T-Mobile US, Inc. ($TMUS)

All right. I think people are settling in, but we'll get started. Hi, everybody. All right. My name is Kutgun Maral. I'm the media, cable and telecom analyst at Evercore ISI, and we're very pleased to have with us at our conference, Peter Osvaldik, the CFO of T-Mobile. Peter, thanks so much for being here today.

Thanks for having us, again.

Absolutely. And I think you have some pointing to do.

Absolutely, yes. There was a disclaimer slide. It looks like it disappeared, but there's the legal [indiscernible], I'll make forward-looking statements, non-GAAP metrics. Please look at our filings for all the risk factors and recons for those.

All right. Perfect. So let's start big picture. I think since the February update, T-Mobile has raised guidance on accounts, EBITDA and free cash flow, and the first quarter results really reinforce the accounts and ARPU framework. From the CFO seat, what are the changes that -- what have been the changes that have given you the most confidence since February to increase that outlook? Is it the demand environment, quality of net adds, monetization, free cash flow conversion or anything else that you call out?

Yes. I mean, fundamentally, obviously, since February, there haven't been any big shifts in the strategy or the operational. I mean so it really is just doubling down and executing against that strategy that we laid out, which most simplistically put is the best network, best value and best customer experience strategy. And that's really what amounts to having industry-leading NPS and being able to have a very consistent profitable model that delivers. And really, that's what you saw in Q1. It was really execution against that, that allowed us to put up not only just industry-leading, but really multiples of the industry results, whether that's in postpaid account net additions that were 217,000, whether that was service revenue of 11% growth, core EBITDA, 12% growth, or what I always look at very, very important to me, certainly from a long run perspective, is how do you translate that growth into free cash flow [indiscernible]. And we delivered 24% conversion of service revenue to free cash flow after investing sufficiently in the business. That's important. You can always cut your free cash flow in the short run, but that will hurt you in the long run. So after you've invested to keep the value creation alive, we have 24% free cash flow margin. That was in Q1. If you look across what was the demand environment, we saw strong demand across every [indiscernible], whether it's our top 100 markets when that's cohort 1, 2 and 3, as we kind of subsegregate the top 100 markets, smaller markets and rural areas, broadband continues to be an absolute area of strength for us, and we grew our 5G broadband product faster in Q1 than we did even a year ago. And our total broadband net additions were over 500,000 for the quarter, and I say that. But of course, we all know volume isn't just volume for the sake of volume. You actually have to convert it into quality. And so when you saw what was going on in Q1 certainly, some of our competitors were going out there with pretty aggressive device-centric promotions. And we chose to not match a lot of those. And when you have a proposition that focuses on value beyond just the device that you upgrade or buy new once every few years, even longer these days, you can actually come up with very effective promotions that don't have to go so deep into the device-centric offers because customers get a lot more value, whether that's the network side or value elements or the experience, again, all invested in NPS. And what we actually saw is we chose not to match those offers because we saw those as CLV not attractive for us. But what we saw was not only the delivery of the volumes that we had in Q1, but also really encouraging signs like port-in ARPUs that were 20% higher than port-out ARPUs in the quarter. And that tells you about the quality of the customers that we're taking in. And we can get into more of that a little bit later. But all of those signs point to the model and the strategy that we laid out, which is, again, the best network, the best value, the best experience coupled with underpenetrated market segments and new opportunities continues to be something that we're very excited about, certainly for the balance of this year, but actually over the long term as well.

That's great. And you talked about this a few times, but let's maybe [ zero in ] on customer lifetime value a little bit as to your point, promos come in and out, and I think from the outside, it's a little hard to figure out how much of any operator's performance is really coming from aggressive promotional activity versus something that's a little bit more tangible and durable as you look out in the next few years. So you gave a few examples. But how do we think about the CLV that you're getting from your new customers? And what are the kind of internal proof points that give you confidence that you're headed in the right direction as opposed to just others maybe hitting a quarter?

