Broadband Inflects for the First Time Since 2020, Peak Dilution Passes, and We Upgrade

Results vs. Consensus

Q1 2026 Scorecard

The 117,000 improvement is the company's stated figure, computed on unrounded subscriber counts. The rounded table values would imply 118,000; we use the company's number rather than the arithmetic on rounded inputs. Note also that this release reports residential broadband only, where prior releases also broke out business broadband, so every broadband comparison here is residential to residential.

Year-Over-Year Comparisons

Comcast publishes pro forma reconciliation only for revenue and adjusted EBITDA, so EPS, operating income and free cash flow are shown on the as-reported basis alone and overstate the underlying decline. Versant contributed $1,769 million of revenue and $834 million of adjusted EBITDA in the year-ago quarter, which is the entirety of the gap between the two bases.

Quarter-Over-Quarter Comparisons

Versant sat inside the Media segment, so Connectivity & Platforms is substantially comparable across the two quarters (management noted a few smaller restatements within C&P in the March trending schedules). On that comparable basis, C&P adjusted EBITDA rose 5.4% sequentially and Residential margin recovered 250 basis points from Q4's trough of 34.6%. That step is, however, almost exactly the prior year's seasonal move: C&P EBITDA rose 5.4% from Q4 2024 to Q1 2025 as well, and Residential margin recovered 270 basis points then against 250 now. On a year-over-year basis the constant-currency decline has in fact widened in each covered quarter, from 3.7% to 4.5% to 4.7%. The sequential improvement is real but is not by itself evidence of a trough.

Revenue assessment. Pro forma revenue growth of 10.9% is the best headline Comcast has produced in years and is also the least informative, because $2.2 billion of it was a seventeen-day sporting event calendar that recurs on a four-year cycle for the Winter Olympics and never in the same quarter for the Super Bowl. Stripping it out, growth was low single digits. The more durable read is inside Connectivity, where total revenue fell 1.0% but the mix continued shifting: domestic wireless service revenue grew 15.0%, wireless equipment revenue grew 52.9%, and international connectivity grew 9.5%, against domestic broadband down 5.1%.

Margin assessment. Residential Connectivity margin of 37.1% is down 160 basis points year over year but up 250 basis points sequentially from Q4's 34.6% trough. Programming expense fell 7.8% on continued video decline while non-programming expense rose 5.7% on device costs and marketing, including the Legendary February spend. The important structural point is that the year-over-year margin decline of 160 basis points is roughly the same magnitude as Q4's 140 and Q3's 140, meaning the rate of deterioration has stopped accelerating even as the absolute level keeps falling.

EPS assessment. Adjusted EPS of $0.79 against $1.09 is a 27.5% decline, but the comparison is contaminated by Versant and by peak NBA dilution landing in the same quarter. Below the line, interest expense of $1,094 million was broadly flat and the tax expense fell to $706 million from $1,196 million. The cleanest way to think about the earnings power is that this quarter absorbed the single worst combination of items the year will produce: no Versant, peak NBA amortization, no rate increase, and maximum free-line dilution. Everything after this gets structurally easier.

Segment Performance

Connectivity & Platforms

Residential Connectivity & Platforms revenue detail

Comcast introduced a new disclosure this quarter, grouping domestic broadband and domestic wireless service into a "domestic convergence revenue" line and separating wireless equipment revenue. The reason is strategic: management wants investors to judge the household relationship rather than the broadband line in isolation. Convergence revenue fell 2.8% and convergence ARPA, average revenue per account, stands at roughly $85.

"Our convergence ARPA, or average revenue per account currently stands at roughly $85. For context, our telecom competitors are roughly double this amount on the same metric. This really underscores the significant growth opportunity in front of us, especially as we stabilize broadband and look to accelerate growth through wireless."
— Jason Armstrong, Chief Financial Officer

Assessment: The new disclosure is genuinely useful and the ARPA gap is the most compelling long-run argument management has made in four quarters. A household paying Comcast $85 against roughly $170 at a telecom competitor is either evidence of enormous monetization headroom or evidence that Comcast sells a narrower bundle at a lower price. It is probably both, and the free-line strategy is precisely an attempt to convert the first interpretation into revenue. It is also worth noting the framing serves management's interest: reframing the metric at the moment broadband ARPU turns negative shifts attention to a number that can grow.

Business Services Connectivity

Revenue grew 5.8% to $2,640 million with EBITDA up 3.8% to $1,476 million and margin down 110 basis points to 55.9%. Growth continues to come from enterprise solutions, with small and medium business under sustained fixed-wireless pressure. The T-Mobile MVNO for business customers is due to launch and will extend mobile into the mid-market and enterprise base.

Assessment: The margin decline moderated, from 150 basis points in Q4 to 110 here, which is the first improvement in that trend across our coverage. This remains a $10 billion-plus business compounding mid-single digits at 56% margins into a stated addressable market above $60 billion, and along with Theme Parks it is one of only two segments that has grown revenue and EBITDA in every quarter we have covered.

