Comcast is a vertically integrated media and technology conglomerate that derives revenue from two fundamentally different engines: a regulated-adjacent infrastructure business (cable / broadband / wireless) and a content / experiences business (NBC, Peacock, Universal Studios, theme parks, Sky). The infrastructure arm is a quasi-monopoly over its service footprint; the content arm competes in one of the most dynamic, capital-intensive industries on earth. That structural duality is both the company's greatest strategic asset and its most persistent analytical puzzle for investors.
Comcast completed the spinoff of Versant Media Group (CNBC, MSNBC, USA, E!, Syfy, Golf Channel, Fandango, Rotten Tomatoes) on January 2, 2026. What remains after Versant is a cleaner, higher-growth asset mix. The segment breakdown below reflects the continuing business on a 2025 annualized basis:
| Segment | ~Revenue | % of Total | Adj. EBITDA Margin | Growth Trend |
|---|---|---|---|---|
| Residential Connectivity & Platforms Broadband, video, wireless (Xfinity) |
~$60B | ~48% | ~38–40% | Declining subs; ARPU pressure |
| Business Services Connectivity Comcast Business, SMB to enterprise |
~$10B | ~8% | ~57% | +5–6% YoY; high-margin growth |
| Media NBC, Telemundo, Peacock, Bravo (retained) |
~$23B | ~19% | Low-mid teens | Peacock losses narrowing fast |
| Studios Universal Pictures, DreamWorks Animation, Focus Features |
~$10B | ~8% | ~15–20% | Cyclical / IP-driven |
| Theme Parks Universal Orlando, Hollywood, Japan, Beijing, Singapore; Epic Universe (May 2025) |
~$9–10B | ~8% | ~35–40% | +19% YoY Q3 2025; Epic Universe inflection |
| Sky (International) UK, Italy, Germany — video, broadband, streaming |
~$18B | ~15% | Mid-high teens | FX-pressured; structural shifts |
Roughly 70–75% of revenue is subscription-based or recurring — broadband, wireless, business services, Sky. These customers pay monthly, often under implicit lock-in enforced by infrastructure dependency and bundle stickiness. Transactional revenue (box office, advertising, theme park admissions) makes up the balance and is more cyclical. The company's FCF-to-revenue ratio of approximately 15% reflects the high fixed-cost nature of infrastructure but excellent cash conversion once built.
Comcast raised broadband prices consistently for a decade, with ARPU growing 3–4% annually even while adding modest volumes. That era is now under direct challenge. In a strategic pivot announced in late 2025, management is holding broadband prices flat in 2026, bundling free mobile lines as a defensive retention weapon. This is a significant sign that the legacy pricing-power moat has eroded in its core residential broadband segment, even if business services retains it. Theme parks and Studios IP retain strong pricing power.
The United States accounts for approximately 75–78% of revenue. Sky adds European exposure across the UK, Germany, and Italy. No single customer represents more than 1–2% of revenue; revenue concentration risk is geographic and regulatory, not customer-based. The entire domestic cable footprint — roughly 63 million homes and businesses passed — represents a structural monopoly-duopoly in most markets, though that is changing rapidly as fiber overbuilders expand.
| Metric | FY2021 | FY2022 | FY2023 | FY2024 | FY2025E |
|---|---|---|---|---|---|
| Revenue ($B) | $116.4 | $121.4 | $121.6 | $123.7 | $123.7 |
| Adj. EBITDA ($B) | $34.6 | $36.3 | $37.6 | $38.1 | ~$38.5 |
| Adj. EBITDA Margin | 29.7% | 29.9% | 30.9% | 30.8% | ~31% |
| Operating Margin | ~17% | ~16% | ~17% | ~17% | ~16.7% |
| Net Margin | ~13% | ~13% | ~13% | ~13% | ~16% (Hulu gain) |
| C&P EBITDA Margin | ~37% | ~38% | ~39.5% | ~41.5% | ~40% (pricing reset) |
Margin trend: Connectivity & Platforms margins expanded sharply through 2024, hitting a record 41.9% in Q2 2024. The 2025–2026 broadband pricing reset has begun reversing this, with Q3 2025 margins at 39.7%. The reset is intentional and strategic — sacrificing short-term ARPU to stabilize the subscriber base — but it will suppress EBITDA growth through at least 2026.
