AT&T Inc. (T) Earnings Call Transcript & Summary

Good morning. My name is Colby Synesael. I'm the telecom services analyst here at Cowen. Welcome to day 3 of the Cowen TMT Conference. For this fireside chat, we have Igal Elbaz, who's the AT&T -- excuse me, the SVP of Wireless Technology and Experience Delivery at AT&T. This fireside chat is structured for 30 minutes. I'm going to go through my questions. There is an opportunity for those who are listening to ask their questions at the bottom of your screen, and that will show up on another screen that I have up, and I'd be happy to go through those. But before we get into the conversation, I think Igal has a disclosure that he'd like to first bring up.

Yes. First of all, good morning, Colby, and thank you for having me. Before we get started, let me try to share this. I want to call your attention to our safe harbor statement. And what it says is that part of what I'm going to share today might be forward-looking statement, and as a result, they are subject to risks and uncertainties. If you have any additional information that you're looking for, you can go to our website, at the SEC, at the Investor Relations site. And also, this is the quiet period for Auction 105, CBRS. So unfortunately, I will not be able to answer any question on that topic.

Okay. Great. Well, thank you. I guess as I think investors don't get as consistent of an opportunity to hear from someone like yourself, can you first, I guess, start by telling us what your role and responsibilities at AT&T are? And as part of that, where are you spending most of your time these days?

Sure. So I'm the Senior Vice President for Wireless Technology. My team is responsible for the architecture, design and technology road map of our wireless network. So think anything from LTE, FirstNet, 5G, our architecture, our technology certification, the RAN, the mobility core, the devices, anything that has to do with our wireless network and new technologies that we're introducing is with my team responsibilities.

Great. Well, obviously, a very, very important role. Want just to start with COVID-19. I guess, what has AT&T's COVID-19 experience been? And how have you reacted to that? And I guess we're going to go through it in a little bit more, but I want to talk more about macro versus small cells and CapEx. But just broadly speaking, maybe just to kind of get us an intro.

Yes. So first of all, let me just share with everyone our pride in the AT&T network and how it's performed before and during a pandemic. Connectivity during epidemic is essential now more than ever. So thinking about our first responders, our health care professionals. We all work from home, our kids are going to school from home. And during all of that time, the network performed very well. It's a result of last few years of investment in fiber, in spectrum, in FirstNet build, in 5G and software-defined networking. So we're definitely showing the results and the outcome of this before and during COVID-19. Obviously, from a macro perspective, we are still evaluating the near- and long-term impact of COVID-19 on the economy, on our business. We did withdraw the guidance for the remaining of the year. But we remain, and we keep reiterated, that we will continue to invest in the critical growth areas of the company, such as 5G, fiber and broadband and HBO Max, which, by the way, was a very exciting day for us yesterday with the launch of HBO Max. A lot of people between WarnerMedia and AT&T worked very hard to get us to that day, and that was an awesome day for us yesterday.

Yes. My wife and I have already started using it last night.

That's good.

So you mentioned investing in 5G. And I guess to that point, I mean, how much of the network investment right now is being spent on the macro sites versus small cells? I mean, where are you directing a lot of your focus these days?

It's both. Maybe before this, I do want to complete the COVID-19. I want to talk a little bit about what have we seen from a statistics perspective on the network. So first of all, we're well practiced dealing with disasters, think hurricanes and floods and fires. Obviously, they have different impact on the network, but in terms of the processes and how we react and the region network capacity planning that we have in place are very well suited to any type of disasters. We've seen about 20% growth on the aggregate traffic on our core backbone. And on the wireless side, we've seen a shift of traffic pattern between -- moving from urban to suburban. Clearly, we all now, in the morning, rather than driving into the office and being in the office all day long, you're working from home. So obviously, that manifest itself differently on the network. We've seen uptick of about 33% in voice call, 65% in messaging. Video remains the main traffic through the network, around 50% with about 4% growth. We've seen about 400% growth in audio and web conferencing. And I can tell you, again, the network performed very well over the last 5 years. To your question, we invested over $125 billion in the network. So that is capital investment, wireless and spectrum acquisition, the operation that comes with that. In fact, I believe this is higher than any other public U.S. companies in terms of investments. So the -- and that build and that momentum continue into 5G. And our build is continuing both on the macro and on the small cell. If you think, even from 5G, there's different layers that we're introducing. We have the coverage, which is what we called 5G, is the sub-6 that typically on the macro. We are already -- at the end of April, we were already covering 120 million box in 190 markets. And at the same time, we continue to build our 5G+, our millimeter wave build, which is mainly in small cells. And we're already in 35 cities. So the momentum and the build...

