A 10% EBITDA Decline, a Frozen Dividend, and the First Dated Hint of a Trough

Results vs. Consensus

Q4 2025 Scorecard

No widely syndicated point estimate for Q4 domestic broadband net adds surfaced across the outlets checked, so this recap makes no beat or miss claim on that line and compares it instead against the prior-year quarter and the prior quarter, both reported figures. The revenue shortfall of roughly $40 million is inside rounding for a company of this size and one provider characterized revenue as in line.

Year-Over-Year Comparisons

The 54.6% net income decline is inflated by the comparison: the prior-year quarter contained a $1.9 billion income tax benefit from an internal corporate reorganization. Adjusted net income down 17.1% is the fairer read, and it is still the steepest decline of the year. Operating income fell 30.1% because the NBA rights expense sits above the D&A line while the associated revenue is only beginning to build.

Quarter-Over-Quarter Comparisons

Residential Connectivity margin has now fallen 520 basis points across two quarters, from 39.8% in Q2 to 37.2% in Q3 to 34.6% here. That is the pricing reset and the customer-experience spend compounding, and it is the single cleanest measure of what the pivot is costing. The Media swing from $832 million of EBITDA to a $122 million loss is mostly NBA rights amortization arriving before NBA revenue scales, on top of the normal third-to-fourth-quarter step down (Media fell from $650M to $298M over the same stretch a year earlier, with no NBA).

Full Year 2025

The full-year figures foot to the sum of the four reported quarters: revenue of $91,397 million through nine months plus $32,310 million equals $123,707 million; adjusted EBITDA of $29,484 million plus $7,900 million equals $37,384 million; free cash flow of $14,866 million plus $4,369 million equals $19,235 million. Adjusted EPS of $3.46 through nine months plus $0.84 sums to $4.30 against a filed full-year figure of $4.31, the expected one-cent artifact of computing annual EPS on the full-year share count rather than summing quarters. The filed $4.31 is used throughout.

The GAAP-versus-adjusted divergence for the year is stark and worth stating plainly: GAAP EPS rose 30.1% while adjusted EPS fell 0.6%. The entire GAAP increase is the $9.4 billion pre-tax gain on the Hulu sale recognized in the second quarter. On an operating basis, 2025 earnings were flat to slightly down on flat revenue.

Revenue assessment. Consolidated Connectivity revenue grew 3.2% to $11.6 billion in the quarter and 4.2% to $46.0 billion for the year, but the important development is beneath it: domestic broadband revenue fell 1.1% to $6,316 million, the first decline of this cycle. In our October note we modelled domestic broadband revenue turning negative during 2026. It turned negative one quarter earlier than that. With ARPU growth at 1.1% and the subscriber base down 181,000 in the quarter, the arithmetic no longer supports growth, and management confirmed no rate increase is coming in the early part of 2026.

Margin assessment. Residential Connectivity margin of 34.6% is down 140 basis points year over year and 260 basis points sequentially. Programming expense fell 4.3% on video decline, so once again this is not a cost shock but a chosen reinvestment: non-programming expense rose 2.5% on device costs, marketing and customer-experience spend while revenue fell 2.1%. Business Services margin also gave up 150 basis points to 54.2%, its weakest showing of the year, as acquisition mix and elevated fixed-wireless competition in small business both weighed.

EPS assessment. Adjusted EPS of $0.84 versus $0.96 is a 12.4% decline delivered despite a 5.4% smaller share count, meaning underlying profitability fell roughly 17%. Interest expense rose modestly to $1,126 million. The tax line is not comparable, with a $89 million expense this quarter against a $1,111 million benefit last year. GAAP EPS of $0.60 is the number that reflects what the business actually earned in the quarter, and it fell by more than half.

Segment Performance

Connectivity & Platforms

Residential Connectivity & Platforms revenue detail

Domestic broadband revenue declining is the milestone of this quarter. It is the highest-margin line in the company, it has now turned, and management has removed the two levers that could reverse it in the near term by declining to raise rates and by continuing to give away wireless lines that dilute the broadband bill. Domestic wireless revenue growth of 18.0% is genuinely strong and, at $1,403 million, is now 22% the size of the broadband line, up from 19% a year ago.

"These dynamics were reflected in the quarter through dilution to broadband ARPU growth and elevated marketing, product, and customer service expenses, contributing to the 4.5% decline in connectivity and platforms EBITDA. As we have said before, as we continue to invest through this transition, we expect incremental EBITDA pressure over the next couple of quarters until we begin to lap these initial investments in 2026."
— Jason Armstrong, Chief Financial Officer

Assessment: The language moved from "several quarters" in October to "the next couple of quarters" here, which is a tightening rather than an extension, and the reported EBITDA decline of 4.3% (4.5% in constant currency, the figure management cites) came in roughly where the October guide implied. Management is executing its own script accurately, which is worth something. The trough now looks like Q1 or Q2 of 2026 rather than an indefinite slide, and that is the most valuable thing this call produced.

Business Services Connectivity

Revenue grew 5.8% to $2,590 million with EBITDA up 3.1% to $1,405 million and margin down 150 basis points to 54.2%. The pattern management has described for several quarters held: modest growth in small and medium business against elevated fixed-wireless competition, offset by strong momentum in enterprise solutions. A T-Mobile MVNO for business customers arrives during 2026.

