Management Bought Subscriber Stability With Margin, and Would Not Say for How Long

Results vs. Consensus

Q3 2025 Scorecard

Consensus revenue is the $30.68B figure most widely cited; a second provider carried $30.70B, a $20M spread that changes nothing. Adjusted EBITDA and free cash flow do not carry a widely syndicated consolidated consensus for this name, so they are shown against the prior year rather than against an estimate. The broadband consensus is presented as a band because the two available providers disagreed by 2,000 subscribers.

Year-Over-Year Comparisons

The 2.7% revenue decline is an artifact, not a trend. The prior-year quarter contained $1.9 billion of incremental Paris Olympics revenue inside the Media segment. Strip that comparison out and consolidated revenue grew roughly 3%. The line worth staring at is adjusted EBITDA, which was essentially flat despite that revenue growth, because the profit pool inside Connectivity is being spent down on purpose.

Quarter-Over-Quarter Comparisons

Q2 2025 comparatives are taken as originally reported, not from a later filing's restated prior-year column. The sequential table is the cleanest statement of the trade Comcast is making: every subscriber metric improved, and every margin metric deteriorated. Residential Connectivity margin gave up 260 basis points in a single quarter. That is the transaction, priced out in full.

Revenue assessment. Consolidated Connectivity revenue grew 4.2% to $11.5 billion, which is the number management wants investors to anchor on, and it is a fair one: it captures domestic broadband, domestic wireless, international connectivity and Business Services while excluding the melting video and advertising lines. The problem is that domestic broadband revenue, the single largest and highest-margin component, grew 0.5% to $6,433 million. Everything else in the connectivity bundle is growing around a core that has stopped. Convergence revenue (domestic broadband plus domestic wireless) grew 2.5%, which is respectable but is being purchased with free wireless lines that carry cost today and revenue only if they convert.

Margin assessment. The 140 basis point year-over-year contraction in Residential Connectivity margin is the most important number in the release. Programming expense actually fell 3.7% on video subscriber decline, so this was not a cost shock. Non-programming expense rose 3.4% on higher device costs, marketing tied to the April 2025 offers, and customer-experience investment, while revenue fell 1.5%. A margin structure that has historically absorbed video decline through programming savings no longer does, because management has chosen to spend the savings and more. The relevant question for the next several quarters is not whether margin falls but where it stops, and management declined to say.

EPS assessment. Adjusted EPS held flat at $1.12 against a consensus of $1.10, a two-cent beat that a 4.9% reduction in share count more than accounts for. Below the operating line, interest expense rose to $1,128 million from $1,037 million and the effective tax rate rose to 27.3% from 25.8%. GAAP EPS of $0.90 declined 3.4%, with the gap to adjusted EPS driven mainly by $633 million after tax of acquisition-related intangible amortization plus $102 million after tax of Versant separation costs. Those separation costs are real cash and will recur until the spin completes, which is a reason to weight the GAAP number more heavily than usual this year.

Segment Performance

Connectivity & Platforms

Residential Connectivity & Platforms revenue detail

Residential Connectivity is roughly 56% of consolidated revenue and about two thirds of consolidated adjusted EBITDA. It is the segment that determines what this equity is worth, and it shrank on both lines. The internal composition is the story: connectivity revenue grew 3.6% while the legacy lines (video, advertising, other) fell a combined 6.2%, and the growth engine within connectivity is now wireless rather than broadband. Domestic broadband revenue growth of 0.5% is the residue of higher average rates almost exactly offsetting a smaller subscriber base, and management has now signalled that the rate half of that equation goes away in 2026.

"This is a deliberate investment phase, one that will take time and carry a cost as reflected in the 3.7% decline in Connectivity & Platforms EBITDA this quarter. And we expect this decline to build slightly over the next several quarters as we continue to invest in pricing, product and customer experience."
— Michael Cavanagh, President (Co-CEO effective 2026)

Assessment: Comcast is running a controlled demolition of its own pricing architecture to defend subscriber count, and it has told investors the rubble gets deeper before it clears. That is an honest disclosure and a better strategy than harvesting a declining base, but it removes the near-term earnings floor. The reported segment EBITDA decline was 3.5%; the 3.7% management cites is the constant-currency figure, which is the more conservative read and the right one to hold them to.

Business Services Connectivity

Business Services delivered 6.2% revenue growth to $2,576 million and 4.5% EBITDA growth to $1,454 million, with margin down 100 basis points to 56.4%. Management described a $10 billion-and-growing business with a total addressable market above $60 billion, positioned as the leader in small business and the challenger in mid-market and enterprise. Enterprise Solutions is doing the growing; small and medium business is where fixed wireless competition is biting hardest.

"In the SMB segment, we're seeing elevated competition, particularly from fixed wireless. Despite this, we still delivered modest revenue growth by driving ARPU higher through increased adoption of our advanced services like cybersecurity, cloud solutions and Comcast Business Mobile."
— Jason Armstrong, Chief Financial Officer

Assessment: This is the most consistently underrated asset in the company, compounding mid-single digits at 56% margins with a long runway in enterprise. The 100 basis point margin give-back is partly acquisition mix rather than pure competitive pressure. A new T-Mobile relationship for business mobile, disclosed on the call, gives Comcast a second MVNO path and reduces single-supplier dependence on Verizon. Small, but strategically sensible.