Subscriber Metrics

Assessment: Total residential customer relationships grew by 10,000, the first positive quarter in the period we have covered, driven by international turning firmly positive at 104,000. On the domestic side, the broadband number is the one that matters and it did what the pivot was designed to make it do. Video losses widened sequentially to 322,000, which is the counterpoint: Q4's improvement in video did not hold, and video still generates $6,256 million of quarterly revenue. Wireless at 435,000 is a fourth consecutive record or near-record and takes penetration to 16% of the residential broadband base.

Content & Experiences

Media and Peacock

Media revenue grew 60.8% to $7,280 million, including $2.2 billion of incremental Olympics and Super Bowl revenue. Excluding those events, growth was 12.7%, with domestic distribution up 21.3% ex-Olympics on Peacock and international networks up 12.4%. Segment EBITDA was a $426 million loss, which management had pre-announced as the peak of NBA rights dilution.

Peacock revenue surpassed $2 billion for the first time, growing 71%, and its EBITDA loss of $432 million includes the Olympics and Super Bowl costs. Peacock streamed a record 16.7 billion minutes of the Milan Cortina Games, described as more than double all prior Winter Games combined.

"We added 2 million net new subscribers in the quarter with revenue up more than 70%, putting Peacock on track to approach profitability for the first time next quarter."
— Mike Cavanagh, Co-Chief Executive Officer

Assessment: Peacock approaching profitability in Q2 would be a milestone the market has been waiting on since launch, and it is now a stated expectation rather than an aspiration. The subscriber trajectory across our coverage is striking: 41 million flat for three quarters, then 44 million, now 46 million. The loss widening sequentially in absolute terms is entirely event cost, and the Q2 setup is favourable with the NBA playoffs and the FIFA World Cup on Telemundo and Peacock. The risk we flagged in January stands: sports-acquired subscribers are the most churn-prone cohort, and Q3 is the honest retention test once the event calendar empties.

Studios

Studios revenue grew 21.2% to $3,426 million and EBITDA more than doubled to $555 million, driven by content licensing up 36.8%. Theatrical revenue fell 59.2% on a tough prior-year comparison. The Super Mario Galaxy movie has crossed $750 million globally, described as the biggest title of the year worldwide, with the franchise now at $2 billion.

Assessment: The EBITDA doubling needs a caveat the segment table does not carry. Management disclosed that the content-licensing strength was led by the renewal of a major series on Peacock, which is an intercompany transaction: it books revenue at Studios and cost at Media, and shows up as a $140 million elimination loss at the Content & Experiences level versus a $14 million benefit a year ago. Studios' standalone result is therefore flattered relative to the segment's contribution to consolidated earnings. Total C&E EBITDA of $331 million is the number that matters, and it fell 46.0%.

Theme Parks

Theme Parks grew revenue 24.2% to $2,331 million and EBITDA 33.3% to $551 million. Management disclosed that adjusting for roughly $100 million of Epic pre-opening costs in the year-ago quarter, EBITDA growth was over 7%. Orlando continues to drive higher per-capita spending and attendance across the resort. Offsetting this, Osaka faces China-related inbound travel pressure and Beijing a more challenging macro environment.

Assessment: Management volunteering that the 33% figure is really 7% adjusted is exactly the kind of disclosure that builds credibility, and it substantially reframes the segment. Underlying growth of 7% on a park still ramping toward full capacity is solid rather than spectacular, and the international softness is a genuine new negative that did not exist in prior quarters. We are moderating our enthusiasm here slightly: this remains a high-quality growth asset, but the headline growth rates through 2026 will flatter it as the Epic comparison base normalizes.

Key Topics & Management Commentary

Overall Management Tone: The most confident call of the four, and the first where the confidence rested on results rather than plans, with a running theme of a company that believes it over-corrected into pessimism and is now proving otherwise. Management volunteered the unflattering half of nearly every good number, quantifying the one-off contribution to the broadband improvement, the event contribution to revenue growth, and the pre-opening-cost distortion in Parks growth. That self-correcting instinct, sustained across three consecutive calls, is now a meaningful part of why we take the forward statements seriously.

1. The Broadband Inflection, and Exactly How Much of It Was Real

Residential broadband net losses of 65,000 against 183,000 a year earlier is the first year-over-year improvement in more than five years. Management described the milestone precisely and then immediately qualified its durability.

"First, despite what remains an incredibly intense competitive environment, broadband net losses improved by more than 100,000 year-over-year, the first year-over-year improvement since the fourth quarter of 2020. We also delivered the best wireless net additions of any quarter in our history. Together, these are early signs that the strategic pivot we've made in our connectivity business is underway."
— Mike Cavanagh, Co-Chief Executive Officer

The CFO then attributed the majority of the gain to a specific, non-recurring marketing event.