Yes. I think it's important. If I back up a little bit, is when I see the most astute investors in this really look at it the same way we do, which is it's not about just 1 KPI, for example, postpaid phone lines was always the hyper focused on KPI. But it's also what's your ability to translate unit growth on 1 KPI into actual value creation that matters for investors in the long term. So can you take that unit growth and translate it into service revenue growth? Are you running the company efficiently enough where that service revenue growth then actually accretes down to core EBITDA growth? And of course, are you investing sufficiently? These are CapEx-heavy industries? Are you investing sufficiently and yet still able to convert that service revenue into free cash flow. And that's where, again, we have fundamentally multiples of what the other competitors in this industry have because structurally advantaged and we'll have that for a very long time into the future. So it's -- those are the external things that I would look at as an investor because it's hard to get a sense of, well, what were CLVs of customers. I can tell you, again, internally, what we're looking at, of course, is what are unit economics looking like? What are things like I just mentioned, port-in APRUs looking like, 20% higher porting ARPUs than port-out ARPUs. From a CLV perspective, in Q1, what we actually saw, again, and this is fundamentally, as you have the best product, the network and more and more consumers are hearing about it, believing it, Q1 happened to be of recent switchers to T-Mobile, the highest-ever percentage of those cited network quality is the reason for switching. So that flywheel of network perception that has a long, long runway ahead of us is starting, but you also have the best value. What you get for what you pay, certainly, the network is a big part of that. But all the other day-to-day benefits that you get as a T-Mobile subscriber. So those are the things that bring in high-quality subscribers, high CLV subscribers, again, we talked about we'd have to match deep -- what I see is CLV negative device offers because we have all of these other tangible things that are bringing customers in. And we actually saw CLVs in Q1, up double digits year-over-year. And that's actually continued into Q2. So that's the focus that we have. It's not just the volume but the value creation for investors. And again, you see that manifest itself in that differentiated service revenue, core EBITDA, free cash flow, those are the things to really focus on.

Yes. And I think you're helping shape the way that investors look at the story by having switched disclosure around a little bit and I think off towards better understanding around the ultimate economics of the business. I think folks are still settling into some of the new account and ARPA disclosure framework. So what do you think that the Street is maybe still getting wrong in sets as we look into the new disclosure. In particular, how should we interpret account churn when newer accounts, broadband-only accounts and acquired accounts all carry different churn profiles?

Yes. I mean I think what some individuals are still maybe not seeing holistically as just pan out for the hyper focus on 1 KPI and so much more focus on are you translating that into tangible value creation, right? Service revenue, EBITDA, free cash flow, the things that we talked about. I think that's the most -- and fundamentally, the most important thing to look at as to is this the right investment? And how is every company differentially able to create that value. When you look at just account churn, in particular, I know everybody has been used to postpaid phone churn for the longest time, and we did disclose that in Q1. That was only up 3 bps year-over-year as we saw more of a normalization from depressed periods in the past. Account churn is structurally always going to be higher than postpaid phone churn. It's really just a math equation for us. And there's a few underlying reasons for it. One of those being as accounts come on board and tenure, they -- not only tenured accounts have lower churn, but tenured accounts, particularly with MPS being where it is, also tend to grow in size. Certainly, other products beyond phone but also grow phone lines on a per account basis. So what you see happening there is as an account is tenured and churns less, well, they have more postpaid phone lines in that account than a new account that's churning higher. So that's a difference that you see between postpaid phone churn and account churn. The other thing is the success that we've had with our -- not only our 5G broadband offering, but bringing on Lumos and [indiscernible] customers on board. Broadband obviously has a structurally higher churn profile in the U.S. than mobile. And so that's a math equation of the success of 5G broadband you're going to have higher account churn than postpaid phone churn. But it came in exactly as we expected, as we anticipated, it's a trend I expect to continue for the balance of the year, certainly, -- that doesn't mean that every day, we don't wake up and say, well, how do we get account churn even lower, both short 10-year accounts as well as longer tenure accounts. But I wouldn't read anything more into it than structural math reasons.

Yes. Perfect. And maybe switching gears a little bit to postpaid ARPA. Growth in Q1 was well ahead of expectations. You pointed to a roughly 2% growth in Q2 before reaccelerating in the back half and landing at 2.5% to 3% for the full year. What gives you confidence that the second half acceleration reflects underlying drivers like premium mix, more products per account, the success you're seeing with business, broadband attachment, for example, rather than just some of the easier comps that you have.