Subscriber Metrics

Assessment: Read this table in two directions and it says two different things. Sequentially, everything improved sharply: broadband losses more than halved, wireless set a record, video posted its smallest loss in nearly five years. Year over year, broadband losses widened (104,000 versus 87,000) and total customer relationships deteriorated badly (a 210,000 loss against 29,000 a year ago). The sequential improvement is partly the normal back-to-school seasonal benefit, which management acknowledged. The pivot is helping at the margin. It has not yet turned the direction of the franchise.

Content & Experiences

Media

Media revenue fell 19.9% purely on the Olympics comparison; excluding the $1.9 billion of incremental Paris revenue in the prior-year quarter, it grew 4.2%. Domestic advertising rose 2.6% ex-Olympics, described as the best result year to date and driven by the 20th and highest-grossing season of Sunday Night Football. Domestic distribution grew 1.5% ex-Olympics, with Peacock up 25%, partially offset by declines at the linear networks. The segment's 28.0% EBITDA growth to $832 million is almost entirely a Peacock story.

Peacock contributed $1.4 billion of revenue and an adjusted EBITDA loss of $217 million, versus $1.5 billion and a $436 million loss a year ago, though the prior-year figure carried Olympics costs. Paid subscribers were flat at 41 million, with management crediting a strong late-quarter content slate and distribution initiatives for offsetting churn from the in-quarter $3 price increase.

Assessment: Halving the streaming loss while holding subscribers through a price increase is a genuinely good quarter for Peacock, and the loss curve is the metric that matters more than the subscriber count at this stage. But flat at 41 million is the third consecutive quarter without meaningful growth, and the NBA arrives as a cost before it arrives as a subscriber. Management was clear that straight-line amortization of the new sports rights creates upfront dilution, particularly in the first season. Media EBITDA growth of 28% this quarter is unlikely to be repeated in the next few.

Studios

Studios grew revenue 6.1% to $3,000 million on content licensing (+9.1%) and theatrical (+4.6%), the latter carrying Jurassic World Rebirth, which grossed nearly $900 million worldwide year to date and pushed the franchise past $7 billion cumulative. EBITDA fell 21.9% to $365 million on higher programming and production costs tied to content licensing sales, plus marketing spend against a larger film slate.

Assessment: Studios earnings are lumpy by construction and a down quarter on marketing timing ahead of a heavy Q4 slate, including Wicked: For Good in November, is not a signal. The 12.2% EBITDA margin is low for the segment, and the release timing means the spend and the return sit in different quarters. We would not extrapolate this quarter in either direction.

Theme Parks

Theme Parks grew revenue 18.7% to $2,717 million and EBITDA 13.1% to $958 million on the first full quarter of Epic Universe, which opened in Orlando in May 2025. Management reported higher per-capita spending and attendance across the entirety of Universal Orlando, with less cannibalization of the two pre-existing parks than expected. Operating expenses rose 22.0%, outpacing revenue, as the new park absorbed its opening cost base.

"The idea was to have Epic head us towards a week-long vacation type of experience. And what we're seeing as Epic is now in the market is that it's driving higher per cap spending and attendance across the entirety of Universal Orlando. So -- and one of the nice things is that lesser cannibalization of attendance from our 2 pre-existing parks than we had expected."
— Michael Cavanagh, President (Co-CEO effective 2026)

Assessment: This is the cleanest growth story in the portfolio and the one with the most visible near-term operating leverage, since management explicitly framed Epic as still capacity-constrained and deliberately throttled on ride throughput while they work out the operational kinks. Revenue grew 18.7% while EBITDA grew only 13.1%, so the margin is currently being paid to open a park rather than to run one. As throughput scales over the next year, that gap should reverse. Capital spending here is also falling now that construction is complete, which helps consolidated free cash flow.

Key Topics & Management Commentary

Overall Management Tone: Management was candid to the point of being unusually forthcoming about near-term pain, repeatedly volunteering that EBITDA pressure would continue and ARPU would worsen before either improved. Confidence was high on the destination and deliberately vague on the timeline, and no one on the call attached a quarter or a metric to the phrase "the other side." The posture was that of a team early in a multi-year repositioning that wants to be judged on 2027 rather than 2026, delivered alongside a leadership transition that reinforces the same message.

1. The Broadband Pivot: Trading Price for Stability

The central fact of this quarter is that Comcast has abandoned the annual rate-increase model in residential broadband and replaced it with nationwide everyday pricing, four clear speed tiers, an included gateway, unlimited data, and a one- or five-year price guarantee. Rolled out at the end of June, the new architecture is the direct cause of both the improved subscriber trend and the compressed margin. Management framed the choice as unavoidable given the competitive end-state they now expect.