"We estimate these specific offers accounted for over half of our year-over-year improvement in subscriber losses."
— Jason Armstrong, Chief Financial Officer

Assessment: The honest arithmetic is that roughly 50,000 to 60,000 of the 117,000 improvement is repeatable and the balance was bought with a once-in-four-years event calendar. That still leaves a genuine underlying improvement, and management separately cited three durable drivers: connect volumes up for the first time in more than four years, lower voluntary churn, and a step-up in gig-plus take rates. The disclosure discipline here is the notable part. A management team spinning this quarter would have let the 117,000 stand unqualified.

2. Broadband ARPU Turns Outright Negative

Broadband ARPU declined 3.1%, having grown 2.6% in Q3 and 1.1% in Q4. The deterioration was pre-announced and management named the drivers and the expected duration.

"Broadband ARPU declined 3.1%. This is consistent with the pressure we signaled on our fourth quarter call, and reflects the absence of a rate increase at the beginning of the year, our new go-to-market pricing, including the legendary February offers and the impact from strong adoption of free wireless lines, which initially has a dilutive impact on broadband ARPU."
— Jason Armstrong, Chief Financial Officer

Assessment: Domestic broadband revenue fell 5.1% as a result, its steepest decline of the cycle, and this is the cost side of the subscriber win. The trade Comcast has made is now fully visible in the numbers: it is buying roughly 117,000 subscribers per quarter of improvement with a 5% decline in its highest-margin revenue line. That trade only pays if the subscriber base stabilizes and the free lines convert, which is precisely why the conversion evidence below matters more than any other item on the call.

3. Free-Line Conversion Gets Its First Data Point

After two quarters of asserting that a majority, then a "vast majority," of free wireless lines would convert to paying relationships, management finally reported observed behaviour from the earliest cohorts reaching their twelve-month anniversary.

"We're early in that role. But as I mentioned, we're really focused on life cycle management, managing those customers all the way throughout. And in the early cohorts, we've seen a significant majority of those customers rolling to paid. So we're -- we feel that will continue as we move forward and more of the lines roll in the back half of the year."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Management added that conversion will directly benefit broadband ARPU through revenue recognition as lines roll to paid in the second half.

Assessment: This is the item we have flagged as the load-bearing assumption in every prior note, and it has moved from claim to observation. It is still not a number: "significant majority" of "early cohorts" is not a conversion rate, a churn assumption, or a disclosed ARPU. But observed early-cohort behaviour is categorically different evidence from an intention, and it arrives alongside a second useful data point, that roughly 30% of line net additions now come from existing mobile customers adding lines, which is the behaviour of a satisfied base rather than a promotional one. We are raising our modelled conversion assumption on this.

4. Peak NBA Dilution Confirmed and Passed

Management had stated in January that Q1 would carry roughly 50% of the NBA season's games and represent peak EBITDA dilution. It did, and the CFO confirmed both the peak and the improvement expected from here.

"The first quarter was the peak volume with about 50% of the games played and the corresponding costs flowing through. So as a result, this quarter represents our peak EBITDA dilution from NBA costs."
— Jason Armstrong, Chief Financial Officer

Assessment: Two consecutive quarters in which management pre-announced a specific mechanical outcome and then delivered exactly it. Media EBITDA of negative $426 million is the trough by construction, and the second half of the NBA season plus the lapping of the first-year straight-line amortization make the comparisons progressively easier from here. Combined with Peacock approaching profitability in Q2, the Media drag on consolidated EBITDA has a defined and short remaining life.

5. Legendary February: Reach Converted Into Connectivity Marketing

The seventeen-day stretch containing the Milan Cortina Winter Olympics, Super Bowl LX and the NBA All-Star Game reached more than 225 million Americans and drove roughly $2 billion of advertising sales. The Olympics averaged 23.5 million viewers, the most-watched Winter Games since Sochi. Super Bowl LX averaged 125.6 million viewers, the most-watched programme in NBCUniversal's hundred-year history and the second most-watched ever. The NBA All-Star Game drew its largest audience since 2011.

"And third, we had a real company-wide moment with legendary February. We outperformed across audience, engagement and monetization. And importantly, we leveraged this massive reach to market our connectivity products at scale, a proof point that when we really lean in, we can move the needle."
— Mike Cavanagh, Co-Chief Executive Officer

Assessment: This is the first concrete demonstration of the conglomerate synergy argument that has justified keeping media and cable together, and it is genuinely persuasive on its own terms. It is also the reason more than half the broadband improvement is not repeatable, and management said as much. The strategic question it raises is whether Comcast can manufacture smaller versions of this effect regularly. The 2026 calendar offers the World Cup on Telemundo and the NBA playoffs; 2027 offers considerably less.

6. A Third Competitive Vector: Satellite Joins Fiber and Fixed Wireless

Management named satellite as a newly promotional competitor for the first time, alongside continued fiber overbuild and fixed wireless, and disclosed that fiber now overlaps about 55% of the residential footprint.