Assessment: Still the most reliable compounder in the portfolio, but the margin trend deserves watching. Business Services margin has now declined in each quarter we have covered, from 56.4% in Q3 to 54.2% here, and the year-over-year decline widened from 100 to 150 basis points. Part is acquisition mix, but SMB competitive pressure from fixed wireless is real and is the same force damaging residential broadband. This segment is not immune to the structural issue; it is earlier in it.

Subscriber Metrics

Assessment: Broadband deteriorated on both comparisons, widening to 181,000 from 104,000 sequentially and from 139,000 a year ago. Some of the sequential move is the reversal of Q3's back-to-school seasonal benefit, but the year-over-year widening is not seasonal and is the honest read. Wireless net adds also slipped from the Q3 record to 364,000, and management attributed this to wireless competition stepping up late in the quarter, which is a new competitive vector on top of the existing fiber and fixed-wireless pressure. The bright spot is international turning positive at 22,000 after a 95,000 loss in Q3. Full-year wireless net adds of approximately 1.5 million to over 9.3 million lines and above 15% penetration was the company's best year ever in that product.

Content & Experiences

Media and Peacock

Media revenue grew 5.5% to $7,620 million on Peacock, higher international networks revenue (+19.0%) and domestic distribution (+5.3%), with domestic advertising up 1.5% against a tough political-advertising comparison. Segment EBITDA swung to a loss of $122 million from a $298 million profit, entirely on NBA rights expense plus elevated sports costs on the linear networks. Peacock contributed $1.6 billion of revenue and an EBITDA loss of $552 million, versus $1.3 billion and a $372 million loss a year ago.

Full-year Peacock losses improved by more than $700 million despite absorbing the NBA from the fourth quarter onward, and full-year Peacock revenue grew 10% to $5.4 billion. Subscriber growth of 3 million sequentially and 8 million year over year broke a three-quarter plateau at 41 million.

"Peacock revenue grew more than 20% to a record $1.6 billion, supported by strong distribution revenue growth of over 30% as paid subscribers increased 8,000,000 year over year and 3,000,000 sequentially, reaching 44,000,000 as of December 31."
— Jason Armstrong, Chief Financial Officer

Assessment: This is the most genuinely encouraging development of the quarter and it directly rebuts the concern we raised in October, which was that Peacock had stalled at 41 million and the NBA would arrive as a cost before it arrived as subscribers. It arrived as both, simultaneously. Adding 3 million subscribers in a quarter while raising price earlier in the year is a real result. The loss widening to $552 million is the expected shape, not a deterioration, and management quantified the amortization profile precisely enough that Q1 dilution is now modellable rather than guessable.

Studios

Revenue fell 7.4% to $3,027 million on lower content licensing (-8.3%) and theatrical (-20.2%), with EBITDA down 38.4% to $351 million. Management attributed the decline to tough comparisons against the prior-year slate, which included Wicked and The Wild Robot, against current-quarter titles Wicked: For Good and Black Phone 2, plus content-licensing timing and higher marketing spend on a larger film volume. The Wicked franchise has now grossed over $1.3 billion worldwide.

Assessment: Comparison noise rather than franchise deterioration, and the 2026 slate management outlined is unusually strong. Studios EBITDA has now declined year over year in both quarters we have covered, but the causes differ each time (marketing timing in Q3, slate comparison in Q4), which is the normal texture of a film business. We continue to treat this segment as non-extrapolable in either direction.

Theme Parks

Theme Parks grew revenue 21.9% to $2,893 million and EBITDA 23.5% to $1,035 million, crossing $1 billion of quarterly EBITDA for the first time, with growth accelerating from Q3's 18.7% and 13.1%. Management disclosed that Orlando hotel average daily rate is up 20% and occupancy up 3%, with 2,000 rooms added, and that Japan delivered its second-best EBITDA year on record.

"We are really pleased with what we are seeing from Epic, which continues to drive higher per cap spending and attendance across the entirety of the resort. While we are not yet operating at full run rate capacity, we have made meaningful progress expanding ride throughput, and we remain focused on scaling further over the next several quarters, with higher attendance, stronger per caps, and additional operating leverage over time."
— Jason Armstrong, Chief Financial Officer

Assessment: Exactly the operating leverage we flagged in October as the setup here. In Q3, revenue grew 18.7% against EBITDA growth of 13.1%, meaning margin was being paid to open a park. In Q4 that reversed, with EBITDA growth of 23.5% outpacing revenue growth of 21.9%. Epic is expected to be fully ramped by the end of 2026, with a UK park approved, a Frisco kids resort opening this year, and a first outdoor coaster at Universal Studios Hollywood. This pillar is performing better than we underwrote.

Key Topics & Management Commentary

Overall Management Tone: Notably more energetic and more specific than October, with an urgency framing running through the prepared remarks and a new Connectivity chief given room to set out a plan explicitly built on the assumption that competition never eases. Management continued to volunteer bad news early, but for the first time paired it with a timeframe, and was visibly more willing to quantify (game counts, penetration, hotel rates, network upgrade percentage). Where it stayed vague was precisely where it stayed vague last quarter: no EBITDA guide and no free-line conversion rate.

