T-Mobile US, Inc. (TMUS) Earnings Call Transcript & Summary

Yes. Well, let me do the uplink thing first, John, and then we'll come back to the comparatives on propagation because that seems to be a topical discussion.

It really is. Yes.

I'm up again this week, I think I don't have to -- not to do the tutorial again, right? But the weakest link on wireless transmission is from the mobile device back to the base station because of the power constraints, et cetera, that exist in the mobile device. And so we talk about networks being uplink limited. And if you've got a better source of spectrum to support that uplink with higher propagation, for the same volume of power, you can get further, right? It's just basic physics and math. So for us, one of the things, one of the key reasons why we looked at building out this subtending 5G layer is a tremendous uplink to support mid-band spectrum, like 2.5 and like C-band. And so TDD 2.5 and mid-band in C-band, you can concentrate the vast majority of your spectrum because it's asymmetric, right, you can divert the majority of your resources for downlink transmission. You have to keep some uplink for signaling, et cetera, and some work. But the vast majority of your spectrum gets committed to downlink compared to FDD where it's 50-50. So you can do a lot more on downlink with 100 megahertz of spectrum in a TDD system than you can do in an FDD system. FDD is 50. The TDD system, you could probably do 80 megahertz on the downlink or more. And then you bring in a low-band uplink. And that can extend. I mean, we've done extensive testing on this, John. That can extend your coverage by up to 30%. And so if you can imagine that uplink constraint goes away, you're now better. And what happens is you become downlink constrained on your mid-band spectrum. You can't match the full capability of the low-band transmission, but you become downlink limited. And so that's a tremendous opportunity that we've been working on incredibly hard. We're driving, I believe, we're still driving the industry on this one. I'm surprised that there hasn't been more industry momentum around this. I think back to Verizon on their Analyst Day, they were like carrier agg with low band. What's that? Like my head was exploding listening to that. So it's a huge opportunity. It's coming in a material way this year. We already have devices, capabilities. There's a lot of work ongoing right now. That's one of my big objectives and targets inside '21 is that we can deliver that more broadly across the footprint that we have. And so that's another way for us to, even with all this build and the scale and pace that we have on the build, to really continue to extend the range and the performance of our 5G offering, especially in that mid-band layer. So that's a really exciting area. Our 2 vendors in Ericsson and Nokia are making good progress. We still have a bunch of work to get done, but we're very confident on what can be achieved in 2021. So that's one piece. Then let's come to the fun debate, 2.5 versus C-band. So of course, I'm a C-band fan. We spent the best part of $9-plus billion on C-band in the auction for a 40 megahertz slice in the key markets and areas where we believe it can be a really great supplement and addition to our 2.5 gigahertz layer. And so again, this is an important piece, John, in that this is another layer in our cake almost. So our big geographic mid-band coverage layer is going to be 2.5. And then we can come in with C-band and we can use that where we need to, where the capacity hotspots exist and where the propagation characteristics of the spectrum match the topography. And so we can be selective where we deploy it. Of course, our competition has to deploy C-band kind of to build a contiguous mid-band layer. So that's a very different proposition and challenge. And of course, with C-band versus 2.5, the physics, you can't bend the law of physics, right? We've all been doing this a long time. You know the further you move up the band, the less the propagation of the spectrum, right? We've seen that from low band to AWS to PCS to 2.5 to 3.5. It's obvious somewhat. And so we spent an inordinate amount of time pre-auction testing 3.4, testing the C-band versus the 2.5 gigahertz performance. And we know. I mean, we've gone out. We've tested this with equipment that is state-of-the-art that's being deployed today or will be deployed en masse. Of course, we own and are deploying rapidly a 2.5 gigahertz asset. And what we see is that the C-band spectrum, good spectrum it is, but it doesn't propagate as far as 2.5. With 64 MIMO, the same thing, we're using both. And I think all of us will use 64x64 massive MIMO technology. In C-band, we're using it in 2.5. So common kind of antenna and a framework and opportunity there. And 2.5, I mean, we look at density of the network. If you have to build out a C-band network to match a 2.5 built mid-band network, you're looking at probably up to a 50% increase in density. And the one thing I'll call out, John, just to be clear, because my competition likes to kind of somehow cite this, they point to a section of FCC rules, where the rules talk about the power constraints and limits that are available. And there's a set of rules for C-band. There's a set of rules for 2.5. And in C-band, it's actually set. There are set limits on power in urban and rural areas. And in 2.5, it's actually a curve. And the curve and the position on the curve for your power is dictated by the bandwidth of your transmission, of your antennas. And so I think even this week, the Verizon guys pointed to the fact that, if you're using 90-degree beam width antennas, then C-band can actually support higher power than 2.5 gigahertz. Well, that may be the case. But John, it's completely irrelevant because nobody's been using 90-degree antennas for years. We use 65 and have them for, I don't know how long, as long as I can remember. And of course, with massive MIMO and beam forming, we're using like 50 beam width antennas. And so when you go to a more concentrated beam of transmission with your radio and these massive MIMO antennas and capabilities, you're not using 90 degrees to spray all over God's given earth with, and that's why there's a power limit there, right? They don't want you spreading massive capabilities all over God's given earth. So you concentrate the beams, you use smaller antenna beam widths, and you can support and concentrate higher power there. And if you look at some of the comparisons that are there in the rules, actually, 2.5 gigahertz power limits, they can be multiples of C-band in rural environments, like 4x. So I scratch my head with this whole thing. I mean I defend them ever since. But come on guys, we all know that it's higher banded spectrum. Yes, there are power rules. Don't try and point to elements of rule sets which are completely irrelevant and are historic in terms of where the technology is today. So we're very confident in our position. Don't get me wrong, we spent a lot of money on C-band. We'll be deploying that to supplement on top of a great 2.5 gigahertz layer. But it's hard to believe that you're going to be able to -- with the density and the numbers that have been talked about by our competition, you're going to be able to compete with the footprint and the breadth of coverage that T-Mobile is going to provide on 5G.