"Our assumption is that the market will stay highly competitive, fiber, fixed wireless and now satellite is getting more promotional. And what we focus on is what we can control and what matters for the customer anchored by the following. We have a great network that is on par with fiber and it does exceed the capabilities of fixed wireless and satellite, both of which are capacity constrained."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Assessment: The capacity-constraint argument is technically sound and is the same one Comcast has made about fixed wireless for two years, which has proved broadly correct: on the January call management described fixed wireless as having stayed "pretty consistent," with stability there, even as this quarter's remarks have it still marketing aggressively. Satellite is likely to follow a similar path, serving price-sensitive and rural segments rather than contesting the high-usage core. The more consequential disclosure is the 55% fiber overlap, which quantifies for the first time how far the overbuild has progressed against management's stated expectation that most of the footprint eventually gets overbuilt. That is the structural bear case, and it is roughly halfway realized.

7. Wireless: Record Adds, Premium Traction, and a Product Launch

Record net additions of 435,000 took total lines to 9.7 million at 16% penetration. Premium Unlimited, launched a year ago, now accounts for roughly 30% of postpaid phone connects with the premium base up roughly fivefold. Management launched Mobile+ the day before the call, including lifetime device protection at no additional charge, described as an industry first.

Management also cited the structural economics: roughly 90% of Xfinity Mobile traffic is offloaded onto Comcast's own network, two MVNO relationships now cover consumer and business, and acquisition costs are lower because the company sells into an existing broadband base.

Assessment: The premium data is the most encouraging wireless disclosure to date, because it addresses the historical criticism that Comcast could only compete at the low end. Thirty percent of connects at the premium tier, with the premium base up fivefold, means the product mix is improving alongside the volume. Combined with 30% of line additions coming from existing mobile customers adding lines, the wireless business looks less like a promotional land-grab than it did two quarters ago.

8. Free Cash Flow Falls 28%, Exactly as Warned

Free cash flow of $3,901 million fell 28.0% from $5,421 million, driven by the three headwinds management named in January: the non-recurring 2025 cash tax benefit, a significantly lower tax-legislation benefit, and the removal of Versant's cash flow. Net leverage ended at 2.3x, with management noting it will tick up as Versant exits the trailing-twelve-month calculation.

Assessment: No surprise, which is itself the point. This is the third consecutive quarter in which a negative outcome arrived exactly as pre-announced. Total capital returned fell again to $2.5 billion, from $2.7 billion in Q4 and $2.8 billion in Q3, and the buyback stepped down to roughly $1.3 billion from $1.5 billion in each of the prior two quarters. That is the second buyback reduction since the $1.7 billion of Q2 2025, and a third consecutive quarterly decline in total capital returned. It remains the weakest part of the capital-allocation story.

9. Management Says the Stock Is Undervalued

Asked about consolidation and structural options, the co-CEO made an unusually direct statement about the share price.

"I think we're undervalued, frankly, and the negativity on the business is something we need to work on changing people's sentiment towards a period full stop. And I think doing that by continuing to run the play that Steve just articulated really well is Plan A."
— Mike Cavanagh, Co-Chief Executive Officer

Assessment: Executives rarely say this on the record, and it is worth noting the tension with the third consecutive buyback reduction. If management genuinely believes the shares are undervalued and the operating trough is passing, the capital-allocation response would be to buy more, not less. The reconciling explanation is the free-cash-flow step-down, which is real. But an investor is entitled to weight actions above words, and the actions have been consistently more cautious than the commentary for three quarters.

10. Cable Consolidation Replaces Media Consolidation as the Open Question

The strategic question shifted this quarter from media M&A to cable-sector consolidation. Management reiterated a high bar while explicitly preserving optionality and naming partnership as an intermediate path.

"I think, in addition to that, we've got plenty of opportunities and have worked with others in the industry to partner around video or mobile or otherwise. So there are ways to benefit ourselves through scale in partnership terms, and we're open to doing that. And then ultimately, there's always bigger ideas that, as Brian said, open to strategic possibilities to create value."
— Mike Cavanagh, Co-Chief Executive Officer

Assessment: The framing is materially better than the media-M&A version from earlier quarters, because cable consolidation would be scale in the business Comcast actually operates rather than diversification away from it, and because partnership around video and mobile achieves part of the benefit without balance-sheet risk. Management's demonstrated willingness to walk away from Warner Bros. Discovery on price gives this optionality more credibility than it would otherwise have. We would view a disciplined in-footprint or adjacent cable transaction differently from the media deal we warned about in October.

11. Network Position and the AI Operating Layer

Monthly data usage on the network rose 10% in the quarter. Management described running hundreds of models with thousands of attributes to optimize acquisition, upsell, win-back and retention, alongside an enhanced marketing technology stack. Real-time 4K was launched during the tent-pole events, delivering live sports at lower latency than competitors.

Assessment: Usage growth of 10% is the single most durable argument in the bull case, because it is the mechanism by which capacity-constrained alternatives eventually fail to satisfy the high-value household. The AI and marketing-stack detail is harder to verify and easy to overstate, but the checkable outputs cited in prior quarters were real: a 20% reduction in trouble calls and a 35% reduction in repair minutes where the new technology is deployed. We treat the network position as genuinely differentiated and the operating-layer claims as plausible but unproven.