"Over time, though, we believe that the vast majority of the broadband market will be comprised of 2 multi-gig symmetrical providers serving most addresses, and we aim to be a winner in this segment, with the rest of the market likely being served by capacity limited alternatives. We've been seeing this end state begin to take shape. Fiber expansion continues at a steady pace. And as we've said before, we expect most of our footprint will eventually be overbuilt."
— Michael Cavanagh, President (Co-CEO effective 2026)

Assessment: The admission that most of the footprint will eventually be overbuilt is the most consequential sentence on the call, and it is not new guidance so much as a public acceptance of a structural change. A monopoly or duopoly broadband asset with pricing power is worth a very different multiple from a duopoly asset that competes on price. Comcast is now explicitly managing toward the latter. The pivot is the correct response to that reality; it is also confirmation that the reality has arrived.

2. ARPU: Growth Decelerating by Design, With a Rate Freeze Ahead

Domestic broadband ARPU grew 2.6% in the quarter, decelerating from prior trend because of the new everyday pricing structure and the free wireless line offer. Management then guided that number down twice in the same breath: a step-down of more than a point in Q4, and continued pressure in early 2026 because no broadband rate increase is currently planned.

"The rollout of our new everyday pricing structure at the end of June, combined with the success of our free wireless line offer, caused a deceleration in our broadband ARPU growth, resulting in 2.6% growth this quarter. As we continue to transition customers to more consistent pricing and ramp up free wireless line additions, we expect ARPU growth to step down more than 1 point in the fourth quarter, and we expect continued pressure on ARPU in early 2026 as our current plan is to not take a rate increase in broadband in the early part of next year."
— Jason Armstrong, Chief Financial Officer

The head of the Connectivity business was blunter still when asked directly about the 2026 trajectory, saying it was "unlikely that we'll be able to grow ARPU in 2026, especially in the early part of the year."

Assessment: With subscribers declining and ARPU heading toward zero growth or worse, domestic broadband revenue is very likely to decline in 2026. That is the highest-margin revenue in the company. Every other line in the model has to work harder just to hold consolidated EBITDA flat, and management has already said it will not be flat. Investors should model domestic broadband revenue down in 2026 and treat any stabilization as upside rather than base case.

3. Connectivity EBITDA: A Decline Management Says Will Deepen

C&P adjusted EBITDA fell 3.5% as reported (3.7% in constant currency), the first quarter in which the pivot's costs are fully visible. Both the President and the CFO independently guided to further deterioration, using nearly identical language about continued pressure over "the next several quarters" until the transition laps.

"As we've said from the beginning, this pivot carries several costs, including rate reinvestment through pricing simplicity, which carries revenue dilution as well as investment in customer experience, which carries additional operating costs. This quarter is the first quarter where these impacts are reflected in our financial results."
— Jason Armstrong, Chief Financial Officer

Assessment: "Several quarters" without a defined endpoint is the crux of our Hold. If management had said four quarters, an investor could underwrite a trough and buy the bottom. Instead the guide is directional and open-ended, and it arrives alongside two other EBITDA headwinds management named in the same section of the call: NBA rights onboarding and the loss of Versant's earnings. The company is simultaneously absorbing a self-inflicted margin reset, a new sports cost, and a divestiture, and it reduced the buyback rather than lean into the weakness.

4. Wireless: A Record Quarter Built Substantially on Free Lines

Domestic wireless added a record 414,000 lines, taking the base to 8.94 million and penetration above 14% of domestic residential broadband customers. Wireless revenue grew 14.0% to $1,246 million. Management disclosed that nearly half of residential postpaid phone connects in the quarter came from customers taking the free line, and that the first full quarter of the new premium unlimited plan (priced at $40 per line on a two-line plan, with twice-yearly device upgrades) saw strong uptake at the high end.

"Wireless net additions hit a new record at 414,000, and nearly half of our residential postpaid phone connects came from customers taking a free line, which is a great way to bring new customers into the ecosystem."
— Jason Armstrong, Chief Financial Officer

Assessment: A record with an asterisk. Comcast is buying wireless lines with a twelve-month giveaway, absorbing the network cost today against a monetization event in the second half of 2026. The strategic logic is sound, since the company owns the customer relationship and the incremental cost of an MVNO line is modest, but the accounting reality is that nearly half of residential postpaid phone connects were free lines. Penetration of 14% against a captive broadband base does leave genuine runway, and this is the one place where the pivot is producing a metric that compounds rather than merely stabilizes.

5. The Free-Line Conversion: The Single Load-Bearing Assumption

Management's entire case for why current margin compression is an investment rather than a permanent reset rests on converting free lines to paying relationships beginning in the second half of 2026. They stated the intention plainly and described early usage data as encouraging.

"Looking ahead, in the second half of next year, many of the free lines will come up for monetization. Our intention is to convert the majority to paying relationships, which should provide a significant tailwind to convergence revenue growth at that point."
— Jason Armstrong, Chief Financial Officer

Asked how customer quality is being protected so the conversion actually lands, the Connectivity chief pointed to a history of promotional management and a deliberate focus on high-value segments, without offering a target conversion rate.

Assessment: No conversion-rate target was given, no churn assumption was shared, and no disclosure framework was promised. This is the number that determines whether 2026 is a trough or a plateau, and investors will not be able to verify it until it is already in the results. Management noted that customers are activating and using the lines, which is a real leading indicator, but promotional conversion in wireless has historically run well below "the majority" across the industry. We treat this as the primary item to monitor, and it is our clearest upgrade trigger.