Yes, it's -- there's a lot of quarter-over-quarter noise. And again, kind of a function of math and that makes the quarter-over-quarter year-over-year comparative is a little bit difficult. And that's why it's important to pan out to that 2.5% to 3%. Couple of the reasons why the quarters are a little bit different from a year-over-year perspective are, one, in early Q2 of 2025, we did our last round of rate plan optimization. So Q1 of this year was the last meaningful kind of year-over-year benefit from that. The other thing that you saw is very highly accretive decisions like acquiring U.S. Cellular. That came with mathemically lower ARPA as did the Metronet and Lumos acquisitions of customers as we bring them on and then you can expand that relationship over time. And so until we lap those things in Q3, you're going to see comparatives that are a little bit odd. And that's why the 2% for Q2 fundamentally is the right number when you look at the oddities of the math. But then you need to step out and say, well, okay, what's driving the 2.5% to 3% full year growth? And those are the things that we're most excited about. And some of those are, again, structural advantages that we have. First being our front book to back book advantage. And you know the story. I mean partly the beauty of having grown up as the value leader, being seen as a far distant third or fourth in terms of network is you had value seekers come to you. People that generally chose you for price, and that means that we have a structurally lower back book than the competitors do. Now that you've arrived at the station where you have the best network and are attracting network seekers that tend to be higher CLV, higher ARPA customers, well, they're coming on board. And so every ARPA account that's coming on board is actually structurally accretive to you because of this back book, front book dynamic that we have. And you see because those network seekers are coming to us, you have really high loading on our premium plans of lines on new accounts that came in Q1, 60% of those continue to be on premium plans, which is about double the base. So that continues that flywheel of new accounts coming on board, increase ARPA. The second, of course, is -- if you have an NPS that's completely differentiated and your customers trust you, then you can grow that relationship over time, whether that's postpaid phone lines that are added, whether that's 5G broadband, other connected devices that allows you to grow ARPA over time. And those are things that we continue to see happening as well. And then third, and probably the least on our waterfall is we're going to continue to be thoughtful around rate plan optimization. And under a more-for-more construct, where can we see legacy rate plans potentially moving up from a pricing perspective. And so that's how we think about those 3 items, and they're all -- the first 1 is certainly very unique to us, having that back book, front book structural advantage, and that's how you get the 2.5% to 3% growth year. So we're very excited about that.

Yes. That's fantastic. And maybe just 1 more on the postpaid side for now. When you think about the next few years of account growth, can you maybe rank the major opportunities between network seekers, small markets in rural areas, T-Mobile for Business, U.S. Cellular, broadband, maybe not just by headline TAM, but where you're seeing incremental sales and marketing dollars generating the best returns?

Yes. This is 1 of the most exciting parts I think about the company is the fact that just like we laid out in February, it's across all of these categories. U.S. Cellular is now, we consider it an integrated part of which segment it's in. But there's fundamentally a few things we're looking at that drive growth. One of those being network seekers that we talked about. Now that you're at the point where you have the best network and more and more people out there, not to their own customers, but prospect, future customers, as we call them, believe that we have the best network, you start getting a flywheel of that's 20-plus million accounts out there that historically chose either AT&T or Verizon because they believe they have the best network. And at the time, they did, we didn't. But we're in a totally different place where years ahead, we have all the basically components to continue to be years ahead of that. And so continuing to take network seekers as high CLV, high ARPA customers is a very exciting area for us to grow. And that's across smaller markets, rural areas, our top 100 accounts or markets across the board, smaller markets and rural areas. Now with the acquisition of U.S. Cellular, we've hit about 24% share, but there's a long runway there. And what's really interesting, both network seekers and smaller markets and rural areas, if you have 5%, 10% share, it's really actually in a fish to get that flywheel going because you don't have a word of mouth. And the vast majority of customers, when they consider this category and switch are still talking to friends and family as the works for you where I live, work and play. And when you achieve 20%, plus 20%, now you've got a flywheel of word of mouth, particularly when you have an NPS score that's way above the competitive set, and you have that best network perception catching up and starting to move forward. So it makes it even more efficient. You've got 5G broadband, which is an exciting opportunity. We just upped our subscriber target to 15 million subscribers by 2030. You obviously have fiber on top of that. You have T-Mobile for Business, which is just a great opportunity to continue to grow. And it's a great kind of indicator of where the network is because large enterprise, government customers, they test the network. They also want to see more than just pure connectivity of their phones from a network perspective. What they want to see is are you bringing novel solutions to me that can actually help solve business challenges? And if you do that, then I'll sell you the postpaid phone business too, that's no problem. But I don't want just the postpaid phones [indiscernible], I mean if I'm a CIO at a large enterprise, why would I do that? What I want you is to solve my problems. And because of where our network technology is, we're able to offer completely unique solutions there. And so that allows a lot of account growth into the future because you win the hearts and minds of CIOs and large enterprises, large government entities because you're solving problems, whether that's automated balls and strikes or what we just did with the PGA Tour, there are so many examples of unique solutions F1, that we're able to provide to customers to solve real pain points. And then we get the connectivity business with the phone as well. And then beyond that, of course, we're going to look at, to a lesser degree, from a pure account perspective, but how do you grow into smart adjacencies, like financial services, physical AI is a great opportunity for us in the future. That's not even in the plan from a revenue and EBITDA perspective. But those are all the areas we look at. And what's beautiful is we don't have legacy declining businesses that we need to shield. So every area that we're investing in is an area of growth and strength for us, which is, again, one of those unique structural advantages that we have. So we see growth across the board. .

