T-Mobile US, Inc. ($TMUS)

Good afternoon, everyone. I'm Sebastiano Petti, and I cover the telecom, cable and satellite space at JPMorgan. I want to welcome T-Mobile's CEO, Srini Gopalan. Srini, thanks for being with us today.

Well, thanks for having me here. And just before we get started, I'd like to draw your attention to our safe harbor statement that I will be making forward-looking statements and using non-GAAP measures. But delighted to be here.

Thank you. Thanks for joining us. And I think it's your first time at the conference. So great. So Srini, since the February capital markets update which reset multiyear guidance to reflect not only recent M&A, but as well as improving underlying trends, T-Mobile followed up with a strong first quarter print. What are the 2 to 3 priorities you're most focused on executing against, I guess, over the guidance horizon?

Yes. Thanks. Look, the heart of what drives our outsized growth, both revenue but also financial and tumbles into industry-leading free cash flow, is differentiation, right? Our NPS is today, 20% higher than that of AT&T and Verizon. And the heart of that is being able to bring together best network, best value and best experience. So what I'm really focused on is widening that differentiation in networks that's going from 5G to 6G that's driving 5G advanced. In terms of protecting our position on best value, that's providing value not just once in 2 years, but on an everyday basis. And on best experience, the heart of our best experience has been our people. Now we're augmenting them with digital and AI. And the reason that differentiation is so central to the way I see the world is all of the outsized growth. And I'll just give you a few examples. We've got 20 million businesses and families who are still using AT&T or Verizon because they think it's the best network. In small markets and rural areas, our market share is 24%. We're in a very, very early innings there. Enterprise, our market share is in the 10% to 15% range, massive opportunity there. Broadband, we talked about at the Capital Markets Day, we have the opportunity to get to close to 20 million customers by 2030. And then there's all the new businesses we're incubating, things like financial services. Also, the possibilities with edge AI and physical AI. So there's a lot of growth ahead of us. And the heart of that growth is stoking differentiation. So what I'm focused on is making sure that we say streets ahead on best network, best value, best experience.

Yes. The network perception, you called out the 20 million network seekers, right, that aren't on your T-Mobile subs. In first quarter, you saw the highest ever share of switchers citing network quality. So beyond the passage of time and marketing and ad spend, what specifically closes the residual perception gap? And help us about maybe realistic time line to when you could possibly fully neutralize that?

That journey is well on its way. And you say with every passing day, we get more recognition for it. Things like the Ookla Speedtest, which is 0.5 billion data points across the country. This is not kind of a drive test where you have 50 people driving around and trying to figure out who the best network is. This is billions of data points from actual customers. You've seen J.D. Power, where after 22 years, we claimed the best network position. So we're getting more and more recognition from it. The thing we're learning that we're now scaling is ultimately, changing network perception is about what the network means for you, right? Because nobody really cares about America's best network beyond a point. Network differentiation happens when you experience it, when our network is demonstratively superior where you work, where you live and where you play. And we're getting to the place now we're able to target people at an individual level and demonstrate to them, network superiority. And that's really what creates this acceleration. I don't look at this as a point of we will arrive at some point, which is the truth. The reality is we will make a lot of share gain along the way. 20 million accounts, families, businesses is a lot of runway. Just think of it this year with our guidance, you're talking 1 million and change accounts. That's a lot of years of growth left for us.

And any update in terms of maybe what you're observing in May or since the earnings call in terms of just promotional intensity across the ecosystem? You are still confident in customer lifetime values being maintained at appropriate levels. But I guess, what's underpinning the confidence that led you to raise your postpaid net add guidance after the first quarter?

I think we spend a lot of time obsessing about promotions and competitive intensity at the margin. Sure, it's important. But if you think of it, right, there is a flow of where the water is going, and then there's kind of perturbations on top. The flow of where the water is going is based on differentiation. When you have 20% higher NPS, when you have best network, best value, best experience, there's a natural demand pull that's heading in our direction. That, at its core, is what gets us comfortable with raising postpaid guide. You said something really important there, though, because our true north on how many nets we do in terms of accounts in any quarter gets driven by our view of CLV. And you saw first quarter, 6% higher accounts than last year. We didn't talk about this at great length, but also better value. So our port-in ARPA was 20% higher than our port-out ARPA. And so when you look at competitive intensity, yes, sure, December-Jan was more intense. But as we work through the first quarter, we saw value in going after more accounts because the CLVs were strong. We were looking at our port-in ARPA versus our port-out ARPA. All of that filled us with confidence that we were booking stuff that had really strong CLVs. And quarter 2 is going really well.