6. Video: The Best Subscriber Result in Nearly Five Years

Domestic video lost 257,000 customers, an improvement of more than 100,000 year over year and, per management, the best result in nearly five years. Churn was described as at record lows, credited to segment-appropriate packaging and the entertainment operating system, including a Multiview feature for simultaneous live events.

Assessment: Genuinely encouraging and almost entirely overlooked in the market reaction. Video still generated $6,591 million of revenue in the quarter, more than domestic broadband, and it declined 5.0%. Slowing that decline meaningfully changes the trajectory of both revenue and programming expense, which fell 3.7% precisely because the video base is shrinking. There is a trade-off here that management did not dwell on: retaining video subscribers slows the programming-cost tailwind that has been quietly offsetting broadband weakness.

7. Free Cash Flow and the Buyback Step-Down

Free cash flow of $4,945 million rose 45.2%, bringing the nine-month total to $14,866 million against $9,283 million a year ago. The CFO attributed the increase to a cash-tax tailwind from new legislation and favorable working-capital timing, partially offset by higher capital expenditures of $3,071 million. Net leverage ended the quarter at 2.3x. Capital returns totalled $2.8 billion, comprising $1.2 billion of dividends and $1.5 billion of buybacks, down from $1.7 billion of buybacks in Q2.

"We do have some near-term headwinds, namely the EBITDA impact from our broadband repositioning, the onboarding of NBA rights and the spinoff of VERSANT and its associated EBITDA and free cash flow. For those reasons, we reduced our quarterly buyback pacing a touch to $1.5 billion in the quarter."
— Jason Armstrong, Chief Financial Officer

Assessment: The candour is admirable and the signal is negative. A company trading at a 52-week low with a balance sheet at 2.3x leverage and $14.9 billion of nine-month free cash flow chose to buy back less stock, not more. Management framed this as prudence ahead of three known headwinds, which is defensible. It is also the clearest statement of internal conviction available to an outside investor, and it does not read as confidence that the shares are mispriced.

8. Capital Intensity Is Rising Inside Connectivity

Consolidated capital expenditures rose 5.4% to $3,071 million, but the mix moved sharply. Connectivity & Platforms capex rose 19.5% to $2.3 billion on scalable infrastructure, home passings and faster deployment of the new XB10 gateway, which is now bundled into every broadband package. Content & Experiences capex fell 19.9% to $714 million now that Epic Universe construction is complete.

Assessment: Including the gateway in every package is a product decision with a capital consequence: Comcast is now capitalizing equipment it previously rented out at a margin. That converts a modest revenue stream into a capital outlay, which is another quiet drag on the connectivity economics that does not appear in the EBITDA line. The offset from Epic completion is real but temporary, and the underlying direction of connectivity capital intensity is up while connectivity revenue is flat.

9. Media Strategy: Sports as the Load-Bearing Asset

Management devoted substantial airtime to an exceptional run of sports inventory: the NBA returning to NBC and Peacock (with the opening doubleheader drawing the largest audience since 2010), the NFL and college football seasons, and February bringing the Super Bowl, the Winter Olympics and NBA All-Star Weekend, followed by the World Cup on Telemundo in June. A Sunday Night Basketball franchise launches in February, modelled on Sunday Night Football, which has been the top prime-time show for 14 straight years and averages roughly 25 million viewers.

Assessment: The sports portfolio is a durable competitive asset and the concentration of events in the first half of 2026 gives Media an unusually favourable setup. The cost side deserves equal weight. Management confirmed straight-line amortization of the NBA rights, meaning expense recognition is smooth while the audience and advertising benefit is seasonal, creating dilution that is heaviest in the first season. Media EBITDA grew 28% this quarter against an easy Olympics-inflated cost comparison; the next several quarters carry a new nine-figure cost line that was not there before.

10. Leadership Transition and the Versant Spin

Comcast announced that Steve Croney will become CEO of Connectivity & Platforms at the start of 2026, with Dave Watson moving to Vice Chairman after eight years leading the business. Separately, Michael Cavanagh was named co-CEO alongside Brian Roberts. The Versant separation, which carves out the cable networks, remains on track and carries transaction and separation costs of $116 million in the quarter that are excluded from adjusted EBITDA.

Assessment: An internally promoted operator taking Connectivity in the middle of a pricing reset provides continuity at exactly the moment execution matters most, and the Chairman explicitly credited Croney with driving the pricing, packaging and AI changes already in flight. The co-CEO appointment formalizes a succession that the market had largely assumed. Neither change alters the investment case, but together with the Versant spin they mean the Comcast reporting in 2026 will not be comparable to 2025 without adjustment, which raises the burden on investors to track the underlying business rather than the headline.

11. M&A: A High Bar, With a Door Left Open

Pressed on speculation around Warner Bros. Discovery, management set a deliberately high threshold while declining to rule anything out, and specified the shape of what would be considered.

"I think we've said repeatedly, and I'll say it again, that the bar is very high for us to pursue any M&A transactions given how strongly we feel about the businesses we have, the strategies we're pursuing and the opportunities we have ahead of us."
— Michael Cavanagh, President (Co-CEO effective 2026)

He added that any complementary media assets Comcast might consider would be "streaming assets and studio assets," and suggested that post-Versant, "more things are viable than maybe some of the public commentary that's out there."