And the good thing for my seat is it's not -- these are not plans that you have on a board deck, the success you've already demonstrated in some of these top 100 markets, like whenever I hear the market share for [indiscernible] I think it is, boggles my mind. And when you think about small markets and rural areas, there's no reason why you shouldn't have, at a minimum, your fair share from a market share [ perspective ].

Yes. I mean you mentioned New York where we have over 50% market share and growing, because now we're attracting those network seekers on top of the value seekers that have come to us. And you're right, when you have that formula, the question is, well, why can't you get there? Now that's not our plan. That's not underpins the guidance that we gave you out there, but it certainly is a long run opportunity for us.

Yes. So we talk about postpaid accounts and ARPA, maybe in the interest of completeness, maybe you could talk a little bit about prepaid and the wholesale markets as well and what you're seeing over there?

Yes. I mean we think of our postpaid and prepaid brands. And I'll get to wholesale in a second, but we think about prepaid and postpaid very complementary products. We have leading brands in both of those categories. And for the longest time, we've always been able to engage with prepaid customers. And some portion of those customers have always moved into the postpaid brand over time as they saw the value attributes or wanted more of the things that postpaid offers, including more device-centric ability to finance things like that. I think when you pan out, what you're seeing at an industry level is more prepaid customers are choosing to interact with postpaid brands directly than they have in the past. And I think probably they see the value and they're moving there. And so -- that's why, for us, we step back and say, well, what's happening with postpaid and prepaid service revenue together. And when you see that, I mean, we delivered total service revenue of 11% growth. Postpaid was 50% growth in Q1 on a year-over-year basis. And I expect Q2 from a prepaid perspective, we'll probably see the same dynamic we saw Q4 to Q1 play out in Q1 to Q2, and that's because we're seeing these great CLV accretives moves into postpaid and the postpaid brand. And so that's something we'll definitely continue. And I think it will be an industry trend for a while. On wholesale, it's playing out exactly as we foreshadowed. And that's fundamentally trackpone and DISH as MVNO partners moving off and then other growth continuing beyond that. So Q2, again, I expect we'll see the same sequential change in terms of absolute dollars in service revenue there as we saw in Q1 to Q4. And then 2026 is the year where it flattens out to slightly starts growing as TracFone and DISH have fully rolled off by then.

Perfect. Okay. I have a few questions on the broadband side. So maybe fixed wireless. Over there, you've been very clear that this is a fallow capacity monetization story, not a capacity give away. I think we still get the question of, as usage keeps rising and incremental spectrum supply comes into the market, what are the leading indicators that tell you that fixed wireless can keep scaling without crowding out your mobile economics?

