-estate research. The second-biggest threat is the Murdoch family's continued resistance to structural separation, which means the wrapper-discount/premium dynamic persists indefinitely.

Because (a) the underlying segments diverge sharply in trajectory, (b) the structural separation question is gating to value realization and is in the family's hands, and (c) AI's effect on publisher economics is a wide error band over the next decade, my confidence that I can forecast the 10-year picture for this specific holdco is medium at best. The recommendation does not require high confidence to be Sell at 4.77x IV; it would require high confidence to be Buy.

CONFIDENCE: medium