Guidance & Outlook

Implied trajectory: Q2 2026 is the last quarter of incremental pressure by management's own account, and it carries offsetting positives that Q1 did not: Peacock approaching profitability, the NBA playoffs, the FIFA World Cup on Telemundo and Peacock, and no Olympics cost. The second half then carries the free-line monetization and the lapping of the pricing reset. For the first time in our coverage, the forward table has more items pointing up than down.

Guidance style: Three consecutive quarters of pre-announcing negatives accurately and then delivering them. This quarter added the pattern of volunteering the unflattering decomposition of good numbers. We now treat management's directional statements as high-information, which is the primary reason we are willing to act on a trough that has been described but not yet fully evidenced.

Analyst Q&A Highlights

How Much Further Broadband ARPU Must Fall to Hold the Stabilization

The opening question went to the central trade-off of the strategy, asking how much more ARPU has to be surrendered to sustain the subscriber improvement, and whether the gains came against fiber, fixed wireless or both. The answer laid out the timeline and the offsetting levers but declined to name an ARPU floor.

Q: "Your broadband ARPU rate of decline actually moderated sequentially a little bit. I wonder if you could just elaborate a little bit on how much lower do you think broadband ARPU might have to go to maintain the kind of stabilization that you've seen? And then if you could just broaden the lens perhaps just to talk about where the improvement came from, was it relative to FWA? Was it relative to fiber? Was it relative to all of the above?"
— Craig Moffett, MoffettNathanson

A: "We do see some incremental pressure in Q2, but we do expect relief as we exit the year, and we talked about it. The primary drivers of the decline include the absence of a broadband rate increase, free wireless lines and migration to our simplified pricing I talked about it earlier, we were not competitive enough. We need to adapt our approach and pivot the business, and our focus is on getting to the other side of as soon as possible."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Assessment: No ARPU floor was given, which is the one genuinely unanswered question in the bull case. The admission that "we were not competitive enough" is unusually blunt for an incumbent and is the clearest statement yet of why the reset happened. On the competitive breakdown, the response was that improvement was seen across all competitive environments, with both connects and disconnects improving, which is a broader-based answer than a fixed-wireless-only recovery would have been.

Whether Wireless Momentum Can Extend Into Larger Households

A question on the durability of the wireless acceleration, the coming business MVNO, and whether reduced carrier subsidies would improve Comcast's win rate produced the most detailed disclosure on wireless economics of any call in our coverage.

Q: "I'm curious if you could expand further on some of the success you're seeing in wireless in terms of kind of moving up into larger families, you mentioned the business opportunity that's coming up with new MVNO. And also just within this context, what is Comcast doing to simplify the migration process for customers. And if carriers start to pull back on subsidies, your competitors do less on that. Does that help you get a better hit rate to move customers over to Xfinity Mobile?"
— Michael Rollins, Citigroup

A: "About 30% of our connects line net adds are coming from existing mobile customers adding more mobile lines, which is really important for us."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Assessment: The 30%-from-existing-customers statistic is the most valuable number in the exchange and was not in the press release. Existing customers voluntarily adding lines is the strongest available evidence that the product satisfies rather than merely attracts, and it directly supports the conversion thesis, since a household that has already added a second paid line is unlikely to churn a converting free one. Management also indicated subsidies are used selectively rather than as a primary weapon, competing instead on price and value, which protects the margin structure of the growth engine.

Whether the Broadband Improvement Persists Without the One-Off Boost

The most important question of the call put the caveat directly: if over half the improvement was promotional, does the full-year subscriber trend still improve? Management answered yes while restating the caveat rather than softening it.

Q: "From Jason's comments, it sounds like after the benefit of the year-over-year improvement in broadband subs is due to the sort of legendary sea promotions and the half was sort of organic based on some of the efforts you've had. If we expect those efforts to sort of gain more traction through the year. Can we expect the high-speed data subscriber losses for the year to improve versus last year?"
— John Hodulik, UBS

A: "Yes, I would say we do expect improvement year-over-year, but more than half of the benefit in Q1 was tied into the legendary February. We really leaned into that from a marketing investment and an offer investment. It was like I said, it was a great moment, and we took advantage of it."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Assessment: A clean commitment to full-year year-over-year improvement, made while restating the one-off caveat for the third time on the same call. Comcast lost 653,000 residential broadband subscribers across 2025 on the sum of the four reported quarters, so improvement against that is a low bar and management chose not to raise it. The commitment is nonetheless checkable, and it is the specific promise we will grade next quarter.

How Capital Will Be Allocated Across the Post-Versant Media Portfolio

A question on capital allocation across Parks, Studios and Peacock, and on whether Peacock's expected Q2 profitability would prove durable, drew a full articulation of the post-spin media strategy.