1. The First Timeline: Back-Half Improvement, Without a Guide

Asked directly whether the declines improve in the second half of 2026 or whether EBITDA growth returns, the CFO gave the most specific forward statement of the pivot so far, while explicitly declining to formalize it.

"So as you look at the back half of the year, I think one of the things that gives us confidence is, a, we start to lap some of the incremental investments we made starting in 2025. And b, we will get into monetization of what is probably the biggest vehicle we have out there, is free wireless lines moving into paying relationships. So I will stop short of giving a full EBITDA guide. I would tell you in the back half of the year, we would expect improvement."
— Jason Armstrong, Chief Financial Officer

Assessment: This is the single most valuable sentence on the call and the reason the stock rose. In October the guide was "several quarters" with no endpoint, which made the trough unmodellable. Now there is a stated expectation of improvement in the back half of 2026, resting on two mechanical drivers rather than hope: lapping the 2025 investment base, and free-line monetization. "Improvement" from a declining base is not growth, and management was careful not to say growth. But an investor can now build to a trough, which was impossible a quarter ago.

2. NBA Dilution Quantified: Q1 2026 Is the Peak

The NBA rights charge that turned Media EBITDA negative was given a precise shape for the first time, tied to game counts rather than to a dollar figure.

"While the fourth quarter represented about 25% of our total games for the season, the first quarter will be the peak volume period with roughly 50% of our games played, which will also result in peak EBITDA dilution."
— Jason Armstrong, Chief Financial Officer

Management reported early commercial traction, citing roughly 170 advertisers on the NBA with about 20% of them new, and the season "basically sold out." The NBA launch was described as the best season start for any network in 24 years.

Assessment: A Media segment that lost $122 million on 25% of the games implies Q1 2026 is going to look considerably worse, and management has now said so in advance. That is a manageable disclosure precisely because it is arithmetic rather than uncertainty. The advertiser data is the genuine positive: 20% new advertisers means the NBA is expanding the customer base rather than reallocating existing spend, which is what determines whether the rights eventually pay.

3. Domestic Broadband Revenue Turns Negative

Domestic broadband revenue fell 1.1% to $6,316 million, with ARPU growth decelerating to 1.1% from 2.6%. The CFO reiterated that further pressure is coming and named the three drivers.

"Looking ahead, we expect further ARPU pressure for the next couple of quarters, driven by the absence of a rate increase, the impact from free wireless lines, and the ongoing migration of our base to simplified pricing."
— Jason Armstrong, Chief Financial Officer

Assessment: All three drivers are self-inflicted and therefore controllable, which is the strongest argument the bulls have. It is also the case that management has chosen to accelerate rather than moderate the migration, with a stated goal of moving the majority of residential broadband customers onto new pricing by the end of 2026. The pace of the revenue decline is a management choice; the fact of it is not.

4. Broadband Subscriber Losses Widen Despite the Reset

Losses of 181,000 were worse than both the prior quarter (104,000) and the prior-year quarter (139,000). Management attributed this to competitive intensity outrunning the early traction of the new go-to-market strategy, and the new Connectivity chief confirmed fiber competition intensified in Q4 and persisted into January.

"So in reference to the competitive environment, in the fourth quarter, we did see a more competitive environment from fiber. And that remains. It is just, you know, I think we assume that is going to happen continually as we go forward as I already mentioned. You know, from a fixed wireless perspective, it stayed pretty consistent, and we are seeing stability there. I think as we are all aware, the mobile environment got significantly more competitive within the quarter."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Assessment: The most important detail is the third sentence. Fixed wireless has stabilized as a competitive force, which is new and positive, but fiber intensified and mobile became "significantly more competitive." That last point matters disproportionately because wireless is the growth engine the entire pivot depends on, and it is the first time competition has been flagged there. Comcast is now discounting broadband to give away wireless into a wireless market that is itself getting more promotional.

5. Free-Line Conversion: Language Upgraded, Number Still Absent

Management strengthened its conversion language from October's "majority" to "the vast majority," and framed the free-line strategy as economically rational rather than defensive.

"Our free line strategy is a logical and, importantly, a rational competitive approach for us. It adds value to our core broadband product, builds familiarity in a tough-to-penetrate wireless market, and will convert to a paying relationship after one year."
— Jason Armstrong, Chief Financial Officer

Roughly half of residential postpaid phone connects in the back half of 2025 were free lines, and management added a structural argument that had not been made before: 90% of Xfinity Mobile traffic is offloaded onto Comcast's own network, and acquisition costs are lower because the company sells into its existing broadband base.

Assessment: The offload statistic is the most useful new disclosure on the wireless economics and materially supports the claim that the product is "firmly profitable." It does not answer the question. Three months on from October, there is still no conversion-rate target, no churn assumption, and no commitment to disclose the outcome, and "vast majority" is a stronger assertion resting on the same absent evidence. This remains the load-bearing assumption of the 2026 thesis and the clearest thing we would upgrade on.

6. The Dividend Is Frozen, and the Streak Is Preserved by Definition

The annual dividend was held at $1.32 per share for 2026. Management characterized this alongside the Versant distribution as continuing an eighteen-year growth record.

