Comcast (CMCSA) · Deep analysis

Report previewCMCSA.USReport date Last updated
← Back to reports
Report contents

At a glance

  • Business model: Comcast earns most of its cash from monthly Xfinity subscriptions, above all broadband, where the sunk last-mile network makes one more connected household cheap to serve. That advantage is regional, and US broadband penetration is saturated, so growth now depends on price, share, and mobile bundling rather than new households.
  • Financial quality: Operating cash flow stayed between $26.4bn and $33.6bn in each of the last five fiscal years, with free cash flow of $12.7bn-$21.9bn, even as reported net income swung widely. Balance-sheet capacity is real but leveraged: roughly $92bn of net debt, negative tangible book, and financing outflows exceeding free cash flow in both the March and June 2026 quarters.
01

Business model

I. Business model: a mature pipe business wrapped in content and parks

Comcast is a connectivity business with content and destination assets attached, not a media company that happens to own wires. The core product is a monthly subscription to the connection and the services riding on it: high-speed internet, video, voice, mobile, and business connectivity, sold mainly to US households and small and mid-sized businesses under the Xfinity brand. NBCUniversal's studios, networks, and Peacock sit alongside that. Theme parks sell admissions and in-park spending. Sky runs a comparable connectivity-plus-content operation in Europe.

Who pays matters more than what is produced. Residential subscribers pay the largest share, monthly, on rolling terms, which means revenue recurs in practice but is cancellable every month in principle. Business customers sign longer contracts and churn less. Advertisers pay for reach, which is cyclical, and park guests pay per visit. That is a conglomerate anchored by a utility-like subscription core, though not a regulated one: no rate case guarantees a return, and regulators set rules rather than prices. Connectivity and parks are mature and capital-hungry. Streaming is unproven.

What the filing I have does and does not support

The FY2025 10-K, filed February 3, 2026, is the current annual report. The excerpt supplied here is its cover page: 3,588,401,619 Class A and 9,444,375 Class B shares outstanding as of January 15, 2026, and $130.745 billion of Class A stock held by non-affiliates as of June 30, 2025. Treat that dollar figure as a floor, not a market cap. It excludes affiliated and Class B holdings and is seven months old by the filing date.

Segment revenue, subscriber counts, ARPU (average revenue per user, the monthly revenue one subscriber generates), and geographic mix sit in MD&A and the segment notes, which were not provided. So the structure below is reasoning about how the business works, not a quote of mix numbers. Anyone making a mix argument should pull those tables first.

Broadband is the profit pool; everything else is adjacency or diversification

Broadband is where the cash comes from. Once the plant is built, an incremental connected household costs little to serve, so each added subscriber drops through at a high margin. The deeper cause is that the last-mile network is a sunk, locally hard-to-replicate asset. A rival must dig, string, or buy spectrum before it can take a single customer. That is the real competitive advantage, and it is regional rather than national: Comcast's protection is strongest where no fiber overbuilder has arrived.

The same asset base explains the fragility. US broadband penetration is broadly saturated, so growth has to come from share, price, or new services rather than new households, and fixed wireless and fiber now give a dissatisfied household options that barely existed at scale a decade ago.

That produces a genuine price-versus-demand tradeoff. Raising the monthly rate is close to pure margin and lifts revenue per subscriber. It also widens the gap between a cable bill and a mobile carrier's bundled offer, and a household that leaves takes its video, mobile, and eventually business-services attach with it. Comcast's answer is convergence: sell mobile lines alongside broadband so the bundle gets harder to price-check. Convergence only works if the forgone margin on a discounted mobile line costs less than the broadband subscriber it retains. If retention holds only while mobile is discounted deeply, the profit pool is shallower than the subscriber count suggests.

Content and parks run on different economics, and both consume capital

Studios and networks are hit-driven. There is no recurring contract, and content costs are committed before anyone knows whether a title works. Peacock is the unproven piece of the mix: some of its subscribers arrive through the bundle rather than by choice, and a discounted user is a fragile user. Theme parks invert the profile, with high-margin admissions and in-park spending, but heavy, lumpy capital spending and real cyclicality, so a soft consumer year lands against fixed costs. That is the tradeoff between expansion and execution: parks can absorb years of investment before the attendance shows up.

Two classes of stock, one decision-maker

The dual-class structure means voting control need not track economic ownership. That can protect a decade-long network or park build from quarterly pressure. It also means a shareholder unhappy with the content-connectivity mix has little practical leverage. Conglomerates often trade below the sum of their parts, and whether this one deserves that discount depends on whether management's capital allocation beats what a breakup would release. The operating question is narrower: does the ownership structure make the connectivity business slower to respond to fiber and fixed-wireless competition than it otherwise would be?