Understood.

Sorry for all these long answers.

No. Those are great. Those are great. That's great. You're ticking off a number of things on my list. It's perfect. Before we move off the mid-band, can you talk about any appetite you have for additional 2.5 or maybe your thoughts on the 3.45 later this year or just any sort of spectrum need or just your view on the need for additional spectrum below, call it, 6 gigahertz.

Yes. I mean, obviously, I'd start with the fact that we have a great spectrum position today, John, right? So we're in a much stronger position than our competition. We have more mid- and low-band spectrum combined than AT&T or Verizon. I mean, Verizon closed the gap some with C-band, massive expenditure for them to do so. So we're in a very good spot. And back to the earlier discussion, I'm driving as hard as I possibly can to deploy as much of that spectrum on 5G as soon as I can over the next 18 to 36 months before we can reach our crescendo on 5G mid-band spectrum. That said, as a company, I think we've got a strong history of being very disciplined. We're looking at new spectrum opportunities when they come by. And we'll also be opportunistic. There may be chances for us to look at certain spectrum sources as they become available and supplement our position. But I can't give you any basis on obviously what we'll do or not do. On 3.45 to 3.55, which is the next fairly imminent set of auctions that are coming up on the October time frame, which is just around the corner, I honestly think there's still a lot to learn, John, right? I mean, it's a very different auction from C-band. It's 100 megahertz compared to 280. So it's almost 1/3 of the spectrum resources. No single party can walk away with more than 40 megahertz out of the auction. 10 megahertz carriers in there. So there's a lot of different combinations that could fall out of the auction. But the other piece that we're all looking at and we're waiting on bated breath, to be honest, is a lot more information from the DoD, the incumbent on that spectrum today. And they've highlighted concerns on almost 60 million POPs of the country where there are interference concerns. And a lot more information is going to get shared on that, we hope, as we move through the next, I suppose, 6 to 8 weeks. And that will help us all understand how and where this spectrum is important and the opportunity. So more to come there. We'll watch that one, and we'll learn a lot more, I hope, as we go through the summer. And then hopefully, behind that will follow a 2.5 gigahertz white space auction. There's interest there for us. Again, it's kind of primarily rural 2.5. I mean, that's the construct there. It's pieces of the spectrum that were never passed through. It's not BRS. It's EBS. And so yes, sure, there's always going to be interest from us there. But I think that auction is some time off in the future. Unfortunately, I think that one may be delayed into, certainly, it's going to happen in 2022, you think now.