"Consistent with what we articulated at a conference last month, we are maintaining our annual dividend at its current level of $1.32 per share. In addition, our shareholders received a dividend in kind through the distribution of Versant shares and now will be able to participate directly in Versant's capital allocation priorities, including dividends. As a result, our investors should see higher total dividends in 2026, marking our eighteenth consecutive year of dividend growth."
— Jason Armstrong, Chief Financial Officer

Assessment: Comcast's own dividend did not grow. The streak survives only by aggregating a separate company's future dividend policy, over which Comcast has no control, into a metric that has always measured Comcast's own per-share payout. The underlying decision is defensible given the 2026 free-cash-flow step-down management flagged in the same remarks, and coverage is not remotely in question at $4.9 billion of dividends against $19.2 billion of full-year free cash flow. The framing is the problem, not the policy. Taken with October's buyback reduction, this is now two consecutive quarters of capital-return retrenchment presented as continuity.

7. Free Cash Flow: A Record Year That Does Not Repeat

Full-year free cash flow of $19,235 million rose 53.4% and was the highest on record. Management was unusually direct about why 2026 will be lower, naming three specific items.

"As we look towards 2026, it is important to note that one-time cash tax benefits in 2025, including the $2 billion mentioned upfront, will not recur. In addition, recall, we said the benefits from new tax would average about a billion dollars per year for the next five years. The timing of those benefits is lumpy. We saw an outsized benefit in 2025 and expect the benefit to be significantly lower in 2026. Finally, as you can see from their filings, the Versant spin-off removes a significant pool of cash flow from our operations."
— Jason Armstrong, Chief Financial Officer

Assessment: Three named headwinds against the single most attractive statistic in the investment case. The $19.2 billion figure will be quoted widely and is not a run rate. Stripping the roughly $2 billion non-recurring cash tax benefit alone takes the underlying figure to about $17.2 billion, before the further reductions from lower tax-legislation benefits and the loss of Versant's cash flow. The candour is genuinely useful; the implication is that 2026 free cash flow could plausibly land several billion dollars lower, which compresses the free-cash-flow yield that anchors the valuation floor.

8. Capital Spending Holds Flat Into the Largest Investment Year

Total 2025 capital spending including capitalized software and intangibles fell 5% to $14.4 billion, with Content & Experiences down 17% to $3.6 billion post-Epic and Connectivity roughly flat at $10.5 billion. Management guided 2026 total capital spending to be "relatively similar" to 2025 at both segments. Roughly 60% of the footprint is now on mid-split spectrum and a virtualized architecture, with reported operating benefits of a 20% reduction in trouble calls and a 35% reduction in repair minutes where FDX is deployed.

Assessment: A useful clarification that resolves an ambiguity from the prepared remarks. Management called 2026 "the largest broadband investment year in our history," which sounds like a capital commitment but is not. The Connectivity chief confirmed on questioning that the investment "leans much heavier into our go-to-market pricing strategy," meaning it runs through the P&L as revenue dilution and operating expense rather than through capex. Capital spending is flat. The network upgrade metrics are the strongest evidence that the physical asset is keeping pace with the competitive threat.

9. The Verizon MVNO Is Modernized, and T-Mobile Joins

Management disclosed an amended long-term MVNO agreement with Verizon, described as supporting profitable growth for Comcast, Charter and Verizon simultaneously, alongside the previously announced T-Mobile arrangement for business customers.

Assessment: Given that wireless is now the entire growth engine and that management flagged intensifying wireless competition in the same call, securing improved long-term wholesale economics is materially more important than it would have been a year ago. Management declined to give terms, which is normal and expected. The combination of an amended Verizon deal, a T-Mobile business relationship, and 90% network offload is a stronger wholesale position than Comcast had entering 2025, and it directly underpins the claim that free lines can be absorbed profitably.

10. Warner Bros. Discovery: Comcast Looked, and Walked

The M&A question that management deflected in October was answered directly. Comcast evaluated a Warner Bros. Discovery transaction, was attracted to the international reach, and withdrew on price and structure.

"But in terms of Warner Brothers, I mean, what can you say? It is still underway, obviously. But I think we saw an opportunity to see if we could build value for the Comcast Corporation shareholders looking at their international reach, which would have been additive. But once it looked like all cash, we were just not interested in, at these values, stretching our balance sheet to do something like that."
— Brian Roberts, Chairman and Co-Chief Executive Officer

Assessment: This resolves the largest open risk we identified in October, and resolves it favourably. We wrote then that a large media acquisition announced while the core business was mid-reset would be a negative catalyst. Management reached the same conclusion, and the discipline shown in walking away from an all-cash structure at those values is exactly the behaviour a shareholder should want. The residual risk is smaller but not zero, since the door was left open to "creative" structures. Capital-allocation discipline was the concern; this was evidence for it.

11. Versant Completed, and the Portfolio Question Answered

The Versant separation closed on January 2, 2026, with shareholders receiving one Versant share for every 25 Comcast shares. Fourth-quarter results therefore still reflect a full quarter of ownership, and management committed to providing pro forma trending schedules excluding Versant ahead of the Q1 report. Asked whether further structural separation of the media assets might unlock value, management ruled it out.