What to watch

If broadband net losses narrow while pricing rises, the pricing-power argument holds and the bundle is doing its job. If subscriber counts hold only when promotional discounts deepen, Comcast is defending the franchise with giveaways, and the cash engine is weaker than the customer base implies. The 10-K's segment tables, which I was not given, are the first place to test either claim.

Questions worth following

  • Can Comcast grow broadband revenue per subscriber faster than it loses subscribers to fiber and fixed wireless?

    This is the core economics of the profit pool. If price increases outrun subscriber losses, the cash engine stays intact and funds content and parks. If not, Comcast is trading margin for scale in a saturated market and the whole capital allocation plan weakens.

  • Can Peacock convert bundled and discounted viewers into paying subscribers who stay?

    Peacock is the largest unproven asset inside Comcast. If its base is mostly free or heavily discounted, streaming keeps draining cash that broadband generates and the segment is a permanent drag rather than a growth option. A durable paying base would change that math.

  • Will Comcast's theme parks hold attendance and per-capita spending if US consumer demand softens?

    Parks are the most cyclical and most capital-intensive piece of the mix. If attendance and guest spending prove durable, they justify continued investment and offset softer media results. If they prove cyclical, a large chunk of Comcast's capital is committed to an asset that amplifies consumer downturns.

02

Financial quality

Financial quality: solid cash, a profit line doing less work than it looks

Comcast is a mature, capital-heavy connectivity and media business. Over the last five fiscal years the cash engine has been wide and consistent: operating cash flow of $26.4bn–$33.6bn every year and free cash flow of $12.7bn–$21.9bn. Reported profit is a different story, and FY2025 grew for reasons that had little to do with the cable plant.

Five fiscal years, one distorted year of profit
Metric (USD bn unless noted)FY2021FY2022FY2023FY2024FY2025Q2 2026 (3 mo)Assessment
Revenue116.38121.43121.57123.73123.7129.94Steady, low single-digit growth
Operating income20.8222.6223.3123.3020.675.16Peaked FY2023, stepped down in FY2025
Net income14.165.3715.3916.1920.003.53Volatile, driven by non-operating items
Net margin (%)12.24.412.713.116.211.8Adequate; FY2025 figure is inflated
Gross margin (%), gross profit ÷ revenue67.068.569.870.171.872.0Adequate and slowly improving
ROE (%)15.26.118.819.221.84.0Strong on paper, leverage-driven
ROA (%)5.22.05.96.17.31.4Adequate; mid-single-digit asset returns
Liabilities / assets (%)64.568.168.567.664.365.2Heavy, though normal for cable
Current ratio0.850.780.600.68Not disclosedNot disclosedThin; structural for advance billing
Operating cash flow29.1526.4128.5027.6733.648.09Strong and consistent
Free cash flow17.0912.6512.9615.3821.895.19Strong; positive all five years
Capex, implied (% of revenue)12.06 (10.4)13.76 (11.3)15.54 (12.8)12.29 (9.9)11.75 (9.5)2.90 (9.7)Moderate and falling
Operating cash flow ÷ net income2.064.921.851.711.682.29Exceptional, but distorted

*Computed from supplied net income and average equity or period-end assets; the ratio table gives ROE and ROA only for FY2021–FY2024 annuals and the two 2025 quarters. The Q2 2026 column covers three months and is not comparable with the annual columns. Capex is implied by operating cash flow minus free cash flow. The supplied ratio table shows gross margin of 55–58% for FY2021–FY2024, a different cost-of-revenue definition from the income-statement basis used here.

Two movements matter. Operating income fell to $20.67bn in FY2025 from $23.30bn in FY2024 while revenue was flat, yet net income rose to $20.0bn from $16.19bn. The bridge is the $9.5bn gap between EBIT and operating income — most of it landing in the June 2025 quarter, which produced $11.12bn of net income on $19.91bn of EBITDA. FY2022 is the mirror image: net income collapsed to $5.37bn while operating cash flow barely moved, and goodwill fell $11.7bn that year, a pattern consistent with a large non-cash writedown.