How many small cells roughly do you have at this point?

We don't share exactly the numbers and the build. But again, you should expect us continuing building both macro and small cells. And again, from the nature of what we're doing in 5G, we're doing both. And we also completed 80% of our FirstNet build, which is, interestingly, it's close to 1 year ahead of our schedule, we were able to complete 80% of this build.

Great. And then management seemed to suggest on your last earnings call that you guys would probably reduce your CapEx spend in 2020. Essentially, it would be lower than previously expected. Where would those cuts be? Where are the opportunities to delay projects that otherwise would have happened?

I think it's mainly around getting some priorities in the company, as I said, in the critical growth areas of the company, which are 5G, fiber and HBO Max, we will continue to invest, and we shouldn't see any impact in that build.

Okay. And then FirstNet, you touched on this in being at that 80% mark. But can you update us on the FirstNet -- on FirstNet and how it's performing from a network perspective through all of this?

Yes, absolutely, Colby. So first of all, we are honored and privileged to be able to serve our first responders and health care professional during a pandemic and a crisis. As I mentioned, we've surpassed 80% of the build, both on rural and nonrural square mile and coverage, which is about 10 months of -- ahead of the schedule, that the Band 14, this is the spectrum that we got. So very aggressive build. And one of the advantage -- which we shared a couple of time in the past, one of the advantage of that build of FirstNet, that it allows us to be very efficient in our network build through a process that we call ONETOUCH. And what does mean -- what does that mean? Is that every time we climb a tower to deploy the FirstNet spectrum or network, we are doing additional 3 things. One is that we continue to deploy the spectrum that we have. So think of AWS, WCS holding, we just put it to work. We're aligning this up and adding capacity to the network. So that's one thing. The second thing that we're doing is we keep modernizing the network. So think about all of the capabilities from a spectral efficiency that comes with 4x4 MIMO, 256 QAM, the ability to do more carrier aggregation. So all of these done in the same climb. And the third one that we've done over the last 2, 3 years, that every time we touch the network or any piece of hardware, it's already 5G-ready. So when our vendors were ready with their software, for us, it was only a software push because all of the hardware was already in place. From a business perspective, FirstNet, we already have more than 12,000 agencies with FirstNet. We have 1.3 million connections, 1.2 million of them are voice and phones. From what we're seeing on speed test, the way we are analyzing the data out of speed test, FirstNet is actually the fastest network. It's faster than any other commercial network in the U.S. So we are pretty satisfied with the outcome. There's different offers that we offer during COVID-19. There's 50% on new rates for individual first responders. The law enforcement and fire and EMS agencies could get free device for life. So again, it is -- from a network quality and performance perspective, FirstNet is a critical part of our net.

Got it. I wanted to shift a little bit and talk about 5G. So I guess, first off, I don't want to talk about CBRS specifically, but generically speaking, there's multiple, what I'd refer to, as shared spectrum bands that are intended to come to market. Generally speaking, what is AT&T's view on shared spectrum and it's -- where it fits? Or where it will be of value in the marketplace?