"So I think, piling on to what Brian L. Roberts and I had just said, I think I would add to it that we do not really see that there is a strategic advantage or making NBCUniversal stronger by separating it from the cable side of the house or putting it outside of Comcast Corporation."
— Michael Cavanagh, Co-Chief Executive Officer

Assessment: An explicit rejection of the sum-of-the-parts catalyst, and investors should take it at face value. We flagged in October that we would not underwrite a re-rating on structural action, and management has now closed that door directly. Value creation is to come from operating execution alone. The commitment to pro forma trending schedules is a small but genuine improvement in disclosure quality, since 2026 reporting will otherwise be non-comparable.

12. A New Operating Leader Sets a Deliberately Unforgiving Frame

The new Connectivity chief used his first call to define success in terms that assume no relief from the competitive environment.

"When I think about what success looks like, it starts with being honest with ourselves and clearly defining our reality. The market is going to remain intensely competitive. Success is not about waiting for the environment to change; it is about how we perform inside of that environment."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

He committed to improving broadband performance year over year in the near term, returning to revenue and EBITDA growth, driving higher mobile penetration, and improving Net Promoter Scores, describing all of it as within the company's control.

Assessment: The right frame, and a marked improvement over an incumbent posture that waits for competitive intensity to normalize. It is also a set of commitments an analyst can hold him to, and we will. "Improve our broadband performance year over year in the near term" is the most checkable of them, and this quarter went the other way: 181,000 losses against 139,000 in the prior-year quarter. The clock on that promise starts now.

Guidance & Outlook

Comcast issues no formal quantitative guidance. The directional commitments made on this call are meaningfully more specific than October's.

Implied trajectory: The first half of 2026 gets worse before it gets better, and management has now said so with enough specificity to model it. Q1 carries peak NBA dilution at roughly 50% of the season's games, continued ARPU pressure with no rate increase, and the first quarter without Versant's earnings. The back half carries the lapping of 2025 investment plus free-line monetization. A reasonable shape is a first-half adjusted EBITDA decline steeper than Q4's 10.3%, improving through the second half.

Guidance style: Comcast pre-announced every negative on this call before being asked, as it did in October, and every October commitment was delivered. That record of accuracy is now two quarters long and is the reason we weight the back-half improvement statement as a genuine signal rather than reassurance.

Analyst Q&A Highlights

Whether the Shift From Local to National Pricing Is Working

The opening question sought evidence on both intake and retention under the new national pricing architecture, and whether wireless net adds could accelerate further from the free-line promotion. Management answered with a list of leading indicators rather than a subscriber outcome, and stressed that national price points still permit local flexibility.

Q: "If I could dig into the broadband side of the business for a moment. First, in terms of moving from more localized rate plan management to national, can you give us an update in terms of what you are seeing on both the intake and retention of customers? And then secondly, can you discuss more of the wireless opportunity?"
— Michael Rollins, Citi

A: "We have seen year-over-year improvement in voluntary churn, we have seen an active migration of the base to more simplified transparent pricing which has long-term benefits, we have a strong adoption of the five-year price guarantee further stabilizing the base, and we are seeing continued mix shift toward our gig plus tiers, which is a clear differentiator from fixed wireless and satellite."
— Michael Cavanagh, Co-Chief Executive Officer

Assessment: Every indicator cited is a retention metric, and none is an acquisition metric. That is consistent with the subscriber result, where losses widened to 181,000, and with the admission last quarter that the free line produces only a slight uplift in new connects. The strategy is demonstrably improving the quality and stability of the existing base while doing little to win customers from competitors. Whether that is sufficient depends entirely on whether gross additions stabilize once the migration completes, and nobody asked that question directly.

Whether the Consolidation Wave Changes Peacock's Strategic Position

A question on the sector's consolidation activity produced the most candid disclosure of the call: Comcast evaluated a Warner Bros. Discovery transaction and withdrew. Management then argued its post-Versant asset mix places it in a different competitive position from peers pursuing scale.

Q: "First, Brian Roberts, I wonder if you could just reflect a bit on the process that we have seen play out with Paramount, Netflix, and Warner Brothers Discovery and how you think that sort of shapes your thinking about Peacock with respect to scale or partnerships and what have you?"
— Craig Moffett, MoffettNathanson

A: "But once it looked like all cash, we were just not interested in, at these values, stretching our balance sheet to do something like that. So I do not know how much more we can say except that it forced us in the journey to really take a good look at what we have and what we are building."
— Brian Roberts, Chairman and Co-Chief Executive Officer

Assessment: Capital discipline demonstrated rather than asserted, and it directly retires the largest tail risk in our October thesis. The admission that the process "forced us in the journey to really take a good look at what we have and what we are building" is also a plausible explanation for the more decisive strategic posture visible across this entire call. The residual concern is the closing remark that the company remains "always open" to opportunities to build value, so the door is narrowed rather than shut.

Whether Structural Separation of the Media Assets Would Unlock Value

The sum-of-the-parts question was put directly, framed around the observation that public markets have recently established clear standalone values for media assets against the conglomerate multiple Comcast trades at. Management rejected the premise, then walked through the operating levers it believes take Peacock to sustained profitability.