Great. And then you touched on SMRA, share hit 24% in the first quarter. So this is, I think, your 12th consecutive quarter of win share leadership, if I'm not mistaken. So just beyond word of mouth, I mean, what's resonating in these markets? Is it distribution efforts, brand? And I guess, I mean, what's the long-term ceiling if you'd have to frame it for investors?

Can I start with the second piece?

Yes sir.

Because this theory of a long-term ceiling kind of assumes no differentiation. There's this view of the world that if there are 3 players, everyone sits at 33% share. That assumes all 3 are the same. The reality is when you've got best network, best value, best experience, all 3 are not the same. So you look at the city of New York, we have about 50% market share there. That's what happens when you bring the 3 things together. Not suggesting that SMRA is going to get a 50% share, but we're at a very, very early innings in SMRA. In terms of what we do differently in SMRA, what we're seeing is an underappreciated thing in wireless is the network effect. And that's just a bad pun. But what I mean by that is when you get to something like 24%, 25% share, you get more and more word of mouth. And that creates its own upward spiral where you're seeing more and more people talking to each other about the fact that the network is distinctly superior. And what we've done really well in SMRA is combine the network expansion with the distribution expansion and localized marketing, things like Friday Night 5G Lights, where we go into a small town, there's a $1 million award for fixing the football ground. And last year, we had hundreds of thousands of small towns competing for this. Small town called Dierks in Arkansas with a population of less than 2,000 got 2 million votes, right? And that starts giving you a really local presence and local word of mouth.

Interesting. I was not aware of that program. Pretty cool.

It's actually in Dierks, Arkansas with Gronk. That was quite an experience.

Oh, yeah? Yes, I have to pull that up. So you've always anchored on the best network, which we just have been touching on, at the best value. But should we read that the 2026 guide as -- from an ARPA perspective, I apologize. Should we assume that the 2026 ARPA guide embeds a rate action at some point this year? And I guess more broadly, what's the composition of the 2.5% to 3% ARPA growth? And how much is pricing versus mix in terms of features?

So I'm not going to break up the 2.5% to 3%. Let me give you directionally how we think about this thing. We guard our value leadership zealously. Now we're -- we've strategically architected our portfolio so that we have something that's quite rare in the telecoms business. We have a front book that's higher priced than our back book, which means our existing customers actually pay less than our new customers. That just means where -- even volume growth is ARPA accretive. And the way we think about the sources of ARPA growth is we've got 3 distinct sources of ARPA growth. Number one, this front book/back book dynamic, which is even as we grow volume, we actually grow ARPA. Number two, expanding the nature of the relationship with our customers, which is kind of why this 20% higher NPS at 45 makes a big difference because people don't buy more things from you unless they're incredibly satisfied with the things that they've already bought. And the last, and traditionally, our smallest has been thoughtful more-for-more rate optimizations. And that's the formula you should expect to play out not just over this year, but over the years of our thinking.

And then shifting gears to the broadband business and FWA. T-Mobile was again the fastest-growing ISP in the first quarter with 50,000 -- 500,000-plus broadband adds and with FWA reaccelerating year-over-year. I guess, what's driving that inflection? And how is the competitive backdrop evolving now that Verizon appears to be deemphasizing FWA a bit and AT&T kind of scaling their efforts?

Ultimately, it's really simple. It's an incredible product. So you look at this business, right? Over the last 2 years, our number of customers has doubled. Usage per customer has gone up 25%, and our median download speeds have gone up 50%. And they're 50% higher than our nearest competitor. So what you have is an incredibly easy-to-use product. In fact, on some of our newest generation routers on WiFi, right? This is not when you plug in fiber Ethernet into your computer or whatever. On WiFi, we're getting 400 Mbps download, which is why our NPS is actually higher than fiber. So it's just an incredible product powered by an ultra capacity network. That's what's been driving a lot of our volume.