Assessment: This is not a denial. Naming the asset categories and volunteering that more is feasible than the market assumes is the language of a company that wants optionality preserved. For an equity already carrying a capital-allocation question, the possibility of a large media acquisition while the core connectivity business is mid-reset is a risk the market will price, not an opportunity. We would view a major media deal announced in the next twelve months as a negative catalyst absent an unusually compelling structure.

Guidance & Outlook

Comcast does not issue formal quantitative guidance. What management provided this quarter was a set of explicit directional commitments, which are more useful than usual precisely because they are unambiguous and unflattering.

Implied trajectory: Seven of the ten forward commitments point down. Domestic broadband revenue grew 0.5% this quarter on 2.6% ARPU growth against a shrinking base; take ARPU to roughly 1.5% in Q4 and to zero or negative in the first half of 2026, hold subscriber losses at the improved Q3 run rate, and domestic broadband revenue turns negative next year. The wireless line and Business Services can offset that at the connectivity revenue level, but not at the EBITDA level, because wireless carries device cost and Business Services is a fraction of the size.

Guidance style: Comcast has historically been conservative in framing and has not over-promised on the pivot. Management flagged the ARPU deceleration a quarter before it arrived and has now flagged the EBITDA deterioration in advance as well. We take the commitments at face value, which means we take the negative ones at face value too.

Analyst Q&A Highlights

How ARPU Evolves as Customers Migrate to the New Plans

The opening question of the call went straight to the trade-off at the heart of the quarter: how much revenue is being surrendered to buy retention, and whether convergence revenue growth should be expected to accelerate from here. Management answered the first half directly and the second half only in framework terms, committing to an aggressive migration of the existing base rather than a gradual one.

Q: "So first on broadband. Just curious if you could share some more context around the evolution of ARPU and what you're seeing in terms of customers migrating to the new plans and the opportunities to build better retention, but the cost of that coming through on the revenue side. And then on convergence, you referenced the 2.5% growth in the quarter. As you continue to market the new slate of offers and promotions, is there an anticipation that this rate of growth should improve over time?"
— Michael Rollins, Citibank

A: "So given the investments we're making, as we've said, it's unlikely that we'll be able to grow ARPU in 2026, especially in the early part of the year. But part of it is, to your question, Mike, we're going to be very active, and we are active migrating customers to the new pricing and packaging with lower EDPs, an all-in approach and, of course, all eligible for that free mobile line."
— David Watson, President and CEO, Connectivity & Platforms

Assessment: The most important disclosure of the call, and it came in the first answer. A full-year 2026 ARPU decline had not previously been stated so plainly. Management declined to quantify migration volumes, saying only that they "will not be hesitant" to move customers onto the new packages, which means the revenue dilution is self-selected in pace as well as size. The convergence half of the question was answered with strategy rather than a number, and the phrase used was "let's get to the other side as quickly as possible," which is a direction rather than a date.

Where the Incremental Operating Expense Is Going

A follow-on line of questioning probed whether the C&P EBITDA pressure is predominantly a revenue-dilution story or a cost story, and specifically what is being spent on customer premise equipment, sales, marketing and customer service. The answer confirmed that both are in play, and added that cost savings elsewhere are being recycled into the pivot rather than dropped to the bottom line.

Q: "I wanted to just ask a little bit more about the trajectory of C&P EBITDA next year. We obviously talked a lot about the ARPU side. Just on the OpEx side, could you talk a little bit about the investments that you're doing there, whether that be in CPE, sales, marketing or customer service to support the reset here?"
— Michael Ng, Goldman Sachs

A: "I think the team, to their credit, has been fairly aggressive in cost rationalization. You've seen that more recently with some realignment around divisional structures and cutting out layers in the company. So that continues the pace. A lot of that, though, obviously being reinvested in the pivot, which we think is the appropriate thing to do."
— Jason Armstrong, Chief Financial Officer

Assessment: This closes off the most obvious bull argument, which is that self-help cost reduction could cushion the margin reset. It cannot, because the savings are already spoken for. Management has taken out management layers and centralized marketing, data science and customer experience, and the proceeds are funding the giveaway. An investor hoping for a cost-driven EBITDA floor in 2026 did not get one here.

Protecting Customer Quality Ahead of the Free-Line Conversion

The question the entire 2026 thesis turns on was asked directly: what guardrails ensure the free lines are worth converting, so that the promised convergence-revenue uplift actually materializes rather than churning off at the first bill. Management responded with experience and intent, and without a number.