Got it.

Neville, maybe we can move up to the top layer of the cake. Can you talk about plans to build millimeter wave and maybe an update on your small cell strategy? I believe you have about 70,000. I think you have also decided to maybe decommission some of the Sprint small cells. How should we think about the small cell pacing from here?

Yes. Well, 2 pieces in that, Batya, right? I mean, millimeter wave and then small cell and they're not necessarily one and the same thing, right? So let's dissect them. So I mean, we love our millimeter wave and always have. I mean, we deployed one of the most intense millimeter wave networks in New York. We were first out of the gate almost. But our focus is primarily on using millimeter wave as that top layer of the cake where capacity needs really dictate. So we're deploying millimeter wave in stadiums, key venues, locations where we see that capacity and need being generated. I've never been a believer back to the Verizon story of a coverage layer on millimeter wave outdoor. I think we all know that's incredibly difficult to build. The propagation on millimeter wave is limited. Its indoor penetration is incredibly poor. But Verizon has been throwing out a lot of small cells to try and build something at least on 5G. But I think the coverage and capabilities on that layer, folks have tagged it at, I know 1 million or 2 million people may be covered with that millimeter wave layer. So very, very limited coverage. And so that was never our strategy with the layer cake. It's the foundational low-band. Mid-band carries the lion's share of the traffic and activity and then you use millimeter wave in the areas where you see peak traffic and need, where there's massive density of customers coming together. So that strategy is there for us. Now when I come to small cells, I mean, we have, as we do, we have more cell sites than anybody today. And we have more small cells, I think, than anybody else today as the combined T-Mobile and Sprint. And so I have more and more cells than I need. I mean our target framework is about 50,000, we've said from where we are today. So is there some decommissioning in that space potentially? Yes. But we're targeting kind of 50,000. We're going to see where it goes. My strategy today around here is primarily it's a macro-based network in 5G. So we're going to pour huge amounts of spectrum onto the largest and most dense wireless network in the U.S. I mean nobody can dispute we have the biggest network. I mean that's like barking at the moon. I mean we're well over 110,000 sites today, right? So Verizon's, I suppose, more than half of that, but not much more, right? And we're talking targeting 85,000 sites or so was our target end state, which is much more than both the stated end goals of AT&T and Verizon. So today, we have massive density on that macro layer. I'm going to pour huge volumes of spectrum with radio on to those sites and maximize that coverage and breadth. I'm going to use small cells again where there's intensity of, maybe specific coverage need or it's specific capacity need that we need to peel off. But when you have a wealth of macro capability, your small cell need is reduced. I mean it's just again the physics. And I think both companies prior to our deal, we had good macro networks. We had a very dense network at T-Mobile. And we're all looking at small cells to supplement that as a comparison and alternative to macro build. Deal closes, now I've got more macro sites than I really know what to do with. And I've got more small cells than I really know what I need to do with. And so we're looking to collapse and combine and do that most efficiently to build that density and capacity and performance. So don't get me wrong, love small cells. But again, I have more than I need. And so as a company, we're looking towards that final design. Does that change over a long-term horizon? Maybe. I think that, for me, I see our combination of our sites, our covered POPs and our spectrum resources, and I see us in a leading position for years to come. And the competition, I think, is still very nervous out there talking about density and build. They've gone through massive expenditures on spectrum. They don't want to talk very publicly about massive rollout of low-band 5G or massive densification in their networks. And I get it. I get it, right? You can only swallow so much in an investor community in one go. But that will come. I mean, we will force that to come from AT&T and Verizon because if they don't, I mean, they're basically going to cede network leadership hands down to T-Mobile.

Got it. Maybe putting some CapEx dollars around that network deployment, you have a budget for about $12 billion this year. And then further out, I think you've said that it's about $9 billion to $10 billion going forward. But with potentially costing more to build that incremental 100 million POPs in sort of smaller rural footprint, could we see maybe CapEx stay at elevated levels? Or what could really change that long-term guidance?