Yes. So maybe before this, we are extremely happy with the spectrum holding that we have, the low-band, mid-band with AWS, PCS, WCS. And we've been active in the last 2 auctions on millimeter wave band of the 24 GHz and 37 to 39. So we have what we need in order to continue build 5G. And that comes with a lot of spectral efficiencies improvement. I talked about some of the capabilities for deploying, but there's additional spectral efficiency that come with 5G specifically when everything's going to be 5G only and has new spectrum and why the channel's going to show up, going to be able to do things like massive MIMO and build forming. To your question about shared spectrum. So like any other things in the wireless system, shared spectrum has its own place in wireless spectrum, mainly if it's coupled or bundled with licensed spectrum. So today, for an example, we use 5 GHz as part of LAA. That helps us to boost performance and capacity in dense areas. But again, one of the advantage of this is that we are able to couple this with licensed spectrum, which allows us to provide the level of resiliency and reliability that we need in the network. There are use cases that shared spectrum is good for. My personal opinion is that wherever it's possible, it's always better to have licensed spectrum and dedicated spectrum because that allows us to give you -- to provide our customer the level of consistency and reliability that is needed. And as we move forward and the FCC is providing and putting out there more shared spectrum, we're going to look at each one of them individually to see what's the benefit to our network, and we're going to take a decision on a case-by-case basis.

And I guess talking about your robust network and the spectrum which you already have, the company has indicated interest in supporting cable MVNOs. Can you give us an update on the company's current thinking? I feel like I haven't heard as much about that more recently.

Yes. Colby, obviously, I can't share any details about any MVNO deals. But now that we have established that we have best network, fastest network, we have the spectrum, we have the capacity, that indicates that we can have now capabilities and we can have discussion, build discussions, where it makes strategic and economic sense for us because again, we have the capacity to allow this to happen.

So the simple answer, though, would be that you guys remain interested in potentially those types of deals?

Yes.

Okay. And then just talking about spectrum, I mean, can you give us a brief overview of dynamic spectrum sharing? One of your competitors talks a great deal about that these days. And what its potential role in 5G adoption could be.

Yes. So Colby, actually, we were one -- we were among the first to present and champion the concept that now is called dynamic spectrum sharing. In fact, our engineers are very well -- understand that capability very well. And I can tell you that in my view, dynamic spectrum sharing is a very important technology and capability to allow us and other carriers, by the way, to scale 5G. In the basic of this, maybe for the benefit of the audience, what dynamic spectrum sharing allows us to do is to allow both 5G and LTE to operate on the same channel. So the network is dynamically allocating resources, depends if the device is 5G or LTE. What this allows us to do, when you move from one generation to another and you need to move on top of your existing spectrum, typically, what you have to do is to take portion of your spectrum, let's assume one channel. So let's assume you have 4 channels or 4 lines in your world, okay? What you had to do before is to take one channel, clean it up, from the calls or from 4G or LTE and give it to -- make it available for 5G. And what that does is that you take all of your traffic on 4G and now you put it on 3 lanes instead of 4 and you allow a free lane for 5G, which doesn't have a lot of traffic because it takes time until enough people have devices. So that's not an efficient way to manage your spectrum and the transition. Dynamic spectrum sharing just allows us not to turn off any of the channels, but continue to have the same lines and the same channel serving both. And you start with line where you allow both 5G and 4G. And over time, as more devices are coming, you can start doing this on all of the channels. So this was a long answer just to say: One, it's a very important capability; two, you should expect us to have DSS introduced in our network in the second half of 2020.

Yes. I mean, from a finance guy's perspective, so said differently, not an engineer's background, it seems pretty profound. I mean, it seems like it really will allow those who enable it to be much more efficient with the use of their spectrum. And as a result, whereas there's been this continued view that we need more and more spectrum, and I'm not saying that we won't, the magnitude of growth to which we need more spectrum could be flattened a little bit, I would think, just by using what you have in a more efficient way.

Totally agree with you, Colby. It is a profound technology and really allows us to do that transition very efficiently.

Okay. I wanted to talk about devices. You talked about how part of this is getting more 5G devices into the marketplace. When do you expect 5G devices to be available in a robust way, if you will, on AT&T's network? And I guess as part of COVID-19, are we seeing any comply -- supply constraints that could slow that down?

Yes. So as you and the audience knows, we launched 5G. We were the first to launch standard-based mobile 5G at the end of 2018. So since then, we have devices clearly, and that continued. Just recently, we introduced the Galaxy S20 in March that shows up in our stores, and then the LG V60 came out in April, so -- and those devices are capable to run both on the sub-6 and millimeter wave 5G spectrum. And those will continue to show up. And you should expect about 15 devices by the end of the year on our network.

