Rating: Buy — Conviction: High
Comcast trades at ~4.4x model-basis EBITDA and ~6.4x forward earnings. That price discounts a permanent broadband decline the filings don't yet show: Q3 2026 revenue was flat pro forma, wireless lines are compounding, and Peacock has turned profitable. Broadband is losing subscribers faster (144k net losses vs 91k a year ago) and residential EBITDA fell 5.8%. But the price sits closer to our Bear case than to Base, and even the exit-multiple method alone, which assumes no re-rating, clears today's price comfortably. The thesis is that cash generation of this size doesn't need broadband to grow, only to stop shrinking quickly.
Conviction tests (3a-v-c): T1 pass (Base +39.1% at +1pp, +95.2% at −1pp, Buy at both) · T2 pass (Gordon alone +103.5%; exit-multiple alone +23.3%, still Buy) · T3 pass (every FY0 fact and historical series from EDGAR XBRL, the 2025 10-K and the Q3 2026 10-Q/release; no est. inputs) · T4 pass (the NBCUniversal/Sky spin is pro rata and value-neutral to a holder before dis-synergies; no pending ruling or contract moves value >25%) The rule gives High. I'd read it nearer Medium, because the four tests don't challenge the assumption the call rests on most: that broadband's decline stays gradual.
Business overview
Comcast sells broadband, wireless, video and business connectivity in the US (Connectivity & Platforms, with $19.7bn of Q3 revenue, 67% of the total), plus Sky in the UK and Italy. It also owns NBCUniversal: Media (NBC, Telemundo, Peacock), Studios and Theme Parks ($10.5bn). Residential connectivity alone is $17.0bn of revenue and $6.1bn of the $9.1bn quarterly Adjusted EBITDA. The cable networks were spun off as Versant on 2026-01-02, and Sky Germany was sold on 2026-05-31. In June 2026 Comcast announced a spin-off of NBCUniversal and Sky, expected in mid-2027, with Comcast retaining up to 19.9%. Three variables drive earnings: broadband customers and ARPU (broadband revenue −4.7% in Q3), the residential margin (35.9%, −90bp), and how fast wireless (10.6mm lines, service revenue +14.7%) and Peacock (48mm paid subscribers, $261mm EBITDA) offset them.
Competition
The pressure comes from fiber overbuilders and fixed wireless. AT&T reported a record 769k internet connections (fiber plus FWA) in Q3 2026. T-Mobile added broadband customers in the upper-400k range in Q2 2026 and targets 15mm FWA customers by 2030 (Broadband Breakfast, October 2026). Verizon is the third FWA player. Charter is the closest listed comparable and is priced for distress: a $18.3bn market cap, a 2.5x forward P/E (distorted by its collapsed equity value) and 5.2x EV/EBITDA. AT&T and Verizon trade at 7.0x and 7.4x. Competitive damage shows first in price, through promotional ARPU (Comcast's broadband ARPU fell 3.8% in Q2), then in net losses.
Bull case
- Convergence stabilizes broadband — bundled mobile (free lines now converting to paid; ~40% of postpaid phone connects on premium unlimited) cuts churn, so net losses narrow and ARPU flattens as promotions lap. Plays out if Q4–Q2 net losses fall back below 100k a quarter. Model: rev_growth, ebitda_margin, terminal_growth
- Content earns its keep — Peacock is profitable two quarters running, Studios EBITDA rose 18.6%, and Media grew 8.3% excluding the World Cup. Plays out if Peacock profit holds without the event calendar. Model: rev_growth, ebitda_margin
- The multiple recovers from distress levels — separation clarifies two cleaner equities, and a 6x exit approaches the telco multiple. Plays out if spin terms leave RemainCo leverage near today's ~2.2x. Model: exit_ev_ebitda
Bear case
- Structural share loss to fiber and FWA — broadband becomes a melting asset, like video, with losses accelerating as overbuild widens. Plays out if quarterly net losses stay above 150k into 2027. Model: rev_growth, ebitda_margin, terminal_growth
- Price-led defense compresses margin and lifts capex — promotions and network upgrades cut the residential margin ~1pp a year. Plays out if residential EBITDA keeps falling 5%+ while capex rises above 12% of revenue. Model: ebitda_margin, capex_pct_rev
- A concentrated, levered RemainCo gets a higher risk premium — Moody's put the ratings on review for downgrade, citing concentration in broadband after the spin. Plays out if a downgrade lands and spin financing loads debt onto RemainCo. Model: wacc, exit_ev_ebitda
Valuation & balance sheet
| Metric (definition) | Current | Own history (range or 5y avg) | Peers | Source, as-of |
|---|---|---|---|---|
| EV/EBITDA (model basis: OI + D&A, leases excluded, FY2025) | 4.37x (~4.7x on est. ex-Versant base) | n/a (unverified) | lease-inclusive TTM: CHTR 5.24x, T 6.99x, VZ 7.37x (CMCSA 4.79x) | 10-Q, XBRL, stockanalysis 2026-10-22 |
| P/E forward (consensus) | 6.41x | n/a (unverified) | CHTR 2.53x (distorted), T 8.97x, VZ 8.92x | stockanalysis 2026-10-22 |
| FCF yield (TTM, aggregator) | 24.35% | n/a (unverified) | CHTR 24.19%, T 11.70%, VZ 11.63% | stockanalysis 2026-10-22 |
| Net leverage (net debt / FY2025 model EBITDA) | 2.16x | n/a (unverified) | n/a | 10-Q 2026-09-30 |
Model-implied value range (from model-summary.json; generic module, Gordon-growth DCF and exit-EV/EBITDA DCF, midpoints): Bear $12.85 · Base $37.35 · Bull $66.35 per share, i.e. implied returns of −43.8% / +63.4% / +190.3% vs $22.86. These ranges show how the bull and bear drivers translate into value; they are not price targets.