Q: "Can you step back and talk about how you are thinking about your asset portfolio over the next twelve to twenty-four months or longer? And what would need to change for you to consider a different structural approach to the media assets to recognize the value and potentially strategic flexibility? Because for the first time in a really long time, there are clear values for different parts of the media business versus the current conglomerate multiple that you are unfortunately getting."
— Jessica Reif Cohen, BofA Securities

A: "So the advantages we have sitting inside the company do not get stronger by being smaller as a standalone entity is our view. What our view is, as I just said, is to create value by executing against the plans we have."
— Michael Cavanagh, Co-Chief Executive Officer

Assessment: An unambiguous no, and investors holding for a structural catalyst should reprice that expectation to roughly zero. The substantive part of the answer was the Peacock path: price increases that held subscriber growth, advertising growth, NBA advertiser demand, and affiliate renewals building through 2028 that will reprice for content already taken on. That last item is the most under-appreciated, because it means a meaningful revenue stream accrues over several years without further cost.

How the Competitive Environment Evolved and When EBITDA Turns

A two-part question on competitive conditions and the pacing of Connectivity EBITDA through 2026 drew the call's most consequential answer, and the first acknowledgement that mobile competition has intensified alongside fiber.

Q: "And then maybe for Jason S. Armstrong, you referred to the biggest year investment in the broadband business. And it sounds like your point is sort of incrementally accelerating declines in the first half with C&P EBITDA. Are you suggesting or do you guys model out that those declines will improve in the second half of this year or that we can actually get to EBITDA growth? And when do you guys expect that to happen?"
— John Hodulik, UBS

A: "So I will stop short of giving a full EBITDA guide. I would tell you in the back half of the year, we would expect improvement."
— Jason Armstrong, Chief Financial Officer

Assessment: The question asked whether declines merely improve or growth returns, and the answer chose the weaker of the two options deliberately. "Improvement" from a declining base still means decline, just less of it. Read alongside the confirmation that fiber competition intensified and mobile got "significantly more competitive," the honest reading is that 2026 is a stabilization year, not a recovery year, and the market applauded a trough rather than a turn.

Whether Epic Universe Can Sustain Its Second-Year Momentum

A question on the theme parks outlook and Epic's second-year priorities drew the most confident answer of the call, along with the specific Orlando resort metrics that support the halo thesis management has argued since the park opened.

Q: "Was hoping to dig in on the theme parks. Can you expand on the trends that you are seeing there and outlook for the business? Epic seems to be delivering on what you had hoped for in terms of driving higher per caps and attendance across Orlando. You touched on this a bit earlier, but perhaps you can discuss the operational or financial priorities for its second year."
— Kutgun Maral, Evercore ISI

A: "We have, you know, we added 2,000 rooms. Our average daily rate in the hotels in Orlando is up 20% and occupancy up 3%. So we, again, feel great. It is a continuation in the near term."
— Michael Cavanagh, Co-Chief Executive Officer

Assessment: Hotel rate up 20% with occupancy up 3% while adding 2,000 rooms is a far stronger data point than attendance alone, because it demonstrates pricing power across the whole resort rather than admissions at one new gate. That is the halo effect the Epic investment was underwritten on, and it is now measurable. With the park expected to reach full ramp by the end of 2026 and Japan posting its second-best EBITDA year, this segment is outperforming what we underwrote in October.

Whether the 2026 Broadband Investment Is Operating or Capital Spend

The final question probed an ambiguity in the prepared remarks, where 2026 was described as the largest broadband investment year in company history without specifying the form. The answer clarified that essentially all of it runs through the income statement.

Q: "First, just on the comments around the broadband investments this year, I was just wondering if you could just expand on that a little bit more. Is that more in kind of customer relationships and pricing? Is that more on the CapEx side?"
— Michael Ng, Goldman Sachs

A: "And I do think that is the case. You know, investment, it is less the capital side where our network has been steadily doing what we need to do. The investment, you know, language is about putting more value to the customer, getting them on new pricing and packaging in a variety of ways that is just seen in, you know, seen through EBITDA."
— Michael Cavanagh, Co-Chief Executive Officer

Assessment: An important clarification that cuts against the phrase's rhetorical effect. "Largest broadband investment year in our history" sounds like a network commitment; it is in fact a description of how much margin will be given away, confirmed by flat guided capital spending. Investors hearing that phrase should read it as a warning about 2026 EBITDA, not as a statement about network capability. To management's credit, they said so plainly when asked.