And that's -- those are devices that would mean, I guess, 3GPP's definition of 5G that...

Yes, absolutely. Everything is based on the standard. There's no doubt about this. In terms of any supply disruption due the COVID-19. We continue to monitor and closely monitor the impact on our suppliers, but I can tell you that we haven't seen any significant, or at all, any impact. We are -- for the last few years, we worked with our suppliers on geographically diverse supply chain, and we were working with them on this for the last several years anyways.

Okay. And then another topic I wanted to shift to is fixed wireless. Can you remind us of AT&T's fixed wireless strategy and how you guys think about it, both from a technology perspective, but also obviously equally important from an economic perspective in terms of its viability?

Yes, sure. So first of all, we know that fixed wireless works. I mentioned that we launched at the end of '18. All of the trials that we've done in the couple of years before our launch were all on fixed wireless. So we know the technology in work. And the question is, as you said, typically in the economics. We already, as part of our CAF-II, our Connected American Fund initiatives, by the end of this year, we already going to be able to offer fixed wireless to 1.1 million consumer and small, medium businesses. We already covered about 880,000. But in general, moving forward, the way we think about our network build is always mobility first. This is how I think, this is how we think about wireless system. It's all about mobility first. You talked -- you asked before about small cell. A lot of our small cell build are in dense areas, in urban areas. And they are there to kind of provide capacity relief, for capacity offload. Now on top of that build, this is where we build millimeter wave, right? Millimeter wave is that capability that allows you to get into 1- and 2-gig speed. And again, in the heart of this, it's always about mobility experience and capacity relief. Now -- but because the infrastructure is there and because we're deploying millimeter wave on top of this, there's -- it always makes sense in terms of, okay, can we leverage the same technology to provide fixed wireless services? And then you're also trying to look at the economics and the opportunity to offer those services, both mobility and fixed wireless. So it's definitely [ on the moment ], but the thinking is always mobility first. I will tell you that there's a couple of technology advancement that I think will start to show up later this year and that will probably help with the economics. One is improved chipset in terms of the power as well as the ability to have CPEs or antennas that are self-installed. I think that once those capabilities will be introduced, that will make the economics a bit different.

It sounds like, reading between the lines, is that you're interested, you're not against it. But at least from an AT&T corporate strategy perspective, it's not something, at least right now, that you guys are meaningfully pursuing.

As I mentioned, it's part of the build. We are building the network. And if we can -- and when it makes sense, we will offer fixed wireless. But again, the notion is always mobility first.

Okay. I wanted to shift a little bit and talk about what we define as true 5G. And what I mean by that is there's a lot of talk about 5G from a phone-oriented experience and making your phone faster. But there's other components to 5G, there's the IoT aspect and then there's also the latency aspect. And the thought is that, that will enable new use cases for a wireless network that go beyond just the phone. How big of a focus is that for AT&T? And how realistic? Or from a timing perspective, when do we start to see this become more material, do you think, for companies like AT&T?

Yes. So Colby, AT&T is one of the top operator when it comes to -- global operator when it comes to IoT. We already have more than 70 million connections globally. In fact, we are connecting -- every 2 seconds, there's a new device connecting to our network. So that's pretty significant. We feel that we are very well prepared and ready for 5G in terms of IoT. There's a couple of things that we've done over the last several years. We build a nationwide LTE and a narrowband IoT networks, actually U.S. as well as the majority of Mexico. We probably have the broadest roaming agreements globally in terms of IoT on top of those networks. So we feel we're really ready to the point in time where the specification will now continue and the notion of massive IoT shows up. And just to -- maybe to explain to the audience. When I mean massive IoT, that mainly means that moving from the ability of the network to handle thousands of devices in a square miles to million of devices in a square miles. So you take those capabilities, there's a couple of other things that are happening at the same time. There's a lot of effort around miniaturization of the chipset and the devices; and with the edge showing up so you can offload a lot of the computation and the capabilities into the edge of the network, which allows you to have small devices and all kind of shape or form of devices that people can carry, whether it's -- it can be autonomous cars, it can be body cameras, it can be AR or VR devices. Probably what still needs to happen is there's -- besides specification, like every new technology, takes some time to the cost curve of the chipset to go down to get to a point where people can start deploying this in masses and building devices in masses. But I absolutely believe that this is going to happen. And we are very well prepared and in great place to -- and ready for this.