Q: "So as you look at your key assets in Universal Studios, Peacock, Theme Parks, they all seem strategically very important. […] And you said Peacock will be profitable next quarter. But is that -- should we expect consistent profitability?"
— Jessica Reif Ehrlich, Bank of America Securities

A: "And then because of our straight-line amortization of NBA rights as we look to the next season, so to speak, of NBA lapping itself I think the prospect for ongoing and durable profitability for Peacock is what we have our sights set on."
— Mike Cavanagh, Co-Chief Executive Officer

Assessment: The answer distinguishes carefully between approaching profitability in Q2, which is a stated expectation, and durable profitability, which is described as an aspiration contingent on the NBA amortization lapping. That distinction is the correct one and the caution is warranted, because a single quarter near breakeven that coincides with the World Cup would not establish a run rate. On capital allocation, the commitment to recycle Parks cash flow back into Parks above the cost of capital is the clearest statement of priority, and the pipeline named (UK, Frisco, Hollywood) supports it.

Whether Macro Weakness Is Reaching the Parks and Advertising Businesses

A question on consumer sentiment and its transmission into theme parks and advertising demand, paired with a request for the fiber-overlap figure, drew both a measured macro read and the quarter's most concrete competitive disclosure.

Q: "Just any color you might be seeing domestically in the parks or from maybe some of your ad partners if you're sensing any tone shift perhaps in the economic weakness... And then maybe just more of a housekeeping question. I think, Mike, in your prepared remarks, you did say fiber builds are accelerating... any update in terms of where you guys stand today, perhaps on a fiber overlap basis across your residential footprint?"
— Sebastiano Petti, JPMorgan

A: "I think inside the U.S., domestic to domestic, we haven't yet seen any significant impact in the parks business caused by higher oil, but I think that does not mean that it may not happen depending on the duration of the effect on price of gas and the like and airline tickets and so forth."
— Mike Cavanagh, Co-Chief Executive Officer

Assessment: The macro answer is appropriately hedged and matches the reported results, where domestic Orlando was strong and the weakness was international. The more useful output was the fiber-overlap figure of about 55%, delivered almost in passing. That is the first time Comcast has quantified how far the overbuild has progressed, and it lets an investor calibrate the structural bear case: if the endpoint is most of the footprint, the process is roughly halfway through, and the subscriber trend is improving rather than deteriorating at that midpoint.

What the Free-Line Roll-Offs Have Actually Shown

The closing question asked directly for observed conversion behaviour and whether wireless monetization would help broadband ARPU stabilize. This produced the single most consequential answer of the call.

Q: "I just wanted to ask about the wireless line to paid strategy in the second half. First, could you just talk a little bit about what you've seen in the free line roll-offs to date and the strategy that gives you the confidence in the successful conversion later this year. And then second, I was just wondering if you could talk about the related impact from the wireless monetization strategy on broadband subscriber trends. Could this also help broadband ARPU stabilize later this year?"
— Michael Ng, Goldman Sachs

A: "And yes, it will have a direct impact on broadband ARPU based on revenue recognition as those lines rolled to paid in the back half of the year, and that will be a tailwind."
— Steve Croney, Chief Executive Officer, Connectivity & Platforms

Assessment: The mechanism is now explicit: converting free lines does not merely add wireless revenue, it reverses the ARPU dilution those lines currently cause, because the revenue recognition flows back through the broadband ARPU calculation. That makes the second-half ARPU recovery arithmetic rather than aspirational, provided conversion lands. Paired with the observed early-cohort behaviour, this is the strongest version of the bull case management has articulated, and it is the specific reason we are willing to upgrade before the conversion completes.

What They're NOT Saying

1. Still no conversion rate. "Significant majority" of "early cohorts" is observation without measurement. No percentage, no churn assumption, no retained-ARPU figure, and still no commitment to disclose the outcome when it lands.
2. No broadband ARPU floor. Asked directly how much lower ARPU must go to hold the subscriber stabilization, management described drivers and timing but named no level. This is the largest remaining unknown in the model.
3. No 2026 EBITDA guide, for a fourth consecutive quarter. "Relief as we exit the year" is directionally useful and numerically empty.
4. The Studios result was not reconciled to the segment. Studios EBITDA doubled on an intercompany Peacock licensing renewal that simultaneously created a $140 million elimination loss at the Content & Experiences level. Management mentioned the driver but did not quantify the offset, and the segment table shows both without connecting them.
5. Video losses widened and went undiscussed. Domestic video lost 322,000 against 245,000 in Q4, reversing the improvement management highlighted in October as the best result in nearly five years. It drew no question and no commentary.
6. The buyback cut was not explained. Repurchases fell to roughly $1.3 billion from $1.5 billion, the second reduction since Q2 2025 and part of a third consecutive quarterly decline in total capital returned, in the same call where the co-CEO said the shares are undervalued. Neither the reduction nor the tension with that statement was addressed.
7. No sizing of the satellite threat. Satellite was named as newly promotional but not quantified in subscriber, geographic or ARPU terms, and no view was offered on which customer segments are most exposed.
8. Legendary February's cost was never disclosed. Management quantified the benefit precisely, at over half the subscriber improvement, but never stated what the marketing and offer investment cost, which sits inside the 5.7% increase in non-programming expense.