Comcast generated $12.5B in FCF in FY2024 and an estimated $19–22B in FY2025, the latter inflated by the $9.4B Hulu stake sale gain. Normalizing for that one-time item, underlying FCF is approximately $12–14B annually — among the highest absolute FCF pools of any publicly traded company. Operating cash flow consistently and materially exceeds reported net income, with an earnings quality ratio of ~1.68×, indicating high-quality, cash-backed earnings. FCF conversion of adjusted net income runs 80–90%+.
Comcast carries total debt of approximately $98.9B (Dec 2025), with ~$9.5B in cash, yielding net debt of ~$89B. At an annualized EBITDA of ~$38.5B, net debt-to-EBITDA sits at approximately 2.3× — elevated but manageable for an infrastructure business with highly predictable cash flows. Critically, the weighted average interest rate on this debt is just 4%, fixed-rate, with a long average maturity. There is no near-term refinancing cliff risk, though the current portion of debt (due within 12 months) has risen to ~$6B in 2025, warranting monitoring.
While the leverage ratio is manageable, $98.9B in gross debt leaves limited balance sheet flexibility for large-scale offensive M&A. The Altman Z-Score (1.39) places Comcast in "distress zone" by traditional metrics — but this measure is not well-calibrated for large infrastructure businesses with negative working capital by design (subscription payments are received upfront). Do not over-weight this signal, but do not dismiss it.
Comcast spent approximately $11.8B in capex in FY2025. Maintenance capex for the existing cable plant is estimated at $5–6B annually; the remaining $5–7B is growth capex for network upgrades (DOCSIS 4.0, mid-split), Epic Universe construction (now largely complete), and new broadband passings. With Epic Universe open and the network upgrade program scheduled for ~75% completion by end-2025, capex intensity is expected to decline in 2026, meaningfully expanding "clean" FCF.
As a subscription business, Comcast receives cash before delivering most services (positive working capital cycle). Residential broadband, wireless, and Sky subscriptions are largely prepaid monthly. Theme parks operate on a combination of ticketing and seasonal spending. The result is a structurally negative working capital position that acts as a source of cash — similar to Amazon's retail float or Netflix's subscription model.
ROIC (including goodwill) is estimated at approximately 8–10%, broadly in line with the company's weighted average cost of capital. This is arguably the most important long-term warning sign: while Comcast generates enormous absolute FCF, it has not historically earned a substantial spread above WACC. The Sky acquisition (2018, $39B) has been a persistent drag on ROIC, as that business has struggled to meet original growth expectations. Return on equity of ~20% (2025) looks strong but is amplified by leverage.
Brian L. Roberts (age ~67, Chairman & Co-CEO) is the son of Comcast founder Ralph Roberts and has led the company since 1990. He has overseen transformative acquisitions including AT&T Broadband (2002), NBCUniversal (2011), and Sky (2018). Roberts controls the company through a dual-class share structure — holding approximately 1% of equity but 33% of voting power through Class B supervoting shares. This makes the company effectively unsellable against his wishes and insulates management from activist pressure. Roberts owns approximately 23.5 million shares worth roughly $585M at current prices — meaningful but modest relative to his voting control. His net selling of 469,515 shares in late 2024 is not a large insider sell signal, but it is not a buy signal either.
Michael J. Cavanagh (age ~59, Co-CEO) joined as CFO in 2015 and was elevated to President in 2022, now Co-CEO from January 2026. Cavanagh's background is in financial services — he was Co-CEO of JPMorgan Chase's Corporate & Investment Bank and JPMorgan CFO through the financial crisis — rather than in media or technology operations. He oversaw the Versant spinoff and has been the architect of the current strategic pivot. His 2025 compensation of $71.8M (including $60M in stock) is extraordinarily high and signals both his importance to the board and the board's desire to retain him through a difficult transition. Insider selling: net sale of 191,106 shares over the past 18 months — a mild negative signal.