Q: "I wanted to get a sense a little bit on how you're thinking about the conversion of free wireless lines to pay next year. We've sort of seen this in others in the industry. But how are you guys just ensuring that the customer quality is there and making sure you're putting in kind of the right guardrails so that when you get to that point a year from now, we see that nice uptick in convergence revenue growth."
— Benjamin Swinburne, Morgan Stanley

A: "I can't think of too many other things that are as important as making sure quality connects coming in, and there's going to be a real focus around making the transition to paid status for those free lines as they come off next year. So have a lot of experience with promotional activity, very accustomed to proactive and reactive approaches, and we've watched it closely in the marketplace those that have done this."
— David Watson, President and CEO, Connectivity & Platforms

Assessment: Not a dodge, but not an answer either. The response describes process and intent where the question asked for evidence and guardrails. "We've watched it closely in the marketplace those that have done this" is an acknowledgment that peers have run the same play, and the industry record on converting free promotional lines is mixed at best. Until a conversion rate is disclosed, the single largest driver of 2026 convergence revenue is unmodellable from the outside, and we assume a materially lower conversion than "the majority" in our base case.

Whether Large-Scale Media M&A Is on the Table

Persistent market speculation about a large media transaction was raised alongside a question about the MVNO relationship following a leadership change at the wholesale partner. Management set a high bar on M&A without closing the door, and specified the asset categories that would qualify.

Q: "I have to ask about all the speculation about Warner Bros. Discovery. So Brian, if you could just share any thoughts about how you think about those assets, the complementarity of those assets and whether or not, in this political environment, M&A of that kind is even possible."
— Craig Moffett, MoffettNathanson

A: "So in this case, it would be streaming assets and studio assets since there's no other parks assets out there, and that makes us such a unique company ourselves. So I think in light of that, what we'd be looking for and what we're going to look like post the VERSANT spin, I think more things are viable than maybe some of the public commentary that's out there."
— Michael Cavanagh, President (Co-CEO effective 2026)

Assessment: The market read this as leaving the door open, and that is the right read. Volunteering that "more things are viable" than public commentary suggests is an active signal, not a deflection. On the MVNO question, management expressed confidence in the Verizon relationship and separately disclosed a new T-Mobile arrangement for Business Services, which materially reduces wholesale single-supplier risk. That second disclosure was the more valuable one and drew less attention.

Whether Peacock Needs Global Scale to Compete

A line of questioning on streaming strategy asked whether the level of sports and content investment requires international scale, and if so whether that scale can be built organically or requires acquisition. Management argued that the sports advantage is inherently domestic and that the content partnerships already secured make scale-through-M&A unnecessary.

Q: "Given the meaningful sports investment plus the overall content investment at Peacock, what are your plans to scale up, if any, globally? If you are, do you need to participate in M&A to actually do that? And if not, I know there was an earlier question, but what is your view on NBCU's competitive position if you don't participate since this seems like a moment in time?"
— Jessica Reif Cohen, Bank of America

A: "So I don't think M&A is necessary and you think about the nature of sports is fundamentally market by market as opposed to global. We do have the Olympics, but those rights are in the U.S. So I think that's where I'll leave it on that front."
— Michael Cavanagh, President (Co-CEO effective 2026)

Assessment: The argument is coherent and probably right on its own terms, since sports rights genuinely are national and Peacock's differentiation is the domestic sports slate. It also sidesteps the harder question, which is whether a 41-million-subscriber domestic-only service can carry the content costs of a global-scale competitor indefinitely. Peacock has now been flat at 41 million for three consecutive quarters while absorbing a price increase and preparing to absorb NBA costs. Management's answer implicitly accepts that Peacock will be a domestic profit-optimization story rather than a growth-at-scale story, which is a reasonable strategy and a lower-ceiling one.

Whether Convergence Is Measurably Reducing Broadband Churn

The final question sought the empirical payoff of the convergence thesis: is attaching wireless actually lowering broadband churn, and is the free-line offer helping new connects today. Management confirmed the churn benefit qualitatively while explicitly reframing the free-line offer as a long-duration bet rather than a near-term acquisition tool.

Q: "Is there evidence that wireless and convergence in general is lowering churn in the broadband base? And if we focus more just on the free line promotion, is that helping the overall trend as it relates to new connects?"
— John Hodulik, UBS

A: "In terms of -- when we add wireless to the relationship, it's positive, the impact to churn. So it is -- has been, continues to be. From the impact of the free line plan, part of the deliberate strategy that we have is really a longer-term bet. On the Connect side, there's some a little bit of an uptick of help, but it's really a longer-term bet around churn."
— David Watson, President and CEO, Connectivity & Platforms

Assessment: The most useful answer of the Q&A, and quietly the most sobering. Management conceded that the free line is producing only "a little bit of an uptick" in new connects, meaning the record wireless number is not translating into broadband gross additions. The value is asserted to be retention over time. That is a defensible bet, but it means the pivot's return arrives later and is harder to observe than the cost, which arrives now and is fully visible. It also implies that the improved Q3 broadband number owes more to seasonality and reduced churn on the existing base than to competitive share gains.