Market Reaction

• Pre-print setup: The stock closed at $29.37 on April 22, up 4.9% year to date against an S&P 500 up 4.3%, and down 14.8% over trailing twelve months. The 52-week closing range of $24.97 to $36.33 shows a stock that had recovered roughly 18% from its low but remained far below the prior year's level. A materially less distressed setup than either prior print.
• Reaction session (April 23, before-open reporter): Opened at $30.50, a 3.8% gap up, traded a range of $30.40 to $32.08, and closed at $31.64, up 7.7% or $2.27. Unlike January, the stock did not fade from its highs.
• Volume: 46.5 million shares against a 32.9 million thirty-day average, 1.4x normal.
• Benchmark: The S&P 500 fell 0.4% on the session, so the entire move and then some was company-specific.

The largest reaction-day move of the four quarters we have covered, and the cleanest. The stock gapped up 3.8% on the press release, which contained the broadband and wireless numbers, and then added further ground through and after the 8:30 call rather than giving any back. That pattern is the inverse of January, when the shares opened down on the release and recovered on the call, and it indicates the market read both the print and the commentary as confirming.

What the market bought. Not the earnings, which fell 27.5% year over year, and not the cash flow, which fell 28.0%. It bought the inflection in the metric that has driven the de-rating for five years, plus the confirmation that the worst quarter of the transition is now in the rear-view mirror. This is a stock trading on second-derivative improvement, and this quarter delivered the first unambiguous instance of it.

Context for the level. At $31.64 the shares are up 15.8% from the $27.32 close on the day we initiated in October and up 8.2% from the January close of $29.24, against an S&P 500 that has been roughly flat over the most recent stretch. The stock remains 14.8% below where it traded a year earlier and roughly 13% below its 52-week closing high of $36.33, so the re-rating is early rather than complete.

Street Perspective

Debate: Is the Broadband Improvement a Trend or a One-Quarter Event?

Bull view: The bull case on the Street holds that the underlying drivers are structural rather than promotional: connect volumes rose for the first time in more than four years, voluntary churn keeps improving, gig-plus take rates stepped up, and roughly 40% of the base has migrated to simplified pricing with the majority to follow by year-end. Legendary February amplified a trend that already existed; it did not create one.

Bear view: The bear camp contends that management itself said over half the improvement was bought with a non-repeatable event, that broadband ARPU fell 5.1% at the revenue line to deliver it, and that a company spending its most valuable marketing inventory of the decade to reduce quarterly losses from 183,000 to 65,000 has demonstrated the cost of stabilization rather than its achievement. The Winter Olympics do not return for four years.

Our take: The bears are right that the headline overstates the run rate and the bulls are right about direction. Strip the one-off and the improvement is roughly 50,000 to 60,000, which on a base losing 178,000 a quarter as recently as December is still the first genuine turn in five years. What decides the debate is the second half, when free-line conversion adds revenue without marketing spend. We think the balance of evidence favours the bulls, but we would hold the view lightly until the Q2 print shows improvement without an Olympics behind it.

Debate: Does the Convergence ARPA Gap Represent Headroom or a Structurally Cheaper Product?

Bull view: A growing consensus view is that convergence ARPA of roughly $85 against telecom competitors at roughly double is the single largest quantified opportunity in the story. Comcast has 16% wireless penetration of its own broadband base and a capital-light MVNO with 90% traffic offload, so each converted household carries very high incremental margin. Closing even part of that gap transforms the earnings profile.

Bear view: Some sell-side desks argue the gap exists because Comcast sells a different product: fewer lines per household, lower-tier plans historically, and a customer base that chose Xfinity Mobile precisely because it is cheaper. Reframing the metric at the exact moment broadband ARPU turned negative is presentational, and the gap may simply describe a value-segment position rather than latent pricing power.

Our take: The bears have identified real selection bias, but the premium data undercuts it. Premium Unlimited now accounts for roughly 30% of postpaid connects with the premium base up roughly fivefold, and roughly 30% of line additions come from existing mobile customers adding lines. Both indicate the mix is moving up, not sitting still at the value end. We read the ARPA gap as genuine headroom that will close slowly, over years rather than quarters, and we would not underwrite rapid convergence to competitor levels.

Debate: Is the Trailing Multiple Real, or an Artifact of the Versant Spin?

Bull view: The bull argument is that at roughly 5.6x trailing EV/EBITDA, 7.9x trailing adjusted EPS and a 4.2% dividend yield, the equity is priced for permanent decline in a business that just posted its first broadband improvement since 2020, with peak dilution passed and free-line monetization ahead. Any re-rating toward even 6.5x EBITDA is a substantial move.

Bear view: The bear camp responds that every one of those multiples uses trailing denominators containing three quarters of Versant earnings that no longer exist, so the true post-spin multiple is meaningfully higher. Forward earnings must also absorb a full year of NBA rights, continued ARPU decline, and free cash flow running roughly 28% below the prior year. The stock is cheaper on paper than in substance.