- What's priced in. The price sits 41% of the way from Bear to Base, closer to the Bear case. The market is pricing much of a structural decline.
- Re-rating. Base's 4.5x exit is a 3% re-rating from today's 4.37x. The two methods differ by 65%: Gordon gives $46.52 and exit gives $28.18. The Gordon terminal value at 6.5% WACC and 0% growth implies ~7x Y5 EBITDA (est., my arithmetic), so that method carries a large re-rating, and the Base midpoint leans on it.
- Y1 overlap. Balances are at 30 Sep 2026 while Y1 is full FY2026, so ~$12.7bn of 9M free cash flow is partly double-counted (~$3.6/sh upward bias, est., not adjusted).
- Tail. Because residential connectivity is 58% of revenue, I ran a failure sensitivity: revenue −3% then −3.5% a year, margin to 24%, capex 13%, a 3.0x exit, 8% WACC and −3% terminal growth. It gives a Bear midpoint of $4.23 (−81.5%), which shows how much leverage magnifies a broadband break.
Balance sheet: net leverage 2.16x (gross 2.37x) FY2025 model EBITDA, with the Base path falling to 1.1x by Y5. Coverage 9.2x. Liquidity: $7.6bn cash; revolver availability n/a (unverified). Nearest material maturity: $3.56bn current debt due within 12 months. Total debt is $87.4bn carrying ($71.6bn fair value). YTD repayments were $10.4bn and buybacks are suspended through the separation. Ratings: Moody's on review for downgrade (level n/a (unverified)); S&P n/a (unverified).
Model note: built and verified (LibreOffice recalculation matched all 3,207 formula cells). Unverified inputs: none. Assumptions without basis: none. No scenario CHECKs. FY0 (FY2025) still includes Versant and Sky Germany, and Y1's −2.2% carries the deconsolidation. Debt is two carrying-value lines with a display-only 4.6% blended coupon. The company gave no 2026 guidance. Model Y1 EPS of $3.38 is GAAP-basis, so it sits below adjusted EPS (Q3 adjusted $1.07 vs reported $0.97). The NBCU/Sky spin isn't modeled separately.
| Metric | CMCSA | CHTR | T | VZ | Peer median | vs median |
|---|---|---|---|---|---|---|
| P/E (TTM) | 7.3x | 2.9x | 12.2x | 11.8x | 11.8x | −38% |
| P/E (forward) | 6.4x | 2.5x | 9.0x | 8.9x | 8.9x | −28% |
| PEG | n/a | 0.2x | 0.8x | 1.2x | 0.8x | |
| EV/EBITDA (TTM) | 4.8x | 5.2x | 7.0x | 7.4x | 7.0x | −31% |
| EV/Sales (TTM) | 1.3x | 2.1x | 2.5x | 2.7x | 2.5x | −48% |
| P/B | 0.8x | 0.8x | 1.4x | 1.7x | 1.4x | −39% |
| FCF yield | 24.4% | 24.2% | 11.7% | 11.6% | 11.7% | +12.7pp |
As of 2026-10-22. One source and one definition for every company: aggregator TTM and consensus-forward multiples, EV netting reported cash and including lease liabilities in debt. Not the model's own EV basis (see the report's valuation table). Pages fetched 2026-10-22 after the 4:00pm ET close; CMCSA reported Q3 that morning. Dashed line = peer median. Source: stockanalysis.com /stocks/<ticker>/statistics/, fetched 2026-10-22.
Notes. PEG as given by the source; the source shows none for Comcast (consensus 3-year EPS growth is negative, -3.78%). Charter's P/E reflects a collapsed equity value on a heavily levered balance sheet, not unusually high earnings quality. CMCSA's 4.79x uses the source's $33.54bn TTM EBITDA; on the model's basis (FY2025 GAAP operating income + D&A, leases excluded) it is 4.37x.