No. I mean, we're very confident in the profile that we announced at our Analyst Day. And so the intensity right now is '21 is a very busy year for us. '22 is going to be another hugely busy year. And so by the end of '22, I'm going to have a big, big chunk of this network deployment done, not finished. And so you see that drop off in '23 as the pace and volume of site activity starts to reduce. But that's all based on, obviously, a great 18 months ahead. We were able to secure tremendous capital efficiencies with our vendors and in our deployment. And so I look at those lofty CapEx numbers that Verizon and AT&T have. I know they have fiber and other pieces sitting in their capital lines. But there's a lot of spend there, and it doesn't always look that efficient to me based on the pace that we're moving at. Obviously, I'm trying to hit these sites. I'm trying to minimize multiple visits to sites. I'm trying to deploy in one go and minimize all those zoning and upfront costs as well as the physical construction work that goes on into these sites. But right now, I mean, we are running at an incredible pace. I mean, literally, every week, Batya, hundreds of sites are being upgraded with 600 megahertz and 2.5 gigahertz radio. I have tens of thousands of people running a move across the U.S., deploying hardware every day of the week. And that pace and intensity, as I said, will be sustained fully as we move into '23. But that capital line, even though it starts to dip a little, it's still a very strong investment profile for the company.

Great. Any interest in owning fiber assets as opposed to leasing them?

Not right now. And people ask me this question a lot, right, about are there better kind of owner economics out there on fiber. And I think you have to sit back and look at what's happening in the marketplace today. I mean, on our math, I'd say that the one company of our competitors that has the strongest owner economics is AT&T. I mean, we probably estimate half of their sites, they have their own fiber. Verizon's numbers probably, I mean, Verizon has been selling off fiber businesses, right? We've all watched that. I think their site base, probably maybe 30%. And so the vast majority of like Verizon sites, they use the same model that we do. Why do we use that model? It's because we can secure great rates in the marketplace. And so the economics is a simple equation for us, Batya. We look at, do we want to run fiber? Do we want to light it? We want to take all those costs into our business ourselves. Or can we buy those services cheaper in the marketplace? And right now, I can assure you that math works for us. We're driving multi-gig backhaul. We were, again, this sounds hard to believe, but we were one of the first, if not the first national wireless carrier that drove fiber to the cell site. We started on that journey as we were building out LTE and our competition was still scaling T1s. And so we put in a huge volume. We have some microwave, but the vast majority of our sites are fiber fed. And we've been able to leverage and build a marketplace that isn't just fed by AT&T or Verizon. The cable players are in there, utility companies, independent competitive fiber, both in terms of performance and pricing to the marketplace, especially in metro America. There's a lot of fiber out there, and there are a lot of choices. So that marketplace for us right now is very dynamic and very effective. And I look at our, I'm not going to give you numbers, but I look at our cost per site, and we're in a very comfortable position scaling our network and our backhaul capability for the future with massive multi-gig capability to the site, let alone back into the core of the network. So it's an area of the business that if we decide at some point in time those economics don't work, we would change. But right now, the economics absolutely work for us with the model that we have.

Neville, can you talk about your thoughts on the Open RAN architecture and maybe plans to adopt it at T-Mobile?

Yes. For me, John, I think this is another one of these hot topics that and you know has a lot of kind of pointed opinions on this stuff and I've whacked away over time. I'd say this, sitting a year on from when this debate started, we just have to be patient. I think that's the one statement I would use. The ecosystem in O-RAN is going to take some time to develop. You look at the radio business today and you have 3 major incumbents, right, one of them Chinese. And it's a massive scale game. And it's a complex business. I think people look at point pieces of O-RAN, but there's massive R&D in the space to future-proof your networks. There's a whole debate about who build and develop this stuff and who owns the IP, right? And we see that almost every day, right, you can see battles in that space. And that's just not the radio vendors, then you've got Qualcomm and many others are involved in that piece. And so the industry has kind of settled into a rhythm, you could say, an uncomfortable one with the state of play with 3 majors, but it's delivering. And so for us, when I was putting all our deals together and starting this massive build-out last year, to be quite frank, I mean, O-RAN capability, it couldn't meet or match just the basic needs that we needed to go and deploy with. Those capabilities were simply not there, let alone future-proofing on features and capabilities that need to be delivered '21, '22, '23, et cetera. So it was an easy call for us. But obviously, we did a full view, and we continue to do that. We're engaged in the process. We're supportive on O-RAN capability. But my piece is at the right time, and we're in a position where, obviously, we have great solutions. We're pushing incredibly hard in the radio space today. And if some of my competition or some of the new entrants want to push on O-RAN, that's great, the features and capabilities. I'm pushing, our job are around carrier aggregation to give me real, tangible, valuable benefit potentially and from a network perspective, simplify our network. Most of our competition aren't even talking about that. I'm trying to drive those pieces which I know will deliver and secure material benefit for this company. And O-RAN doesn't necessarily help or support in a meaningfully differential way those opportunities. And I think everybody looks at the O-RAN and the pricing and the capital and how much differentiation is there going to be on ultimately the buy, the purchase and the integration and deployment of these systems with O-RAN. And I think it's a question mark. We just have to see how it develops. But we're there, we're watching. We are all for open standards and capabilities, absolutely. But right now, I think we just have to be patient and let this ecosystem, this new ecosystem that's being talked about, let it mature and it's going to go through a massive journey. When I started deploying radio, John, they were like 10 vendors I could choose from, right? We all saw what happened over the last decade. Now O-RAN will take us on a different cycle, but that ecosystem is going to have to form and there's so many different things that have to be figured out as new entrants come in and try and play and try and compete against scaled competitors and especially one scaled competitor, which is very, very hard to compete with on a global basis.