Yes. You talked about the edge. And there's a lot of intermingling of words between 5G and edge. And the way we thought about is that 5G is dependent on edge, but edge isn't necessarily dependent on 5G. What is AT&T's edge strategy? And how do you think about its importance?

So first of all, I read your research that you and your colleagues wrote, and I think it's a fantastic paper. I recommend everyone on the audience to read it.

Thanks.

I do think there's an area there that probably collectively, as an ecosystem, we should think about is how do we look at this from the eyes of the developers, now that we're looking into a distributing cloud and how you're think about your development pipeline? I think the whole notion of placement and stuff on the eyes of the developers is something that we're probably going to have to go and pay attention to. Edge is a critical part of 5G. In my mind, there's 3 technology that comes together: its software-defined networking, it's 5G. And one thing that we don't talk enough about 5G, and we keep talking about low latency and high throughput, the one thing that we don't talk much is that 5G is a cloud-native, software-based architecture. And now you take this...

What does that mean. So I'm not an engineer.

Which mean that a lot of the functions of the network now can be deployed very easily on cloud infrastructure. And it's disaggregated and it allows us to expand our network into the edge. We can now take some of our network function from our mobility pole and put it very close to our customers. And it's not only the edge of the network, it's only -- it's also in enterprise premises. So think about our strategy in 2 ways. One is what we call an AT&T network edge. This is where we keep pushing out in distributing network close to the customer to allow that high-reliability, low-latency use cases. We're typically doing this, allowing third-party cloud to co-locate with us. We've made announcement with the Microsoft Azure and recently with Google Cloud. And that convergence of compute, storage and networking is in the essence of what we're doing in edge and 5G. So that's one way to think about the strategy. The other thing is that...

Would you actually put the -- in that situation though, like you mentioned Microsoft, so we'll use them as the example. And you're putting those into various MEC locations. Would parts of the AT&T 5G network reside, in theory, on Azure's cloud stack? And that would be a more efficient way for you to run it? Or is it really about enabling third-party applications and use cases that would sit on that Azure stack, but just happens to be in an AT&T MEC, and therefore could get access, if you will, to those end devices or customers on that AT&T wireless network in that given area?

It's the latter. It's mainly about third-party application running on public cloud that is co-located because we have the technical spaces. Our central offices are spread all over the country. So we have the location to be able to host not just our network, but also third-party clouds. The second way of thinking about is we have a lot of enterprise customers. And interestingly enough, a lot of our enterprise customers showed a lot of interest from day 1 in 5G and in MEC, what we call multiaccess edge compute, that resides on enterprise premise because it provides a lot of benefits for those customers. First of all, now for us because of this new architecture, it's much easier for us to extend our wireless network into the enterprise. Rather than building a completely new LTE or 5G network, we can just extend the network into the enterprise. And it's the same form, you have compute and storage converge with this. And now enterprise will think about automation in manufacturing, right? Now I can connect to robots and sensors, everything that requires extremely low latency and reliability. The connection is staying on-premise, from a privacy and security perspective, the data doesn't leave the premise. So that provides a lot of benefits to enterprise. So edge is an important part of our strategy. Think SDN, 5G, edge, think about us getting out both to support low-latency use cases at the edge of the network: Cars, drones, mobile gaming, et cetera; and at the same time, doing the same thing working with our enterprise customers, extending our network into their premise to allow them to do some of those use cases that we've discussed.

Awesome. With that, Igal, we are out of time. Thank you so much for doing this with us. We very much appreciate it. I look forward to seeing you, hopefully in person, sometime in the not-too-distant future. And be safe, and again, thank you.

Thank you so much, Colby, for having me.

Thank you.