Jason Armstrong was promoted to CFO in early 2024 after serving as Deputy CFO. His financial communications with Wall Street have been notably direct, flagging ongoing competitive pressures without spin. Armstrong's $27,742 share sale in recent months is immaterial.
Board independence is structurally compromised by the dual-class share structure. Brian Roberts controls the outcome of any vote, regardless of institutional shareholder sentiment. The Chairman/Co-CEO role is not separated. This is a perpetual governance discount that investors in Comcast must accept. In practice, Roberts has been a reasonably disciplined capital allocator — avoiding ruinous overpayment on deals outside Sky — but the lack of checks creates tail risk if judgment deteriorates.
Roberts' 2025 pay of $35.1M and Cavanagh's $71.8M are high in absolute terms. However, pay is substantially equity-linked, and Cavanagh's $60M stock award at these prices represents real alignment — he needs the stock to recover for that compensation to be worth its headline value. The board approved a $15B share repurchase authorization in early 2025, suggesting alignment with shareholder returns over empire-building.
The NBCUniversal acquisition ($30B, 2011) has proven value-creative in hindsight — particularly Studios, Peacock infrastructure, and theme parks. The Sky acquisition ($39B, 2018) is more mixed: Sky generates consistent cash flow but never delivered the synergies promised, and the valuation paid was arguably top-of-cycle aggressive. The Versant spinoff of legacy cable networks in January 2026 is a rational, disciplined move — shedding structurally declining assets while retaining the high-quality ones. Share buybacks have been substantial ($7.2B in FY2025) and at prices well below intrinsic value estimates, creating meaningful per-share value.
Yes — but it is fracturing in its core residential broadband segment, while remaining robust in Business Services and increasingly valuable in Theme Parks and Studios IP.
| Segment | Moat Type | Strength | Trend |
|---|---|---|---|
| Residential Broadband | Geographic scale, infrastructure | Medium | Eroding — fiber and FWA |
| Business Services | Switching costs, bundled contracts, network scale | Strong | Widening |
| Theme Parks | Intangible assets (IP), location, capital barriers to entry | Strong | Widening with Epic Universe |
| Studios (Universal) | IP (Fast & Furious, Jurassic, Despicable Me, Wicked, etc.) | Strong IP | Stable but cyclical |
| Peacock / NBC | Sports rights (NFL, Olympics, NBA), broadcast scale | Medium | Stable; NBA rights accretive |
| Sky | Aggregation, brand; weaker in OTT transition | Weakening | Structural transition risk |
The residential broadband "natural monopoly" moat — built on the physical reality that fiber deployment is expensive and time-consuming — is being systematically eroded by two forces. First, fiber overbuilders (AT&T, Frontier, Crown Castle-backed startups) are now competing in a growing share of Comcast's footprint. Second, fixed wireless access (FWA) from T-Mobile and Verizon offers "good enough" broadband for a large price-sensitive segment. Comcast passed 63 million homes and businesses; its fiber competitors' combined footprint in those markets is smaller, but growing at 20%+ annually. The company's domestic broadband base declined by approximately 750,000 subs from peak through 2025. The loss rate improved to 65,000 in Q1 2026 — the first year-over-year improvement in six years — which is a critical positive data point, but recovery is not guaranteed.
Business Services is the most under-appreciated moat in the Comcast portfolio. Comcast Business generates ~$10B in annual revenue at ~57% EBITDA margins and is growing mid-single digits. Enterprise and mid-market customers have high switching costs (multi-year contracts, proprietary network integrations), and Comcast's footprint of 63M+ homes and businesses passed gives it dense metro connectivity that fiber-only competitors cannot replicate economically. Universal's franchise IP (the Fast & Furious universe alone has grossed $7B+ globally; Despicable Me/Minions has crossed $5B) has durable theme-park translation value that creates physical moats through irreplicable park experiences.