Got it. And we're running out of time. One last question on fixed wireless access. I mean just how big of an opportunity is this for T-Mobile? What are the maybe the benefits that T-Mobile has? And then what are the impediments to rolling this out and scaling that business up as quickly as possible?

Yes. I mean, I love that opportunity for us, right? And this again goes back to when we were looking at and modeling the capabilities of this combined T-Mobile and Sprint network, John, and this was all sort of work on models and stuff and spreadsheets or whatever. But when you look at that combination of spectrum and site density and 5G efficiency, right, one of the things we never talk about is 5G is a way more efficient bit pipe spectrally than LTE, same as LTE was better than 3G. But you take those things, spectrum, site density and spectral efficiency and you can generate massive capacity on this network. So the math we ran and we still stand behind it, we're seeing we have a road map to is, if you took T-Mobile before the deal and you cast out its future capacity, and then you look to the combined entity and what we can do, the multiplier is more than 10x, it's 14x. So we can deliver this huge enormous 5G factory with the site assets and spectrum and the spectral efficiency of 5G. And so what happens when, and I'll tie a few of the questions and discussion we've had earlier on, when you move into, I don't know, rural town in America. So to Batya's question earlier on, I deploy a 2.5 gigahertz radio. It can support 160 megahertz of bandwidth that I have, but I don't need that all for wireless. I may need a fraction of that for wireless. But my deployment costs are effectively the same thing. And so I have this capital-neutral model where I generate all of this capacity, which is excess to wireless needs, so now we can go and move that towards a home broadband solution. And so the opportunity that we have as we look at that net math, what do we need, what can we generate and what do we need on wireless. There's a huge opportunity space in there, which is unique to us at this point in time. So we're not offering that broadband service everywhere across the nation. But the numbers that we talked about, 7 million to 8 million households that we want to bring into the T-Mobile family as customers, John, that's a very real objective. And we have a path from a network perspective to generate that opportunity at very low incremental cost to the business. When you go into many of these rural parts of the U.S., we all know the story often there's god awful service, no competition, high pricing. And so there's a lot of interest from many areas of the country about the service that we're bringing. And early and initial feedback on the service is still early, but is very, very positive. So I'm excited about that space. We have a lot of capacity we're generating. It's another avenue of great growth and opportunity for the business. And we never got to use cases, but 5G for us, think about some of the things we've talked about today. We can go and we can start to penetrate because of the 5G benefits in capacity, we can go and attack a home broadband market. We can go and attack a small town rural geography, where over 130 million people live in the U.S. Those are real tangible opportunities for us with the powerhouse network that we have. They're traditional business models, right, traditional wireless use cases, but we can run at them harder. Then all of the stuff that comes behind on 5G, wherever it may go, wherever it may be, that's all incremental opportunity for us.

Got it. Well, Neville, as always, thank you very much, been very generous with your time, and we appreciate you participating. Hopefully, we can do it again next year.

All right. John, Batya, thank you. Good to see you.

Thanks, everyone, and thank you all for joining today. That wraps up our conference. Talk to you all soon.