What They're NOT Saying

1. No end date on the EBITDA decline. "Several quarters" was used repeatedly and never converted into a number. Management knows when the pricing transition laps because they control the migration pace; declining to share it is a choice, and it prevents investors from underwriting a trough.
2. No free-line conversion target. The entire 2026 convergence-revenue uplift depends on this rate. Management said "the majority" without defining majority, without a churn assumption, and without committing to disclose the outcome. Peers who have run this play have reported conversion outcomes; Comcast has not committed to.
3. No quantification of the NBA cost. Straight-line amortization was confirmed and the dilution described as heaviest in the first season, but no dollar figure, no per-quarter shape, and no offsetting advertising estimate were given. This is a known, contracted, calculable expense.
4. No 2026 framework of any kind. With three named headwinds converging (pricing reset, NBA, Versant), management offered no consolidated EBITDA, free cash flow or capital-expenditure framing for next year. The capital allocation priorities were restated as unchanged, which is not the same as a plan.
5. Broadband ARPU is discussed; broadband revenue is not. Management guided ARPU down and subscribers as improving, but never stated the obvious arithmetic conclusion that domestic broadband revenue is likely to decline in 2026. Framing the discussion around the broader "connectivity revenue" aggregate, which grew 4.2%, keeps attention off the line that actually carries the margin.
6. The gateway capitalization trade-off went unaddressed. Bundling the XB10 into every package was presented purely as a product win. Nobody asked, and management did not volunteer, what the associated equipment revenue give-up and capital intensity increase amount to.
7. No comment on why the buyback fell rather than rose at a 52-week low. The reduction was explained by forward headwinds. It was not reconciled against a balance sheet at 2.3x leverage generating $14.9 billion of nine-month free cash flow, or against the implicit view that the shares are cheap.

Market Reaction

• Pre-print setup: The stock closed at $28.53 on October 29, the bottom of its 52-week closing range of $28.53 to $45.14. It entered the print down 24.0% year to date against an S&P 500 up 17.2%, down 32.5% over trailing twelve months and down 9.2% over the trailing thirty days. Expectations were low and positioning was defensive.
• Reaction session (October 30, before-open reporter): Opened at $26.09, an 8.6% gap down, traded a range of $25.75 to $28.10, and closed at $27.32, down 4.2% or $1.21. The intraday low marked a 9.7% decline before roughly half the move was recovered.
• Volume: 64.0 million shares against a 25.5 million thirty-day average, 2.5x normal.
• Benchmark: The S&P 500 fell 1.0% on the session, so roughly three quarters of the move was company-specific.

Why the stock fell on a triple beat. The market was not trading the print. It was trading the forward commitments, which arrived in the CFO's prepared remarks before the Q&A even began: ARPU stepping down more than a point in Q4, no rate increase in early 2026, and C&P EBITDA decline building over several quarters. A beat on trailing metrics paired with a guide-down on the forward margin trajectory resolves to a lower stock almost mechanically, and the reduced buyback removed the one offset management could have provided.

The shape of the session is worth noting. Opening down 8.6% and closing down 4.2% means buyers stepped in materially through the day. At $25.75, the intraday low, the stock traded at a level that implied severe terminal-value impairment in broadband. The recovery of roughly half the gap suggests the market found the subscriber improvement and the free-cash-flow strength worth something, even while marking down the earnings trajectory. On 2.5x volume, this was a genuine repositioning of the shareholder base rather than a thin drift.

Context for the level. Closing at $27.32 on the day it printed a broadband beat and a record wireless quarter tells you the equity is being valued as a melting ice cube rather than a cash-generative duopoly asset in transition. That is either the market's most important insight about this company or its largest error, and this quarter did not settle the question.

Street Perspective

Debate: Is the Pricing Reset Strategic Discipline or Value Destruction?

Bull view: The bull case on the Street holds that Comcast is doing what incumbents almost never do, which is accept a lower price point voluntarily and early rather than defending ARPU into an accelerating subscriber decline. The Q3 subscriber improvement is early proof the reset works, and a stabilized base at a competitive price is worth far more than a shrinking base at a premium price.

Bear view: The bear camp contends this is the first step down a price ladder with no bottom, because the competitive dynamic driving it (fiber overbuild plus fixed wireless) intensifies rather than abates. Management has already conceded most of the footprint will eventually be overbuilt. Cutting price to hold subscribers in a market with structurally more capacity coming is the definition of a value trap, and each subsequent reset will be met by a competitive response.

Our take: The bulls are right about the decision and the bears are right about the destination. Given the competitive end-state management itself describes, resetting price is clearly the correct move, and doing it decisively is better than doing it grudgingly over eight quarters. But the honest implication is that broadband's terminal margin is lower than the market previously assumed, and no amount of execution changes that. The right response is to lower the multiple, which the market has already done, and then ask whether it has lowered it too far. That question is about valuation, not strategy.

Debate: Does Convergence Actually Create Value, or Just Move It?

Bull view: A growing consensus view is that wireless is the genuine second act. Penetration of just 14% against a captive base of 28.9 million broadband households implies years of runway; the incremental cost of an MVNO line is low; the churn benefit of a bundled relationship is real and management confirmed it; and the free lines convert to a significant revenue tailwind in the second half of 2026.

Bear view: Some sell-side desks argue convergence is an accounting migration rather than value creation. Comcast is discounting broadband to give away wireless, so a customer paying less for internet and nothing for a phone line is not obviously worth more. Wireless carries device cost and wholesale network cost, so its margin structure is far below broadband's. Trading high-margin broadband ARPU for low-margin wireless revenue destroys EBITDA even if it grows the top line.