Well, as you know, we apply a very unique model to it, the fallow capacity model, as you call it, where we project forward in a 36 million [indiscernible], we disaggregate the country in the 36 million hexbins. And for each of those hexbins,look at -- what is the capacity that's going to be created by the network there? How much of that is going to get utilized by mobile phone traffic because we have new subscribers coming on, but also mobile phone usage growth. And then where we have excess capacity using that model is where we approved the sale of fixed wireless. And then we further cap it, by the way, for market share, make sure we're basically assuming conservative rational market share percentages. And that's how you get to be allowed to sell fixed wireless in haxbin. And so what you should look at externally is, well, what are speeds doing? That's the great proxy for capacity. You've seen what the 5G business -- 5G broadband business is done in terms of customer growth, let's say, over the last 2 years, significant growth in customers, yet you've seen speeds on the network increase by 50%. Of course, usage per customer has gone up by about 30%, a little over 600 gigabytes a month, which is exactly what we planned, but speeds have increased by 50%, coincidentally, they're also faster than 50% faster than the next nearest fixed wireless competitor. Now the average download speed we're seeing over WiFi is about 250 megabits a second. But what's really fun is on our latest gen routers because this is a technology, both from a network perspective and a router perspective, that is rapidly evolving. Our latest gen routers are putting 400 megabits a second down over WiFi, which is effectively fiber lived experience over WiFi. This is why it has the highest NPS as a category over all others, including fiber. When you add in the ease of installation, the experience, the value, it's just fundamentally such an exciting business to be in. So look at speeds, look at customer growth and look at continued this business, giving us a lot of tailwind.

Yes. It doesn't seem to be decelerating at the moment. You've talked about fiber. And let's talk a little bit about fiber. You've -- I want to get away from the debate that is, at least for now on whether or not convergence is real or is it a myth, we'll get to that in a second. But maybe a more relevant question for T-Mobile is the return on capital that you're seeing with fiber. So maybe you could talk a little bit about what you're seeing on a penetration perspective, customer quality, and operating cost that supports the double-digit IRRs in these JV structures that are somewhat unique to you?

Yes. Well, of course, it's early days. We're about a year into some of these. We're seeing, what I would consider very encouraging results, basically playing out the thesis, which was when you combine the expertise of the fiber players and the partners that we have in terms of actually building and laying fiber, coupled with the brand, the distribution, the customer relationships that T-Mobile brings that we'll be able to actually get penetration better and faster over time. And what we're seeing in the early cohorts, again, it's very early, is 1-year 20% penetration rate, which is pretty encouraging for greenfield fiber builds and you're seeing us start rolling out that T-Mobile playbook for example, good, better, best rate plan tiers and having customers self-select up those rate you bring them in with the advertising and then they see the value in the higher tiers and self-help up. So again, very early days, but we're encouraged by what we're seeing from a thesis perspective playing out. .

That's great. And maybe staying with the fiber conversation. And we've certainly done this. I think everyone has done this in terms of modeling out passings and subscribers. But think for a lot of us, the [ Retailco and Infraco ] economics are a little harder to translate to consolidated EBITDA and free cash flow. So maybe you could help us think about the biggest swing factors between a fiber JV being strategically attractive and actually being free cash flow accretive for T-Mobile.

Well, there's no, I would say, disconnection the 2 for us because what we're looking at is what are businesses that can continue to ultimately be free cash flow accretive for us, the ultimate value creation engine for investors. And so when we think about -- well, what does a JV partner have to look like for us to be attractive? And why we've been so disciplined in what we've done here is, one, it starts with geography, like they have to be able to build in geographies that we can get first or very near first to fiber because that allows you to get the penetration rates using the T-Mobile brand, and that's the #1, I think, factor penetration rates, obviously, in ARPUs in terms of what are your ROIC is going to look like on this. The second is how can you get the right partners in this space. We're not experts at laying fiber, certainly not yet, right? So how do you partner with the right people we bring the brand customer distribution marketing know-how, they bring the fiber build know-how and together, it creates the magic that the JVs are if you have the right partnership model. And the third to your point is -- and this is unique to the structure, which is, hey, if you're building very early, it's very capital intensive. So why don't you do a JV like structure? First reason is the partners want to stay in, right? They're excited about the opportunity they want to stay in as well. And doing the JV structure is really smart from an ability to lever up and provide the funding needed for this early build and get again, first to fiber, get the T-Mobile fiber machine working on that to get the penetration rates on it. we're growing very fast now. So obviously, what you're going to see now, like in any subscription business is customer acquisition costs come on board as you're scaling that platform really rapidly. And then ultimately, it turns to free cash flow accretion in the later years. But we think about those, there is no deal that we've ever looked at that says, well, I want this deal for this reason, and it won't be free cash flow accretive to investors in the long run. It absolutely has to be.