The U.S. residential broadband market is near-saturation: penetration exceeds 85% of occupied households. Growth comes from ARPU increases and network quality upgrades (multi-gig, symmetrical), not from gross adds. The structural headwind is that for the first time in 20 years, residential broadband customers have credible alternatives — fiber (AT&T, Frontier) in major metro areas and FWA (T-Mobile, Verizon) in suburban and lower-density markets. The cable/fiber competitive battle will play out over 5–10 years, not quarters.
The streaming industry is undergoing a profitability rationalization after years of subscriber-at-any-cost spending. Peacock — with Netflix at $20/month, Disney+ bundled with ESPN and Hulu, and Warner Bros. Discovery's Max — operates in a hyper-competitive, crowded field. Peacock's differentiation is live sports (NFL, Olympics, NBA starting 2025–26) and NBC's broadcast brand. The streaming TAM is growing, but Peacock is a distant top-4 player by subscription scale.
The global theme park market is one of the strongest secular growth stories in experiential leisure. Disney and Universal have demonstrated that pricing power in premium experiences is exceptional — per-capita spending at parks has risen 30–40% post-COVID. Epic Universe's opening doubles Universal Orlando's capacity and targets international tourism. Universal has announced a UK park (target 2031) and expanded in Las Vegas. The competitive dynamic here is genuinely good: Disney's capital allocation to parks has slowed, while Universal is accelerating.
The top competitors and their threat level:
Broadband is among the most recession-resistant consumer services — households cut streaming services and restaurants before they cut internet. During 2020, Comcast's broadband segment actually accelerated, adding 2.2M customers in 12 months as remote work and school created demand. The Studios and theme parks businesses are more cyclical, with Parks suffering a near-complete revenue collapse in 2020 that recovered fully by 2022. The current downcycle in broadband is competitive, not cyclical — a meaningfully worse structural situation than a recession-driven slowdown.
CMCSA has fallen approximately 28% from its 52-week high of $34.66 and is down nearly 40% from its all-time high of ~$61. The decline is driven by three compounding forces: (1) structural broadband subscriber erosion — the market is pricing in a secular decline in Comcast's most profitable and predictable revenue stream; (2) EBITDA margin compression from the pricing reset — management voluntarily sacrificed broadband ARPU to stabilize subscribers, compressing near-term earnings; and (3) multiple compression across the entire cable/telecom sector as rising rates in 2022–2024 compressed high-debt infrastructure businesses. All three forces have hit simultaneously.
| Segment | EBITDA (~Annual) | Multiple | Implied Value |
|---|---|---|---|
| Connectivity & Platforms (C&P) | ~$32B | 6× (peer: Charter ~7×) | ~$192B |
| Business Services | ~$5.5B | 8× (high-growth SaaS-adjacent) | ~$44B |
| Theme Parks | ~$3.5B | 14× (peer: Disney Parks ~15×) | ~$49B |
| Studios | ~$1.5B | 10× | ~$15B |
| Peacock / NBC / Bravo (Media) | ~$2.5B | 5× (declining broadcast) | ~$12B |
| Sky | ~$3B | 5× (structurally challenged) | ~$15B |
| Enterprise Value (SOTP) | ~$327B | ||
| Less: Net Debt | ~($89B) | ||
| Equity Value | ~$238B | ||
| Implied Price/Share (~3.56B shares) | ~$67 |
At $24.93, CMCSA is trading at approximately 37% of a conservative sum-of-the-parts valuation. Even discounting C&P aggressively — applying only a 4× EBITDA multiple to reflect the secular broadband deterioration — the implied equity value is approximately $40–45/share. The market is pricing in a very severe structural decline in the broadband business, or it simply does not believe management can stabilize subscribers.
Using conservative inputs — 1% annual revenue growth (reflecting broadband erosion offset by parks/wireless/business services), stable 30% EBITDA margins, 10% discount rate, and terminal growth of 1.5% — implies intrinsic value in the range of $38–48 per share. Using a zero-growth flat-revenue scenario with margin compression to 28% implies approximately $28–32 per share. At $24.93, the stock is at or below the "everything goes wrong but they don't go bankrupt" scenario, which represents poor base-case risk/reward for a business generating $12–14B in normalized annual FCF.