What They're NOT Saying

1. Still no free-line conversion rate. Three months and a language upgrade from "majority" to "vast majority" later, there is still no target, no churn assumption, and no commitment to disclose the outcome. The assertion got stronger while the evidence stayed absent.
2. No 2026 EBITDA guide, explicitly declined. The CFO used the phrase "I will stop short of giving a full EBITDA guide" while volunteering directional improvement. That is a deliberate choice to give the shape without the number, which is enough to stop the stock falling and not enough to underwrite a recovery.
3. The dividend freeze is never called a freeze. Holding $1.32 flat is presented as maintaining, and the eighteen-year growth streak is preserved only by counting a separate company's dividends. Nobody asked about it on the call.
4. No quantification of the Versant cash flow loss. Investors are directed to Versant's own filings rather than given the figure, on a spin that management concedes "removes a significant pool of cash flow." Combined with the non-recurring tax items, this leaves the 2026 free-cash-flow base genuinely unclear.
5. Wireless competition was flagged but not sized. The Connectivity chief said the mobile environment got "significantly more competitive," and wireless net adds fell from 414,000 to 364,000. No one connected those two facts on the call, and management did not address what intensifying wireless competition does to the conversion economics of free lines.
6. Broadband gross additions were never discussed. Every metric offered was retention-oriented: voluntary churn, price-guarantee adoption, tier mix. The pivot's effect on winning new customers is the untold half of the story, and the widening net loss suggests it is the weaker half.
7. No dollar figure for the NBA. The game-count shape is genuinely useful, but the annual rights cost still has not been quantified, so the 2026 Media EBITDA bridge remains an estimate.
8. Business Services margin erosion went unaddressed. Margin fell 150 basis points year over year to 54.2%, its first sequential decline of the year and the weakest level of the period, in the segment management repeatedly cites as a growth engine. It drew no question and no explanation beyond acquisition mix.

Market Reaction

• Pre-print setup: The stock closed at $28.41 on January 28, up 1.5% year to date against an S&P 500 up 1.9%, and down 24.0% over trailing twelve months. The 52-week closing range of $24.97 to $38.23 shows that the shares made a new low after the October print before recovering into this one.
• Reaction session (January 29, before-open reporter): Opened at $27.63, a 2.7% gap down, then reversed through the day to a range of $27.58 to $30.11 and closed at $29.24, up 2.9% or $0.83.
• Volume: 65.9 million shares against a 38.6 million thirty-day average, 1.7x normal.
• Benchmark: The S&P 500 was essentially flat at -0.1%, so effectively the entire move was company-specific.

The intraday reversal is the story. The stock opened down 2.7%, which is the market's reaction to the press release: revenue slightly light, adjusted EBITDA down 10.3%, broadband losses widening, Media EBITDA negative. It then closed up 2.9%, a swing of more than five points from the open, which is the market's reaction to the 8:30 a.m. call. What changed between those two prices was the call itself, and above all the back-half improvement statement made in response to a question on EBITDA pacing.

Why a worse quarter produced a better tape. Three months earlier Comcast beat on revenue, EPS and broadband and fell 4.2%, because the guide was open-ended. Here it missed revenue, printed its worst EBITDA quarter of the cycle, and rose, because the consensus bar had been cut to $0.75 and the guide acquired a timeline. The lesson is that this equity is currently trading on the trajectory of the pivot, not on the level of current earnings, and any information that bounds the duration of the pain is worth more than the earnings themselves.

Context for the level. At $29.24 the shares remain 24.0% below where they traded twelve months earlier and only 17% above the $24.97 low set between the two prints. A 2.9% gain on the best guide of the pivot leaves the stock still deeply discounted, which tells you the market is granting a trough date some credibility but not yet pricing a recovery.

Street Perspective

Debate: Is the Back-Half Improvement Statement Credible?

Bull view: The bull case on the Street holds that management has now been accurate about its own trajectory for two consecutive quarters, calling the ARPU step-down and the deepening EBITDA decline in advance and delivering both. A team that pre-announces its own bad news has earned the benefit of the doubt when it finally forecasts an inflection, and the two drivers cited are mechanical rather than aspirational.

Bear view: The bear camp contends that a company confident in a second-half turn would have guided to it rather than explicitly declining to give a number. "Improvement" from a declining base is not growth, and the same call disclosed that fiber competition intensified and mobile got significantly more competitive, meaning the environment is deteriorating faster than the self-help is arriving.

Our take: The predictive record genuinely matters and we weight it, but the bears have the stronger reading of the specific words used. Management chose "improvement" when asked directly whether growth returns, and that word choice was deliberate from a CFO who has been careful all year. The reasonable base case is that the second half of 2026 shows a smaller decline rather than a return to growth, which is a trough rather than a turn. Trough is worth a re-rating from distressed to merely cheap. It is not worth an Outperform.

Debate: Is Peacock's Subscriber Inflection Durable or Sports-Driven?

Bull view: A growing consensus view is that Peacock has broken out. Adding 3 million subscribers sequentially and 8 million year over year to 44 million, after three quarters stuck at 41 million, while raising price and improving full-year losses by more than $700 million, is evidence that the sports-led strategy works. Affiliate renewals through 2028 will reprice for content already acquired, adding revenue with no incremental cost.

Bear view: Some sell-side desks argue the inflection is purchased, not earned. The subscribers arrived in the same quarter the NBA launched and the exclusive NFL game aired, and the quarterly loss widened to $552 million. Sports-acquired subscribers are the most churn-prone cohort in streaming, and the real test is retention through the spring once the NBA regular season ends and before the World Cup.

Our take: Both are right, and the bears are describing a cost rather than a flaw. Sports is the only reliable subscriber-acquisition mechanism left in streaming and Comcast owns more of it than anyone. The disclosure that roughly 170 advertisers bought NBA inventory with about 20% new to the company suggests the rights are expanding the commercial base, not just reshuffling it. The genuine risk is the timing mismatch: costs are straight-lined and certain, revenue is seasonal and contingent. We would watch the Q3 2026 subscriber number, once the NBA regular season and the World Cup are both behind it, as the honest test.