And then I guess a persistent bear case that we continue to hear about less so, but is that FWA can't scale sustainably, just given data usage growth curves. I guess, what leading indicators give you confidence that capacity stays ahead of demand? And I guess, what are the release valves if utilization tightens in specific markets?

So let's talk about -- we talked about 15 million customers by the end of 2030. How did we get to that math? Here's what we do. We divide the country into 36 million hex bins. These are small geographic areas typically covered by a sector or even less than a sector. For each of these areas, we look at the growth in wireless usage based on historic data and then take a conservative view going forward. And we focus that growth entirely between 7 to 9 at night because that's our busy hour. That's what matters as a binding constraint. We take that and block that capacity off. That is only for wireless. Everything else in the network is what we call fallow capacity. The way we got to 15 million was we took that fallow capacity and then capped market shares. We said, "What's a reasonable market share?" Because purely because I can supply, it doesn't mean there will be demand. So we capped that market share. And that gets us to 15 million, assuming no incremental capacity -- no incremental spectrum purchases, assuming no incremental spectral efficiency, right? That's how we get to 15 million, which is why I'm super confident of our ability to deliver on that 15 million because we will end up seeing more spectrum, and we will use the fallow part of that spectrum. We will see spectral efficiency gains. We're already seeing massive spectral efficiency gains with 5G advanced.

Right. That's before we even think about [ fixed income ] with 6G?

Yes.

Got it. So shifting to the fiber JVs. You announced 2 new ones ahead of the 1Q call, adding, call it, roughly 1.8 million passings, I think, with Oak Hill and Wren House. I guess, what made these the right partners? And what's the decision framework going forward as you think about JVs, wholesale relationships and maybe even deploying your own balance sheet for fiber M&A?

So firstly, the way we think about fiber and broadband as a whole is not kind of some defensive rush to scale because we think this myth of convergence is going to suddenly become relevant. It hasn't in the last 10 years. It's built into our run rate. People can talk about hypotheses on how convergence changes customer behavior. You look at the last 10 years, it simply hasn't happened. And so our approach to fiber and fiber JVs has been to look at places where we can make double-digit IRRs, where we can scale fiber, those tend to be first-to-fiber areas. And we do them as JVs because our partners bring incredibly valuable local experience in things like zoning, permitting, digging, et cetera. And that will continue to be our approach to rolling this out. And when you look back -- when you look at our overall plans of 18 million, 19 million, that number will change with these new JVs up until 2030. It's important to remember that FWA is 3/4 of that, right? And that's the -- fiber plays an important role in our portfolio, but it's a role where we see -- where we look for places where fiber can genuinely create equity value and then double down on those with partnerships.

And how have your wholesale partnerships scaled or your partners scaled relative to initial expectations? And is there a preference emerging for JVs over wholesale opportunities?

So look, if I talk about fiber as a whole, it's obviously very early days. But we're tracking well on our penetrations. We're getting 20% first year penetration. Now that's pretty incredible given that the vast majority of this is a greenfield fiber rollout. I'm not converting existing copper customers into fiber, right? That penetration is really strong. And of course, that means I get incremental value rather than ARPU substitution because I'm not shutting down a copper plant to roll out fiber. In terms of preference, we like the JVs, and we will look at places where we get accretive JVs. Wholesale, depending on the structure of the deal, is a good fallback to that. And we tend to be on fiber, very focused on value creation rather than kind of this is the scale we need to get to because the homes passed number is a little irrelevant.

Well, it's a great segue to my next question. You've been clear that cable isn't a priority and that the focus is on taking share from incumbents. Is there a scenario though, whether it be valuation, asset quality, market structure, that would change that calculus? Or is cable structurally off the table?