Understood. Let's talk a little bit about competition, especially on the broadband side. When you think about Verizon and AT&T, leaning harder into fiber and [indiscernible] bundled offers, what are you actually seeing in these overlap markets? And is there evidence that their fiber footprint changes, wireless switching activity, churn or promotional intensity? Or is the impact still smaller than I think what a lot of folks assume?

Yes. So you came back to the conversion, [indiscernible] as we call bundling. We're not seeing anything different there than I think we've laid out this thesis and you see it play out in the results. This has been in the run rate for years and years and years. It continues to be in the run rate. And you see the differentiated volume and financial results that we're able to put out there. And we've mentioned things like, hey, even in places where our competitors have fiber and are building fiber, our share of postpaid continues to grow. And so those are the things to look at. Nothing fundamentally has changed here at all. We continue to grow. We continue to deliver the differentiated financial results. Customers still feel the mobility aspect of this is the [indiscernible] considered purchase. And so there's nothing really new here, nothing that's changing any of the elements of it. But -- and if you step back and look at Q1, it's a great example. You've got AT&T who's been touting convergence and the fiber build strategy probably because they don't want to explain that what they're really doing is spending a lot of capital to upgrade legacy tech to fiber, which makes sense for terminal value and getting a slight delta on ARPUs, but you're not seeing it play out in results. I mean they just posted the largest year-over-year increase in postpaid phone churn of anybody in the industry. So it's not really playing out in terms of some sort of benefit to them, and you're seeing the delivery of the results that we have.

Understood. And maybe switching gears a little bit, but still talking about competition. You could fill in the gaps over here, but I want to talk about satellite. And there are certain satellite operators that sound pretty confident in their ability to compete with broadband against -- not with broadband, but against terrestrial providers in direct to sell. Can you help us understand whether there's a real business case for these site operators and especially as they kind of move towards their next-gen satellite launches and/or close on some of their spectrum deals.

Yes. Well, the short answer is no, there's no effective way for them to compete with a terrestrial network. And they're not limited by money or number of satellites. They're limited by the laws of physics. And here, let me get into that a little bit more, 2 fundamental things that are the problem. One of those being just signal power. So when you're thinking about a terrestrial network, you've got a site that's probably -- we're going to go kilometers now for the second, like a kilometer away from you, obviously, has a lot of spectrum and it's putting out a lot of power, pair that to, for example, a satellite. Remember, these are different constellations in the broadband satellites, a satellite that's 350 kilometers up, which is probably as low as you can go. You go any lower, you're going to have a lot of propellent needed to keep the things stabilized, which means you get in the [ Catch 22 ] cycle of then larger power arrays, which mean more atmospheric drag, which just -- so 350 kilometers is probably the lowest that you can go. So now you're trying to send a spectrum signal. And think about this as like a garden hose, water through a pipe. And you're trying to shoot at 350 kilometers away to a mobile device. It's a lot different than the broadband product, which is shooting to an outdoor large antenna. You're trying to not only shoot a small mobile device, but it's mobile. And the problem is the basic physics of that water through the pipe and through the outlet is what we call open signal loss, right? And this is just the amount of signal loss that you get with no obstruction is not just linear in terms of the distance. It's actually an inverse square. And you cannot put enough power. You can put 100,000 satellites up with all the spectrum of all 3 MNO operators, and you will never be able to project enough power to overcome that open signal loss of 350 kilometers. What that practically means is by the time it down to earth, where the mobile phones are you don't have enough signal to overcome things like roofs, glass, modern vehicles. And so you cannot provide a reliable experience to customers in 70% of the time where they spend is in buildings or in cars, modern vehicles. So you can't service a customer in 70% of the time that they're spending on a mobile device. And again, you can put up all the satellites you want to, the fundamental laws of physics will not allow you to generate a beam signal strong enough to actually overcome obsticles. so that's the first and most basic reason why it cannot be a competitor to terrestrial networks. The second is capacity. And there's a theoretical limit also to the beam size. When we do estimates around 1 of the operators out there who's going to have a V2 D2C or D2D, however you want to think about it, satellite, that beam size is significant. For example, if you think about Manhattan, New York, where [indiscernible] New York a lot, that beam size is roughly multiple, 6 to 15x depending on how you define Manhattan. But multiples the size of Manhattan. And that V2 service is going to be able to provide a decent service outdoor service, see the previous conversation I had to about 10 concurrent users. So think about that. A beam size is larger than the size of Manhattan that can reliably provide service to 10 concurrent outdoor users. It's just not an experience that customers demand. And so for those 2 fundamental reasons, it's just not an overcomable situation. Now that doesn't mean it doesn't have a place. Like we've long said it has a great complementary place in outdoor areas, particularly areas like national parks, where the economics just don't make sense to build the terrestrial network but it can be a great added complementary et for customers. And that's where we see it play out. That's where we see the hotspots are. If I look at the last week of data, D2D traffic on our network is count the 0.002% of traffic. And that's because for the first 2 reasons, I said it's just not where 70% of people are but it does make sense in that place where terrestrial networks just don't make [indiscernible] economic sense. And that was the whole purpose of the JV is how do we make this a win-win for satellite operators, for consumers and make it economically viable for the carriers as well. So how do we stimulate competition in this space, give consumers the coverage in those few limited outdoor areas like national parks where, by the way, by pulling our spectrum together, we can take care of the hotspot needs of [indiscernible] or a Yellowstone National Park, and have it be a great complementary service. That was the whole thought process behind this. But in terms of a competitor to a mobile network operator, it's just not physically possible. And this is also the reason why we said, well, you give them an MVNO, said, well, it doesn't differentiate you. like there's no differentiation. So no, we're not going to give them an MVNO. It doesn't make sense for us to give them an MVNO. The answer is no.