Scenario stress test
Reasoned from the bull/bear drivers above. The model column comes from the scenario overlays (Base case + shock).
| Scenario | Effect | Mechanism | Magnitude | Model Δ value vs Base ($/sh) |
|---|---|---|---|---|
| S1 Fast equity crash | − | Liquidity de-rating of a 0.67-beta, high-yield large cap; fixed-rate debt, no funding channel | Low | −$2.05 |
| S2 Slow bear / recession | − | Advertising, park attendance and studio spend fall; broadband downgrades and FWA switching accelerate. Milder than Base − Bear ($24.50) because it is a two-year cycle, not permanent share loss | High | −$7.24 |
| S3 Rapid rate shock | − | Higher discount rate on a 52%-debt capital structure; ~$87bn debt reprices over time | High | −$5.78 |
| S4 Slow rate grind | − | Gradual refinancing at higher coupons and a higher discount rate | Med | −$2.95 |
| S5 Soft-landing cuts | + | Lower rates help the levered equity; a steady consumer supports parks and ads | Med | +$4.00 |
| S6 Recession-driven cuts | − | Same operating hit as S2; lower rates partly offset | Med | −$3.48 |
| S7a Credit liquidity shock | − | Wider IG spreads raise refinancing cost and complicate separation financing | Low | −$1.64 |
| S7b Slow default cycle | − | Small-business churn in Business Services; higher residential bad debt | Low | −$1.18 |
| S8 Stagflation | − | Programming, labor and park costs inflate while competition caps broadband pricing; rates rise | High | −$5.74 |
| S9a Dollar spike | − | Translation loss on Sky and international studio/parks revenue | Low | −$0.31 |
| S9b Dollar slide | + | Mirror image: translation gain | Low | +$0.31 |
| S10 Melt-up | + | A deep-value, high-yield name re-rates modestly in a risk-on tape | Med | +$3.28 |
| S11 Energy supply shock | − | Consumer squeeze hits parks and advertising; rates rise; no direct energy exposure | Low | −$2.33 |
| S12 Mega-cap/AI derating | + | Rotation from AI leadership toward low-multiple cash yielders; no fundamental channel | Low | +$1.23 |
Currently active/on watch per the playbook: S3 partially active (price-level leg only); S8, S10 and S11 on watch.
What would change the call
Upgrades if: broadband net losses narrow below 100k a quarter with residential EBITDA declines under 3%, and the separation terms leave RemainCo net leverage at or below ~2.5x. Downgrades if: quarterly net losses exceed 150k for two more quarters, residential EBITDA falls more than 7% y/y, or spin financing pushes RemainCo leverage above 3x alongside a downgrade.
Watch items
- W1: Domestic broadband net losses — narrower than Q3's 144k — Q4 2026 release — late Jan 2027 — Model: rev_growth
- W2: Residential Connectivity & Platforms Adjusted EBITDA y/y — better than −5.8% — Q4 2026 release — late Jan 2027 — Model: ebitda_margin
- W3: Separation terms: debt allocation and RemainCo leverage — Form 10 / financing announcements — before the mid-2027 close — Model: none
- W4: Moody's review outcome and S&P action — rating press releases — open — Model: wacc
Sources
- Comcast Form 10-Q, quarter ended 2026-09-30 — https://www.sec.gov/Archives/edgar/data/0001166691/000162828026066863/cmcsa-20260930.htm — accessed 2026-10-22
- Comcast Q3 2026 earnings release (8-K Ex. 99.1) — https://www.sec.gov/Archives/edgar/data/0001166691/000162828026066783/ex991-9302026.htm — accessed 2026-10-22
- Comcast Form 10-K for 2025 — https://www.sec.gov/Archives/edgar/data/1166691/000162828026004994/cmcsa-20251231.htm — accessed 2026-10-22
- EDGAR XBRL companyconcept (Revenues, RevenueFromContractWithCustomerExcludingAssessedTax, OperatingIncomeLoss, DepreciationDepletionAndAmortization, AccountsReceivableNetCurrent, AccountsPayableCurrent, DeferredRevenueCurrent), CIK 0001166691 — https://data.sec.gov/api/xbrl/companyconcept/CIK0001166691/us-gaap/OperatingIncomeLoss.json — accessed 2026-10-22
- stockanalysis.com statistics: CMCSA, CHTR, T, VZ — https://stockanalysis.com/stocks/cmcsa/statistics/ — accessed 2026-10-22
- TheWrap, Moody's places Comcast on review for downgrade — https://thewrap.com/industry-news/business/moodys-comcast-nbcuniversal-split-credit-downgrade-review/ — accessed 2026-10-22
- Simply Wall St, NBCUniversal/Sky spin-off plan — https://simplywall.st/stocks/us/telecom/nasdaq-cmcsa/comcast/news/how-comcasts-nbcuniversal-and-sky-spin-off-plan-will-impact — accessed 2026-10-22
- VectorShift, Comcast Q2 2026 call summary (broadband ARPU −3.8%) — https://vectorshift.ai/research/companies/comcast/earnings/2026-q2 — accessed 2026-10-22
- Broadband Breakfast fixed wireless chart page — https://broadbandbreakfast.com/chart-of-fixed-wireless-access-subscribers/ — accessed 2026-10-22
- Macro log (10Y 5.13%, 2026-10-21) — logs/macro-2026-10.md