Look, I think of this stuff strategically first rather than valuation backwards. Cable is off the table from our perspective, right? The reason for that is kind of threefold. Number one, we're a growth company. Our DNA is the Un-carrier. Our DNA is changing the industry for the good of the customer. We don't play defensive. We don't like being in a place where we're the incumbent. We like being in a place where we're attacking the incumbent. Second piece is just front book, back book ARPUs, where cable's back book is so overpriced that you're spending your entire time chasing your own tail, right? Because then, you compete on the front book and you suddenly see spin down. Even if you get to positive net adds, you're in a place where value leakage happens. And third, you look at the speed at which wireless technology is developing, especially as you look at FWA rolling out in countries like India. Suddenly, the ecosystem becomes really exciting because you're getting CPEs cheaper. You're getting the ability to optimize your technology a lot more. And that combination -- and you'll compare that with where DOCSIS is going. That combination is what takes us to a place of no to cable.

And just a quick follow-up on some of the fiber. You talked about getting to 20% year 1 penetration, which is great. But from a build perspective, I mean, how is that -- is that ramping on schedule? Any early learnings in the markets?

So we're seeing good performance on using our distribution. We're seeing good performance on using the T brand, and it's very early doors on fiber.

Got it. Okay. And then shifting gears. At the CMD in February, you framed a $2.7 billion cumulative cost savings target by year-end '27. And I think Peter touched on the call, weighted maybe a little bit towards later stages of 2026 and beyond. Maybe help us think about from an initiative perspective, what's already in flight versus still to come? And maybe what are the largest buckets of savings opportunities?

Yes. Just one clarification. That was $2.7 billion by '27. That doesn't mean that's going to be the run rate going forward. This is obviously an accelerating run rate, so you should expect even more going forward. In terms of places where we're seeing joy on this broader bucket of digital and AI, I think firstly, it comes from a philosophy of we don't use digital and AI as shaving cost at the margin. We start by saying, what can it do to fundamentally change our experience? What can it do? Because very few customers actually walk into stores or call because they just want to chat to us. They walk into a store or call because they have a problem. What digital and AI, especially with 24 million monthly actives on T-Life, enables us to do is proactively address that problem before they call. Or when they call, be able to deal with it quickly, efficiently, often through our voice bot. And we made a promise at the last CMD of a 75% reduction in our call volume. We're well on track to that. We're getting something like 60% containment with our InstantCX (sic) [ IntentCX ] chatbot. So one big source of savings, and more importantly, the next step in the customer experience is everything we're doing in care and retail. The second is the power of this taken to our software development, whether that's in network or IT, where we're seeing just incredible power in having some of our brightest engineers work with 20 agents, agentic interfaces versus actually having them outsource a bunch of stuff. So we're seeing some real movement there. And thirdly, on the network itself. I mean, when we had Winter Storm Fern, that was the first scale use of AI to actually begin to do our antenna tilt remotely, right, instead of having someone out there in the middle of a winter storm changing your antenna tilt. You're also beginning to see real innovation with things like live translate. So I'd say we're well on track on our -- not just our savings, but using this technology and being at the forefront of this technology to change the experience.

Well, talking about great -- changing the experience, you recently announced an agreement in principle to form a joint venture with AT&T and Verizon aimed at eliminating wireless dead zones by pooling your spectrum and making it available to LEO satellite providers. So I think the stated goals are enabling competition, standardization, helping amongst a fragmented market. But given that a definitive agreement is yet to be reached, maybe help walk us through the strategic rationale? I mean, why does the JV -- what does it accomplish that T-Mobile couldn't achieve on its own through your existing partnership? And does it -- how do you think about the playing field or level of differentiation going forward?