Debate: Does the Frozen Dividend Signal Prudence or Strain?

Bull view: The bull argument is that holding the dividend is straightforward prudence against a 2026 free-cash-flow step-down management disclosed in the same breath, and that with $19.2 billion of full-year free cash flow against $4.9 billion of dividends, coverage is not remotely at issue. Shareholders additionally received Versant shares and will collect that company's dividends.

Bear view: The bear camp responds that this is the second consecutive quarter of capital-return retrenchment, following October's buyback reduction from $1.7 billion to $1.5 billion, and that both were presented as continuity rather than as the changes they were. A company that genuinely believed its shares were mispriced at a 52-week low would be returning more capital, not less.

Our take: The policy is defensible and the presentation is not. Coverage is overwhelming and the 2026 cash-flow step-down is real and was disclosed honestly, so freezing the dividend is a reasonable decision. But describing it as an eighteenth consecutive year of dividend growth requires counting another company's future payout, and that is the kind of framing that erodes credibility with exactly the income-oriented shareholder base a 4.5% yield attracts. We read the substance as prudent and the disclosure as a small negative signal about how management wants this year to be perceived.

Model Update & Valuation Framework

How our October assumptions performed

Revised assumptions

Where the valuation sits

Market capitalization, net debt, enterprise value and the ratios above are derived rather than reported; the EBITDA and EPS denominators are now reported full-year figures rather than the run-rate annualizations we used in October, which makes these multiples cleaner. Note that these are trailing multiples on a business whose earnings management has guided lower for at least two more quarters.

Valuation judgment. At 5.2x trailing EV/EBITDA and 6.8x trailing adjusted EPS with a 4.5% yield, the shares remain priced for continued decline. The multiple has expanded very slightly from October's roughly 4.9x, which is consistent with the market granting partial credit for a trough date. The critical adjustment an investor must make is that all three of these multiples use 2025 denominators, and 2026 EBITDA, EPS and free cash flow will all be lower. On our estimates the forward multiples are meaningfully higher than the trailing ones, which is why the stock is cheaper on paper than it is in substance.

What the Versant spin does to the comparison. Every 2025 figure above includes a full year of Versant, and every 2026 figure will not. Management has committed to providing pro forma trending schedules before the Q1 report, which is necessary rather than generous, and until they arrive any 2026 versus 2025 comparison an investor constructs will overstate the decline. We will rebase our estimates when those schedules are published.

Thesis Scorecard Post-Earnings

Overall: Thesis unchanged in direction, sharper in resolution. The bear pillars performed exactly as written, with broadband revenue turning negative and margin compressing further, but the second bear point has been meaningfully bounded by a stated trough expectation, and the largest tail risk under the third has been retired by management's own capital discipline. Bull 2 is outperforming what we underwrote. Bull 3 has now weakened for two consecutive quarters and is the pillar we would most readily abandon.

Action: Hold. None of our three upgrade triggers fired. Broadband went backwards rather than flat, no conversion rate was disclosed, and the trough was hinted at rather than guided. We are adding a fourth trigger for the year ahead: a Connectivity EBITDA decline that narrows sequentially in Q2 versus Q1, confirming the trough rather than promising it. On the downside, a Q1 2026 EBITDA decline materially steeper than the NBA and pricing math implies would suggest the pivot is failing rather than progressing, and would move us to Underperform.

Bottom Line

Comcast printed its worst operating quarter of this cycle and the stock went up, which is the clearest evidence yet that this equity now trades on the duration of the pivot rather than the level of earnings. Adjusted EBITDA fell 10.3%, Media swung to a loss, domestic broadband revenue turned negative for the first time, and broadband subscriber losses widened on both comparisons. Against that, management cut the ambiguity: the EBITDA decline is now "the next couple of quarters" rather than "several," the NBA drag peaks in Q1 at roughly half the season's games, and the back half of 2026 is expected to improve.

Three months ago we wrote that a business compressing its own margin on an open-ended timetable has no catalyst. That criticism has been partially answered, and it is the reason the shares rose. It has not been fully answered, because a CFO who has been precise all year deliberately chose the word "improvement" over "growth" when asked directly, and because the single number that would let an investor underwrite the recovery, the free-line conversion rate, is now three quarters overdue with the assertion getting stronger and the evidence staying absent.

What genuinely improved this quarter sits outside Connectivity. Peacock broke a three-quarter subscriber plateau and added 8 million paid subscribers year over year while absorbing the NBA. Theme Parks crossed $1 billion of quarterly EBITDA with margin finally expanding, and Orlando hotel rates up 20% show the Epic halo is real and measurable. Management looked hard at Warner Bros. Discovery and walked away rather than stretch the balance sheet, which is precisely the discipline we said we wanted.

What got worse is the part that matters most. Roughly three quarters of consolidated EBITDA still comes from a Residential Connectivity business whose margin has fallen 520 basis points in two quarters, whose revenue is now declining, and which lost more customers this quarter than last. The dividend was frozen and the freeze was described as growth. We stay at Hold, with the first half of 2026 expected to be the worst of it, and we would rather pay a higher price in the second half for a confirmed trough than a lower price now for a well-argued one.