And despite you saying no, we keep getting the same question around, well, they said it this way and they -- and just very clear that there's no interest in the satellite MVNO.

No, no interest.

Okay. Perfect. I think that hopefully is clear enough for folks. With the time that we have a few other topics that I want to make sure we touch on maybe cost savings and AI. You framed the $2.7 billion of cumulative savings target by 2027 and with even more opportunity beyond that. As a CFO, how do you separate gross savings from net savings? And I think what I'm trying to really get at over here is how much of the AI and digital productivity benefits should investors expect to actually flow through versus being reinvested in network, retail, care or growth?

Yes. Well, figures that we gave you, the $2.7 billion in 2027, that's the amount that's actually flowing through to the bottom line and incorporated in that pretty ambitious guidance that we put out there. And like you said, there's way more opportunity even beyond 2027. It's not like $2.7 billion is the ultimate one. But we have conservative adoption assumptions and things to that effect, but that will grow beyond 2030, and we'll see even more gross benefit. But what's even more exciting than what are going to be great, I'm already seeing some of that, some great cost efficiencies and productivity efficiencies come out of that is, what does this mean for both consumers and the TAM and service revenue opportunity for us beyond cost efficiencies. Certainly, you see us leverage the fact that we're so far ahead on network to put early things out there like live translate, right? That's a great AI use case. It's in beta now, that can really provide meaningful [indiscernible] service to both consumers and businesses alike. But the even bigger opportunity that's not in any of our plans from a service revenue, core EBITDA free cash flow perspective. But because we're the network leader from a technology, we're certainly the central ones who can take advantage of this is physical AI. So for the first time, when you think about it, a simple use case is, for example, robotics. And the challenge of robotics manufacturer has is, I really need -- if I have a fleet of robots or drones, I really need space time coherency as [indiscernible] saw I need to be able to know where those things are and when they are, that they're not crashing into each other, they're able to interact, they're able to do the tasks that they need to do. And to accomplish that, I either have to do 1 of 2 things, put a lot of the technology on the device, which means not only sensors, compute but battery drain is a real issue. And that makes the devices really expensive. Or my other alternative is Cool, let me do inferencing in the central cloud, but the problem is the time dynamics, the latency just doesn't solve that problem. So now you're going to have robots and drones crashing into themselves. So this is the first time where networks are going to be in the center of this where you can have inferencing at the edge that allows robotic manufacturers, drone manufacturers. We've got a few of these already working on with other companies to be able to take some of that compute and battery drain, make these things cheaper and put some of that inferencing on the edge because that inherent latency and [ jitter ] and other aspects that are so important are fast enough to be able to do it. And that solves that problem. If I couldn't do it with the central cloud. There's still be plenty of stuff in the central cloud, but you need the central layer of a distributed edge-based inferencing technology that only a network can provide and ours being 2 years ahead of everybody else is in the center of this. So this, to me, is one of the most exciting opportunities of significant TAM expansion for this company, and the industry as a whole later if and when they catch up from a network tech perspective, which they won't. [indiscernible] won't let that happen neither were we, but you've got the ability to connect and being the center of value creation of the next AI revolution, which is physical AI.