So there's -- this is very much in keeping with the Un-carrier ethos, right? What we've historically done is done two things very well. One, got out there and innovated to solve a genuine customer problem. Second, once we've had experience with that innovation, sat back and looked at it and said, "If we future cast this and look around corners, where does this go? And therefore, how do we want to play?" So that's the same story that's playing out with T-Satellite. We worked really closely with SpaceX Starlink. We have a great partnership. We've now extended some of that to broadband to build a new category, which is direct to sell. And we've had lots of experience of how that category is playing out. Long story short, it's clear that this is going to be a fundamentally complementary category. And just to give you an example, we looked at our data in May, and satellite usage is 0.0002% of our total network usage. That's 3 zeroes, right? So it's clearly a complementary use case. And we are seeing it largely focused on the national parks, that sort of territory. So then you look back and say, clear, we've built this product. It provides value. It does help in dead zones, and it's going to be a complementary product. How do we see this industry playing out? You're seeing more and more that there's going to be probably 3 players in space that -- when I future cast this industry, I look at it and go, most consumer wireless, especially the premium offers, are going to land up having satellite connectivity linked to it, right? We've also seen very, very little take-up of a la carte connectivity, right? Pretty much, no one buys satellite stand-alone. They buy it as part of the premium package which gives you a bunch of other benefits, global roaming, ad-free Netflix, et cetera, et cetera. And I look at it and go, this is going to become a standard part of most premium offerings from all players, and there'll be enough supply of it. And this is no longer going to be a source of differentiation. However, there are real problems to be solved. Problems like device ecosystem. Right? Some devices support it, some don't. There isn't a uniform device ecosystem. Problems like IP, right, and standardized interfaces and standardized ways in which consumers can get access to the service. Problems like standardized spectrum. Now you may ask the question of if it's 0.0002%, why do you need spectrum, right? The reality is in hotspots like in the national parks, you do need spectrum. And because of the distribution of spectrum, I might own the right spectrum outside Zion National and Verizon might have it over Yellowstone, right, which doesn't help the customer. And bringing it together creates a uniform spectrum world where we can pool in that spectrum. So all of that standardization plus aggregation of demand, which for the smaller satellite players is critical or for the recent entrants is critical, creates a satellite ecosystem that allows the American customer to get wireless plus satellite as part of a standardized package. Now we will compete, obviously, as providers in terms of how we innovate around that. But as we look around corners, the differentiation doesn't exist, and this is about creating an efficient wholesale infrastructure.

And how do you think about the risk that a D2D provider eventually goes direct-to-consumer, whether it's through a build-out or an MVNO? And I think you've made your views on how an MVNO with a D2D provider may or may not play out.

Let's start with the MVNO first. I think I've been really clear. That's not something we're interested in. We have criteria around how we pick an MVNO. It needs to be an incremental TAM, and we don't see how this gives us an incremental TAM. We deal with the D2D, D2C. I think there's a real danger here that is we get lost talking technology. The real question is, what is the meaningful gap from a customer perspective in this market? Is there a significant large meaningful gap that's not being addressed, right? I think as we bring satellite and wireless together to address dead zones, I think that's something that does address a customer need. But if you think about in the non-dead zone areas, I'd almost frame the question to you, what do you think is the gap that D2D or D2C would address that's not being addressed by terrestrial? And then we can have an informed conversation on technology. But starting technology first, could there be a D2C product? Yes. But look at the take-up of a la carte satellite, right? So I'm more intrigued in what's the problem to solve.

Right. That's fair. And so now shifting to the network. On the Figure AI partnership, what differentiated capabilities does T-Mobile's network or platform bring to physical AI versus peers? And I guess, where do you see the earliest commercial scale use cases playing out?

Yes. Look, I think the commercial scale, I think 6G, we will start the journey somewhere -- at least T-Mobile will because we're well ahead on 5G. We will start the journey, I'd say, '28. And for it to be at scale, you're looking at about '30, 2030. Why we're excited by Figure AI is what our technology does in the wireless world -- and we're already seeing some of this with 5G advanced -- is for physical AI to work, you need kind of 3 things: low latency. Because otherwise, it's really hard. The robots will be banging into each other on the factory floor, which is not a pleasant thought, right? The second piece it does bring is voice built in. So as physical AI scales, the need to interact with robots, humanoids, drone delivery, et cetera, et cetera, will need voice built into the core network. And thirdly, you'd need the drones or the humanoids to communicate with each other, which is only possible with a low latency network. And you'd need this thing we refer to as space time coherence, which is you need to be able to know where any device is at an exact point in time. You need to know both the space and the time that you're communicating with it, right? So you think of physical AI more broadly, it's only going to be possible with a low-latency wireless network. And today, there's only one player across America who has 5G advanced across the country, and that's us. That sets us up in a beautiful place to get to 6G and all of the benefits of physical AI.

So with C-band not yet fully deployed, and I think you still have some mid-band refarming ongoing, how is T-Mobile positioning for the upper C-band auction that's expected in 2027? And I guess, how do you plan on -- just thinking about the further pipeline of 800 megahertz coming to market longer term?