The critical value-trap question: is broadband a temporarily impaired business that stabilizes, or is it in terminal structural decline? The weight of evidence suggests the former — fiber cannot economically overbuild 100% of Comcast's footprint in the next 10 years, and FWA has real capacity and performance limitations in dense urban areas (where Comcast is strongest). The Q1 2026 broadband loss of only 65,000 — the first YoY improvement since 2020 — is a materially important data point that the market has not yet priced positively. This is not the same as a newspaper company whose core product is obsolete; broadband demand is growing, not shrinking.
Comcast pays a quarterly dividend of $0.33/share ($1.32 annualized), representing a 5.3% yield at current prices. The board increased the dividend for 2025 (14th consecutive year of increases). FCF payout ratio is approximately 10–12% of normalized FCF — extraordinarily well-covered and with substantial room for continued growth. Dividend sustainability is not in question given FCF generation.
Comcast spent $7.2B on buybacks in FY2025 and authorized a new $15B program in January 2025. The share count has declined from approximately 4.5B shares in 2020 to approximately 3.56B today — a meaningful ~21% reduction in the float. Buybacks at sub-$30 levels (well below any reasonable intrinsic value estimate) represent highly value-accretive capital allocation. This is one of the most compelling aspects of the investment thesis: management is aggressively buying back cheap stock.
The NBCUniversal acquisition (2011) has been value-creative by most measures — the studios, parks, and now Peacock infrastructure justify the price paid. The Sky acquisition ($39B, 2018) has not delivered on its synergy promises and carries ongoing currency and structural risk, particularly as European linear TV deteriorates. The company resisted AT&T's WarnerMedia bid and has not made empire-building acquisitions since Sky. The Versant spinoff demonstrates rational, non-sentimental asset management — a bullish signal.
Organic reinvestment is focused on six priorities: DOCSIS 4.0 network upgrades, Xfinity Mobile growth, Comcast Business expansion, Peacock content and rights (NBA, NFL, Olympics), Epic Universe + new park pipeline, and Sky streaming transition. The capex cycle is at or near peak — Epic Universe is open, and network upgrades are ~75% complete — suggesting meaningful FCF inflection in 2026–2027 as growth capex converts to operating cash flow.
Management has publicly committed to driving growth through: (1) Domestic Broadband, (2) Domestic Wireless (Xfinity Mobile), (3) Business Services, (4) Peacock, (5) Theme Parks/Epic Universe, and (6) Sky Connectivity. This framing is honest — it acknowledges that the residential broadband gross-add engine is impaired and pivots to differentiated growth vectors.
The most consequential near-term decision: Comcast is holding broadband prices flat in 2026, offering free wireless lines with broadband subscriptions, and shifting to a convergence-first strategy. This compresses near-term EBITDA but, if successful, stabilizes the subscriber base and positions mobile as a value-add retention mechanism and future revenue stream. The Q1 2026 broadband loss of only 65,000 — dramatically better than Q1 2025's 177,000 — is the first real evidence this strategy is working. Management credibility on this execution is "early positive."
Epic Universe opened May 22, 2025 with five immersive worlds and drove 19% theme park revenue growth in Q3 2025 and strong Q2 2025 contributions. EBITDA for theme parks grew 13.1% in Q3 2025. Management guided for over $700M in incremental EBITDA from Epic Universe by 2026. A UK theme park has been announced (target 2031) and Universal Horror Unleashed opened in Las Vegas in August 2025 — expanding the parks footprint beyond the Orlando/Hollywood duopoly.
Peacock EBITDA losses have narrowed dramatically: from an $825M loss in Q4 2023 to a $372M loss in Q4 2024, and by Q2 2025 losses were only $101M — an improvement of $247M year-over-year. Media EBITDA grew 28% in Q3 2025 and 9.3% in Q2 2025. Peacock is on track for EBITDA breakeven in 2026, a major potential catalyst that most analysts have not fully priced. The addition of NBA and WNBA rights (starting 2025–26 season) meaningfully strengthens the sports rights portfolio.