And I think -- I'm not going to push you to provide numbers unless you're willing to. But some telecom investors that I think have -- remember 3, 4, 5 years ago, when they've been made -- there have been some promises made of 5G and what different use cases and business cases could be. And that hasn't necessarily evolved in the way that I think some had hoped. But it sounds like these opportunities are more tangible in near term and you're seeing that. Is that a fair characteristic?

Yes. Remember, we've never put those out as we're the company that has to see the opportunity. It has to be real before we promise investors, "Oh, [ MEC ] is going to deliver $1 billion or $10 billion or whatever it was that hasn't said long ago, we put 0 of that in our plan because there really wasn't this connection of -- okay, what is the value proposition to whom? Like I'm just going to take a software application that's living in the cloud and move it on-prem. We didn't see how that was really going to create a lot of value for anybody. And so we didn't put [indiscernible] into our plan. This is fundamentally different because, again, you either have very expensive robotics, drones, all the things that physical AI is going to bring or you can do it with the cloud or you solve it by making all of this cheaper having battery drain not be a problem and having it solved at the edge. So it's not just bringing some sort of software closer. It's actually providing a value that would never have been brought under MEC.

Perfect. All right, 3 minutes and 2 last questions. One on capital allocation, which, to me, seems to be perpetually underappreciated part of your story, I guess my [indiscernible]. For 2026, your -- you've -- your return authorization stepped up to $18.2 billion. The balance sheet remains a relative advantage. When you think about the next dollar of excess capacity, how do you rank buybacks, dividend, spectrum, fiber JVs and other smaller strategic M&A?

It's -- we really use our very consistent capital allocation framework that I know we've talked about a long time, first being to your point, what is the right leverage for us. We want to have flexibility. We have the best balance sheet in the industry that creates a lot of strategic optionality in the future as well. But what's the right leverage for us. We continue to believe it's 2.5x core adjusted EBITDA. That's a growing core adjusted EBITDA. So that in and of itself creates a lot of capacity. And then you go down the waterfall and say, one, I need to make sure I'm investing sufficiently in the business to take this best network, best value, best experience proposition and make sure that it continues to be in a leadership position. We have no desire to fall behind in our network leadership. We want to grow and things like physical AI or something that's tremendously exciting for us, neither so on value or experience. So that's the first place you invest for long run value creation. The second is you look at things like spectrum acquisitions or smart M&A, you see us do it in the JVs very prudently. And then you look at how do I return to shareholders any of the excess. And that's a balanced, dividend and share buyback place. So we keep running it through the same thought process and philosophy.

Ton of excess. Okay. Last 1 for me. when you talk to investors, where do you think the debate today is still too focused on maybe -- we talked earlier about postpaid phone and whether or not that's a right or wrong metric. Where do you think folks should be spending more of their time on in terms of underwriting the T-Mobile story?

I think it's -- and we've touched on this many times already in the conversation. It's the full left to right of don't focus on just 1 KPI but focus on how we're able to translate growth, but the P and the Q plus new opportunities like financial services, T-Mobile for Business, 5G broadband into true tangible value creation in the form of service revenue, core EBITDA and free cash flow margin and how fundamentally differentiated it is to the others in the industry. And again, we have the structural elements to keep that advantage and keep that growth going on way into the future in a very differentiated way. So I think that's the fundamental piece coupled with particularly this concept of where physical AI could take this company. That's not embedded in any of the guidance that we gave you, that's already exciting and what that means potentially for the future of the company.

Fantastic. All right. Peter, thank you so much for your time today.

Thank you so much for having us. .

Yes. Absolutely.

Appreciate it.