Our take: The bears are arithmetically correct and we have made the same adjustment in our own numbers. Versant contributed $834 million of adjusted EBITDA in the year-ago first quarter alone, so the trailing figure overstates the earnings base by a material amount and the honest post-spin multiple is closer to 6x than 5.6x. That said, 6x for an asset with the network position, cash conversion and optionality Comcast has, at the point its worst quarter is passing, is still cheap. The bear argument narrows the discount; it does not eliminate it.

Model Update & Valuation Framework

How our January assumptions performed

We were too pessimistic on subscribers and too optimistic on price. The net effect on revenue is roughly neutral, but the mix is better than we modelled, because subscribers are a durable asset and promotional ARPU dilution reverses when free lines convert.

Revised assumptions

Where the valuation sits

The trailing-multiple caveat, stated plainly. The enterprise value above is post-spin, while the trailing EBITDA and EPS denominators include Q2 through Q4 of 2025, all of which contained Versant. Versant contributed $834 million of adjusted EBITDA in the first quarter of 2025, the only quarter for which Comcast has disclosed the figure. Scaling that single disclosed quarter across the three trailing quarters that still contain Versant places the true post-spin multiple closer to 6x than the 5.6x shown. We use approximately 6x in our own work and would encourage readers to do the same. We have not modelled Versant's Q2 through Q4 contributions, because Comcast has not disclosed them and estimating them would be exactly the kind of derived figure that ought not appear in a valuation.

Valuation judgment. At roughly 6x forward-adjusted EV/EBITDA with a 4.2% dividend yield, Comcast trades at a discount to its own history and to any reasonable sum of its parts, at the point where its worst operating quarter is behind it. The re-rating case does not require growth. It requires only that the market stop pricing terminal decline in broadband, and the first evidence against terminal decline arrived this quarter. A move from 6x to 6.75x, still a deeply discounted multiple, would be roughly 20% on the equity, with a 4.2% yield paid while waiting.

Thesis Scorecard Post-Earnings

Overall: Thesis materially strengthened. Both bear pillars that had been Materializing since initiation moved to Contained in the same quarter, for the first time. Bull 1 strengthened on record wireless plus premium mix evidence. Bull 3 remains the weak link and has not improved in three quarters. The composition of the thesis has flipped: at initiation the bears were being confirmed and the bulls were back-end loaded, and today the bears are being contained while the bull evidence has begun to arrive.

Action: Upgrading to Outperform from Hold. Our third upgrade trigger, C&P EBITDA visibly troughing with management willing to date the recovery, has fired: management has now bounded the pressure to one more quarter and dated relief to the exit of this year, with peak NBA dilution confirmed and passed. Trigger two, a disclosed conversion rate above 50%, is partially met by observed early-cohort behaviour. Trigger one, two consecutive quarters of flat broadband, has not fired and we are not pretending otherwise.

Revised downgrade triggers: we would return to Hold if Q2 residential broadband losses widen back beyond roughly 150,000, if the free-line conversion rate is disclosed below 50% or management stops discussing it, or if broadband ARPU fails to improve sequentially in the second half. We would move to Underperform if C&P EBITDA decline widens rather than narrows in the second half, which would invalidate the trough thesis entirely.

Bottom Line

For three quarters we have written that Comcast was cheap because management was deliberately compressing its own margin, and that the missing ingredient was evidence the sacrifice would work. This quarter supplied it. Residential broadband losses of 65,000 against 183,000 a year earlier is the first year-over-year improvement since 2020, wireless set another record at 435,000 with the premium mix improving underneath it, and connect volumes rose for the first time in more than four years. Peak NBA dilution arrived precisely where management said it would three months earlier and is now behind.

We are upgrading to Outperform. The trigger we set in October, a Connectivity EBITDA decline visibly troughing with management willing to date the recovery, has fired. The trigger we cared most about, free-line conversion, has moved from a claim repeated in three consecutive calls to observed behaviour in the earliest cohorts, with a mechanism now made explicit: conversion reverses the ARPU dilution those lines currently cause, so the second-half recovery is arithmetic rather than hope.

We are not upgrading on a clean quarter. Broadband ARPU turned outright negative at 3.1%, domestic broadband revenue fell 5.1%, free cash flow dropped 28%, satellite joined fiber and fixed wireless as a third competitive vector, fiber now overlaps about 55% of the footprint, and the buyback was cut for the third consecutive quarter in the same call where the co-CEO said the stock is undervalued. Most importantly, management volunteered that over half the broadband improvement came from a promotional moment that will not repeat at that scale, which puts the run-rate improvement nearer 50,000 than 117,000.

That honesty is a large part of why we are willing to act. Three consecutive quarters of pre-announcing bad news accurately, and now a quarter of discounting its own good news, is the behaviour of a team that expects to be judged on 2027 and is managing the interim expectations rather than the interim narrative. At roughly 6x post-spin EBITDA with a 4.2% yield paid while waiting, the market is still pricing terminal decline in a business that just produced its first evidence against it. We would rather own that setup than wait for the second data point and pay a higher price for it.