Management's guidance on broadband subscriber trends has repeatedly been optimistic relative to outcomes — they have consistently guided for improvement earlier than it materialized. This is a credibility deficit. However, the Q1 2026 result (65K losses vs. expectations of 100K+) is the first time they have beaten on this metric in over two years, and it matters. The pricing reset strategy is transparent and internally consistent — management is not hiding the pain, which is itself a credible signal.
AI is a moderate long-term threat to Comcast in two specific areas. First, AI-powered content generation could reduce Studios' historical advantage in expensive, IP-driven film production — if competitor studios can produce quality content cheaper, the competitive landscape for Universal's slate intensifies. Second, AI-driven optimization of fixed wireless networks could accelerate FWA performance improvements, reducing the speed/latency gap between FWA and cable. Neither is a near-term existential threat, but both are real and worth monitoring over a 5-year horizon.
Comcast has deployed AI in several concrete ways: (1) network telemetry and predictive maintenance — by late 2025, AI-driven tools are diagnosing network issues before customer-facing failures, reducing mean-time-to-repair; (2) AI-powered customer service via Xfinity's virtual assistant, reducing call center costs; (3) content recommendation algorithms on the X1 platform to improve streaming engagement; (4) advertising optimization through AI-based audience segmentation on Peacock and NBCUniversal's ad platform. These are efficiency gains, not transformative revenue drivers, but they do support margin stability.
Comcast's most underappreciated AI adjacency is its broadband infrastructure as the backbone for AI inference and data transmission. As AI models proliferate at the edge (local AI on devices, enterprise AI applications), broadband demand intensity per user is expected to grow significantly. Comcast's network carries more data than any other U.S. ISP — bandwidth demand growth structurally favors network owners. Additionally, Comcast Business's enterprise segment is directly selling higher-bandwidth products to AI-intensive enterprises, contributing to mid-single-digit revenue growth.
Comcast's deployment of DOCSIS 4.0 — delivering multi-gigabit symmetrical speeds over existing coaxial infrastructure — is critical because it provides fiber-comparable performance without full-scale excavation costs. This dramatically changes the competitive economics: Comcast can offer 2–4 Gbps symmetrical to customers at a fraction of the capex required by an all-fiber buildout. If DOCSIS 4.0 deploys on schedule (75%+ of footprint by end-2025, remainder in 2026), the speed/symmetry argument for fiber customers weakens materially. This is the most important technology development for the long-term broadband thesis.
Comcast's data assets are significant: real-time network usage data from 31M+ broadband customers, viewing behavior from Xfinity X1, Peacock streaming data, theme park behavioral data, and NBCUniversal's advertising data platform. These are valuable in the AI economy for advertising targeting and network optimization. However, regulatory privacy constraints (data usage rules, FCC broadband privacy regulations) limit monetization relative to tech-native data companies.
Brian L. Roberts owns approximately 23.5M shares (~$585M at current prices) but controls 33% of voting power through Class B super-voting shares. Combined insider ownership of equity is approximately 1–2% of economic interest. The 12-month trend is net selling by insiders (Roberts: 469K shares in Nov 2024; Cavanagh: 191K shares net; Armstrong: 27K shares net) — a mild bearish signal, though the Cavanagh sale needs to be weighed against his $60M 2025 stock award. No insider has made an open-market purchase recently, which is a meaningful omission at these price levels.
Institutional ownership is approximately 85–90% of the float, with major holders including Vanguard (~8%), BlackRock (~6%), State Street (~4%), and a range of long-only value and income-focused funds. Morningstar estimates a fair value of approximately $41 per share; various DCF models from sell-side analysts cluster in the $31–$45 range. The discount at which CMCSA trades to these estimates suggests sentiment is deeply pessimistic — the question is whether sentiment is appropriately reflecting structural deterioration or has overshot.