Our take: The bear framing is more accurate today and the bull framing may be more accurate in 2027. The critical evidence is management's own admission that the free line produces only "a little bit of an uptick" in new connects and that the real payoff is retention over time. That means the value is currently being created in churn reduction, which is genuine but slow, while the cost is immediate and fully in the P&L. This debate cannot be resolved before the second half of 2026, and that timing gap is precisely why we are not paying for the outcome yet.

Debate: Is the Sum of the Parts Worth More Than the Whole?

Bull view: The bull case argues that at roughly 4.9x EV to trailing adjusted EBITDA, the market is assigning almost no value to assets that would command premium multiples standing alone. Theme Parks grew revenue 18.7% with a brand-new flagship park still ramping; Studios owns franchises worth billions; Business Services compounds mid-single digits at 56% margins. The Versant separation demonstrates management is willing to use structural solutions, and more could follow.

Bear view: The bear camp responds that conglomerate discounts exist for reasons, that Residential Connectivity still supplies roughly two thirds of consolidated EBITDA and is shrinking, and that a spin of the good assets would leave a more levered, faster-declining cable stub behind. Separation also carries real cost, visible in the $116 million of Versant transaction expense this quarter alone.

Our take: The sum-of-parts gap is real, and Theme Parks in particular is materially undervalued inside this structure. But sum-of-parts arguments only pay when a catalyst forces the market to price the parts, and Versant is a carve-out of the weakest asset rather than a crystallization of the strongest. The management commentary on M&A points toward adding assets rather than separating them. We would not underwrite a re-rating on structural action, and we note that a large media acquisition would widen the discount rather than close it.

Model Update & Valuation Framework

Where the valuation sits

Enterprise value, implied EBITDA and the annualized figures above are derived, not reported. The EBITDA figure is backed into from the 2.3x net leverage management disclosed on the call and is consistent with the nine-month adjusted EBITDA of $29,484 million annualizing toward a similar full-year figure. The annualized P/E multiplies a nine-month run rate and is not a substitute for a forward consensus estimate.

Valuation judgment. Roughly 4.9x EV/EBITDA with a 4.8% dividend yield and nine-month free cash flow near 15% of market capitalization is, on any historical screen, cheap. The market has already priced substantial impairment. The problem is that cheapness is not a catalyst, and the specific mechanism by which cheap becomes expensive requires either EBITDA to inflect or the multiple to re-rate. Management has guided EBITDA down for several quarters, and the multiple will not re-rate while the guide is pointing down. The stock can stay at 5x for a long time, and the reduced buyback means the company will not shrink the share count fast enough to force the issue.

What would change the math. Free cash flow is the strongest part of this story and deserves emphasis: even discounting the cash-tax and working-capital benefits the CFO flagged, this business converts a very high share of EBITDA to cash, and the capital intensity peak in connectivity should pass as gateway deployment completes. If broadband subscribers stabilize near flat while wireless continues to compound and Epic reaches full throughput, the 2027 free cash flow profile against today's market capitalization is compelling. None of that is observable in 2026.

Thesis Scorecard: Establishing Coverage

This is our initiation on Comcast, so the scorecard below establishes the pillars we will carry forward and score each quarter, rather than grading a prior thesis.

Overall: The thesis is genuinely balanced, which is why this is a Hold rather than a conviction call in either direction. Both bear pillars are already materializing and the third is emerging, while two of three bull pillars are on track but back-end loaded into 2027. The valuation is the bull case's strongest argument and its weakest, because it is simultaneously the most compelling fact about the stock and the one that has been true, and unrewarded, all year.

Action: Hold. We would upgrade on any one of three signposts: domestic broadband net adds turning roughly flat or positive for two consecutive quarters; a disclosed free-line conversion rate above 50% with retained ARPU; or C&P EBITDA decline visibly troughing with management dating the recovery. We would downgrade if broadband losses re-accelerate past the Q2 2025 pace, if the dividend's coverage comes into question, or if a large media acquisition is announced without a compelling structure.

Bottom Line

Comcast beat the quarter and lost the tape, and it deserved to. The subscriber print was the best in several quarters across broadband, wireless and video simultaneously, free cash flow was excellent, and the market response was still a decline to a fresh 52-week low, because management used the call to guide the most important line in the model lower with no stated end date. That is not the market being irrational. That is the market correctly weighting a forward guide over a trailing beat.

What makes this a Hold rather than an Underperform is that the strategy is right and the assets are real. Accepting a lower price point early, before subscriber decline accelerates, is what a well-run incumbent should do when the competitive structure changes, and Comcast is doing it decisively rather than defending ARPU into a rout. Theme Parks is a genuine growth business with a new flagship still ramping. Business Services compounds at 56% margins. Peacock's losses are shrinking fast. At roughly 5x EBITDA with a 4.8% yield, the downside is defended by cash flow.

What makes it a Hold rather than an Outperform is that every one of those positives is either small relative to Residential Connectivity or lands in 2027, while every negative lands in 2026. The single load-bearing assumption in the bull case, converting free wireless lines to paying relationships, was asked about directly and answered without a number, a target, or a commitment to disclose the outcome. Management reduced the buyback at a 52-week low. When a company this cheap declines to buy more of itself, the most useful thing an outside investor can do is take the hint and wait for the evidence rather than pay for the promise.