I think the pipeline is really exciting. And I think Chairman Carr and the FCC have done a phenomenal job in releasing trapped spectrum across the board and also bringing more spectrum to the market. And that will enable true American leadership in wireless technology. On the specific question of spectrum and auctions, I think there's kind of 3 principles that guide us. Number one, we will maintain spectrum leadership. Number two, we will be 6G leaders. And number three, we'll be extremely thoughtful on how we value spectrum. The way we've always valued spectrum, given that coverage is no longer that relevant, is really look at a build versus buy, the cost of densification versus the cost of buying that spectrum. That's how we will continue valuing spectrum. And when we value spectrum like that, sometimes we look at some spectrum and say we're going to walk away from that, like the EchoStar spectrum. But you should expect us to absolutely be present. And those 3 principles will guide what we buy and how much.

So -- well, that makes a lot of sense, I guess. And we'll see once we have, I guess, visibility into how much upper C-band is coming to market, but it seems like plenty of opportunities and great balance sheet relative to peers to kind of take advantage of some of that. So we'll -- definitely something to watch. So on the 1Q call or just ahead of that, you took -- the Board announced that you took 2026 total stockholder returns authorization up by $3.6 billion now to "as much as $18.2 billion." So as you think about the pullback in shares and the management team's view of intrinsic value, which you touched on, how are you weighing buyback versus -- the buyback pace versus other use cases? And what would shift that priority?

We've had a very consistent and successful algorithm to capital allocation. We start with leverage. We assess what the right leverage is. We're still of the view that 2.5 is the right leverage. Then you look at everything that you need to do from an organic business perspective. And you look at things like M&A and spectrum. And our focus areas on M&A and spectrum have been really clear. M&A is focused on double-digit IRR opportunities in fiber. Spectrum, we've talked about what's coming. We've talked about also kind of specific spectrum deals that we've done. And then that tumbles through to shareholder return. And that's exactly the way we're consistently going to look at it. And that formula hasn't changed, and we don't see why it should change.

So now I would be remiss not to ask about some headlines. So I guess, in regards to the potential -- so there was the Bloomberg article or whomever that cited Deutsche Telekom's interest in pursuing a potential combination. Your comments from the call aside, but I mean, any feedback or maybe what you're hearing from your largest shareholders and what they're telling you about a potential combination? And I guess, more broadly, we've talked about fiber M&A, but I guess, what's the appetite for "transformational M&A" more broadly?

When most people -- I'll answer the second part and then come to the first. When most people ask me about transformation M&A, that's normally code for are you buying cable. And I think I've been really clear on that. So I won't go back over that ground, right? On your question on the Bloomberg article, I'll say what Peter, myself, Tim have said over the last few days, right, which is we won't comment on speculation. And nor is there anything to comment about, really. On the back of the article, we did get a bunch of inbounds on governance. And just to clarify, our governance in -- consistent with Delaware law, any such hypothetical transaction would need the majority of minorities or the majority of what's called the disinterested shareholders. And both DT and our Board of Directors have confirmed this. So that's really all that there is to say.

Got it. Understood. And I guess maybe going back, so talking about maintaining -- being a leader in 6G. Is that -- help us think about the development time line there. Does it require a big new CapEx cycle? Is it more software virtualization-driven just layered on top of the 5G advanced infrastructure?

Look, there's puts and takes. But the answer -- the best answer we have so far is there's no fundamental change in capital intensity. And the puts and takes are fairly simple. You'll need an new air interface. No indication that, that's going to be step function different. You'll end up having some AI compute because this will be the first wireless network that not only processes bits and bytes, but also processes tokens. But the idea would be to look at fallow compute or to build a 6G network which has the same cost structure or an AI-enabled network that has the same cost structure as a 5G network. And I think there's a lot of indication that a lot more innovation will happen on the core, which will balance out some of it. So on balance, I don't see any change in the direction of capital intensity.

Great. Well, Srini, thank you for the time, and thanks for joining us.

Pleasure. Thank you. Thank you so much.

Thanks, everybody.

Thanks.