Quarterly: cash steady, profit lumpy
Quarter (3 mo, USD bn)RevenueEBITDANet incomeTotal assetsTotal liabilitiesNet PP&EOperating cash flowFree cash flowAssessment | Trend
Q1 202529.899.393.38267.77180.47Not disclosed8.295.42Strong | Stable
Q2 202530.3119.9111.12273.85176.39Not disclosed7.825.14Adequate | Profit inflated by a one-time gain
Q3 202531.209.623.33273.00175.58Not disclosed8.695.62Strong | Stable
Q4 202532.317.472.17272.63175.25Not disclosed8.845.09Adequate | Weaker EBITDA
Q1 202631.467.692.17260.00171.67Not disclosed6.894.54Adequate | Smaller balance sheet
Q2 202629.948.853.53257.52167.79Not disclosed8.095.19Adequate | Stable

Net PP&E is not disclosed for any quarter. Quarterly free cash flow for 2025 sums to roughly $21.3bn against the $21.89bn annual figure, so the two definitions are not identical; check the basis before annualizing a quarter.

The 20% return on equity is mostly a capital-structure result

ROE improved from 15.2% in FY2021 to about 21.8% in FY2025, but ROA was only 7.3% and liabilities are roughly two-thirds of assets. Book equity has stayed near $90–97bn since FY2021 despite five years of profits, because dividends and buybacks return nearly all of it. Leverage plus a small equity base, not exceptional asset productivity, produces the return. Interest expense of ~$4.4bn a year against EBITDA of $46.4bn is easily covered. The buyback pause that began in the third quarter of 2026 and runs through the separation will mechanically lift book equity and lower reported ROE without any change in the business.

A below-one current ratio that is not a liquidity problem

The current ratio runs 0.60–0.91, and the quick ratio is identical because Comcast carries no inventory. Current liabilities include deferred revenue — customer cash collected before service — plus $5.4–6.9bn of short-term debt against $7.7–9.7bn of cash. That is normal for a subscription business billing in advance. The real question is the maturity schedule, and the supplied data stops tagging long-term debt after June 2025, leaving a $90bn-plus debt stack only partly visible. The separation of NBCUniversal and Sky, with the Versant spin done in January 2026 and Sky Germany sold in May 2026, is what decides how that debt is divided.

Capex down, free cash flow up — and we cannot yet tell why

Capex fell from 12.8% of revenue in FY2023 to about 9.5% in FY2025 as free cash flow rose to $21.89bn. Net PP&E is not disclosed after FY2022, so there is no way to test whether the network is simply being run harder or whether maintenance was deferred. A free-cash-flow peak that coincides with a capex low deserves scrutiny before it is treated as a run rate.

Red flags
  • What produced the FY2025 non-operating gain, and does it recur? Roughly $9.5bn sits between EBIT and operating income. If similar items accompany the separation, headline EPS and ROE stay flattered; if not, FY2026 earnings could fall sharply even with flat operations.
  • Goodwill and impairment risk. Goodwill was $61.5bn at the end of FY2025 and $53.07bn by June 2026, an $8.1bn step-down in the March 2026 quarter alongside a $12.6bn fall in total assets — consistent with the Versant separation. Against equity of $89.8bn, there is limited room for another writedown.
  • Debt disclosure and refinancing. Long-term debt is untagged in the FY2025 and FY2026 periods supplied, and quarterly free cash flow does not tie exactly to the annual total. Neither proves a problem; both argue for reading the debt footnote in the next filing rather than inferring solvency from the current ratio.

What would change this judgment: operating income returning to growth with a narrow EBIT-to-operating-income spread, and free cash flow holding above $15bn with capex near 10% of revenue.

Questions worth following

  • Does Comcast's roughly 8x P/E rest on a durable earnings base?

    The trailing multiple sits on about $11.2bn of four-quarter net income, not the $20.0bn reported for fiscal 2025, which included a large non-operating gain. Whether that lower base is stable decides whether the multiple is cheap or simply reflects a shrinking profit pool.

  • Can Comcast sustain its free cash flow if capital spending rises to compete in broadband and wireless?

    Free cash flow ranged from $12.7bn to $21.9bn a year over FY2021–FY2025, and the FY2025 peak coincided with capital spending falling to about 9.5% of revenue. If the improvement came from deferred network spending, cash flow would fall as investment resumes; if from efficiency, it is durable.

  • Will the remaining Comcast business carry manageable leverage once NBCUniversal and Sky are separated?

    Liabilities are about 65% of assets, and the supplied data stops showing long-term debt after June 2025. How the separation allocates debt, cash, and the dividend determines whether the remaining connectivity business can fund its network and refinance maturities out of its own cash flow.

Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.

Go beyond the preview

Continue reading the full report

  1. Key risksWhat could break the investment case
  2. Valuation and margin of safetyWhat the current price assumes, and how much room for error it leaves
  3. SummaryOur overall judgement and what to watch next

Free to sign up · 20 research points · unlocking this report uses 10

20 points cover 2 full reports. No card needed.

Research reflects its generation date. Preview the first two modules.

More on Comcast Corp