Short interest is estimated at approximately 2–3% of the float — not elevated. This suggests the current price reflects fundamental pessimism from long holders capitulating rather than a short-thesis crowding. A contrarian reading is that there is limited short-covering upside if the stock were to recover, but conversely, there is no crowded-short squeeze thesis available as a catalyst.
Analyst consensus has shifted to Hold (TipRanks: 4 Buy, 11 Hold, 2 Sell), with the average 12-month price target of approximately $31.93 (TipRanks) to $34.94 (MarketBeat), implying 28–40% upside from current levels. Goldman Sachs downgraded to Neutral in October 2025 with a $30 target (below current levels) specifically on the broadband pricing reset concern. The price target dispersion ($27–$45) reflects genuine uncertainty about whether the broadband business stabilizes.
$15–18/share — Assumes broadband subscriber losses accelerate to 400K–600K annually, EBITDA margin compresses to 27–28%, FCF normalizes to $8–10B annually at compressed multiples (EV/EBITDA of 4–5×). At these levels, the stock would trade at ~3.5–4× normalized FCF and offer a ~7–8% dividend yield — likely a hard floor given FCF coverage.
Timeline: 18–30 months. Broadband losses stabilize at <50K/quarter. Peacock reaches profitability. Epic Universe contributes $700M+ EBITDA. Buybacks at $6B/year reduce share count to ~3.1B. EV/EBITDA re-rates to 6.5–7×. Implied annualized return from $25: 80–120%.
Timeline: 18–24 months. Broadband losses continue but slow. FCF inflects higher as capex declines. Peacock losses narrow but don't fully disappear in 2026. Parks grow 10–15%. Stock re-rates from 4.6× to 5.5–6× EV/EBITDA. Implied annualized return: 35–55%. Plus 5.3% dividend.
Scenario: Broadband accelerates to 300–500K losses/quarter. EBITDA compresses to $32–34B. Debt refinancing costs rise. Multiple remains depressed at 4–4.5×. Dividend potentially at risk if FCF falls below $8–10B. Return from $25: –30 to –40%.
At $24.93, the risk/reward is approximately 2.5–3:1 upside-to-downside in the base case (~$35 target vs ~$18 bear floor), or approximately 4–5:1 in the bull scenario. The dividend yield of 5.3% pays you to wait. The primary condition for this asymmetry to hold is that broadband subscriber losses do not structurally worsen from Q1 2026 levels — and the evidence, as of this writing, is that they are improving. This is above the threshold for an attractively asymmetric risk/reward, but it requires accepting a real and non-trivial structural risk.
Comcast at $24.93 trades at approximately 37% of a conservative sum-of-the-parts valuation, at roughly 4.6× EV/EBITDA against a historical average of 8×, and at a normalized FCF yield of 13–15% — one of the cheapest large-cap infrastructure businesses in the U.S. public markets. The company generates more annual FCF than its entire market capitalization would imply is justified, pays a sustainable and growing 5.3% dividend, and is aggressively buying back stock at deeply discounted prices. The Versant spinoff sharpens the remaining business, Epic Universe is a proven new growth engine, Peacock is approaching profitability, and Business Services is a 57%-margin business growing mid-single digits that is severely under-valued by the market's obsession with residential broadband.
The central risk — accelerating broadband subscriber losses — is real and non-trivial, but the Q1 2026 result (65K losses, the first YoY improvement in six years) is genuine evidence that the pricing pivot is working. Fiber overbuild cannot economically cover 100% of Comcast's footprint in a reasonable investment horizon, and DOCSIS 4.0 narrows the product gap with fiber. The debt load ($99B) is the most serious structural concern and limits upside re-rating until EBITDA demonstrates stabilization.
Trigger for conviction buy: Two consecutive quarters of broadband losses below 75,000, combined with Media EBITDA (Peacock) reaching positive territory. At that point, the multiple re-rating catalyst is quantifiable and the margin of safety is compelling. Current entry at sub-$25 provides adequate margin of safety given the FCF support and dividend floor, but full position sizing is best deployed once broadband stabilization is confirmed rather than anticipated.
This is not investment advice. The above is a research synthesis for your independent decision-making only.