Viasat, Inc. (VSAT) Earnings Call Transcript & Summary

Hi, everybody. I'm Peter Supino, the telecom, cable and satellite analyst at Sanford Bernstein. Thank you for joining us at Bernstein's Strategic Decisions Conference, the 36th Annual Edition, for this conversation with Mark Dankberg of Viasat. Before Mark and I start talking about his business, a few FYIs relating to this year's format. [Operator Instructions] I'm certain you'll find our dialogue exciting and want to ask questions, and I wouldn't want you to be held up by technology. So with that, I want to also remind you about the Procensus poll, which is a sort of give-and-get technology we're using to poll all of our guests at the end of the session. You'll have immediate access to the poll results. The link will also be available on the left side of your screen. If you have any technical difficulties or have any questions, please reach out to your Bernstein sales contact or Bernstein corporate marketing. With all of that out of the way, we can get to the good stuff. It's my pleasure to introduce to you Mark Dankberg, who is the CEO and Chairman of Viasat and who founded the company in 1986. Mark is thought of as one of the leading minds in the satellite industry. And it's really exciting, Mark, to have you with us.

Thanks, Peter.

So I've planned some questions, and we'll start with those. We'll look for the audience's input, and hopefully we'll end up going in some surprising directions. So Mark, this audience at SDC is a blend of portfolio managers with broad purviews and analysts who specialize in telecom. And in that context and with a satellite bankruptcy in the headlines these days, let's begin by discussing how Viasat is unique in the satellite industry? Especially, in context of your history as an equipment vendor.

Okay. Yes. So I'll give a little bit of background there. We started the company as a technology company. So we sold equipment first to the defense departments and then to the commercial satellite market. And what we saw from, let's say, in the 1980s and 1990s, where there was no such thing as broadband, like there were no cable modems. Even in the early 2000s, the whole motion of, would cable modems be successful, was unknown, right? That was unknown to anybody if there's really a market for that. So what we saw in a time that the company grew from the '86 until the early 2000s, we started in my garage and we were up to several hundred million a year by the early 2000s, was yes, broadband was going to be a big market, that the -- that giving people megabits to the home instead of kilobits was going to be transformative in business and also that there is a need for that everywhere; in airlines; government, we could see all that. And what's really interesting about the satellite space is that, satellite assets, everybody's satellite assets in the space were really geared around the broadcast market. So the original thing -- and basically, think of broadcast, the difference between broadcast and unicast is, broadcast is I'm going to send the same bits to everybody. So if I'm a broadcaster and I want low cost of transmission, basically the way I get low cost of transmission is I can make the number of people that get those bits really high. So I think the cost of the bits could be high, but if the number of people receiving them is really high, I don't need a lot of unique bits, and I can have low transmission costs. And that's how the space business evolved. What was interesting was, well, because the peak rates were in the megabits, people could use it for broadband. But when you use it just for unicast and then send it to just one person, that put a spotlight on how expensive those bits were. So the thing that we could see happening in the late '90s and early 2000s was, if space were going to be relevant in broadband, you needed new satellites, that all of the assets that were up in space were designed for broadcast. And think of like in the terrestrial business, it would be like, hey, if I'm a broadcaster, I'm going to put an antenna on a mountain. We're on top of a tall building and use a big signal so that as many people as possible can get that broadcast. But if I'm in a data business, and this really came home with the iPhone in 2007, right, when the iPhone came out, all of a sudden, the cellular networks were jammed. And this is really the value proposition of Qualcomm. Remember, CDMA versus TDMA was, hey, this is way more capacity. I can have smaller cells. The cells don't interfere with each other. So we realized, well, what we had to do in space was to basically come up with new assets that were satellites that had -- instead of having small numbers of big beams, had a large number of small beams. And everything about the technology, that was going to be different, just like everything about the technology between broadcast radio and cellular radio is different. So what we did is, we went to the satellite operators and said, "Look, hey, you need totally new satellites if you want to go after this market," and they didn't want to do it. And so going back to the -- your introduction about generalists and telecom is what we saw from a generic business strategy perspective was this was like a classic disruption. It's like, hey, there was a new dimension of value in space, which was broadband, required totally different assets, which at first weren't going to be super efficient but that you can get on a learning curve and that none of the incumbents wanted to do it. We went to them and say, "Look, here's how you build satellites. Look, here's the design on satellites." None of them wanted it because the upshot of it was to reduce the value of their broadcast assets in space. So nobody wanted to do it. And we said, wow, this is a huge opportunity for us. And so starting around 2008, we decided to enter this space. And going -- and one of the ways that you can just observe what's happened, and it's very timely, Intelsat's declared bankruptcy, Intelsat has 50 satellites, about $2 billion a year in revenue. So think of that as $40 million a year in revenue. So the productivity of those assets isn't super high. We have 4 satellites and about $1 billion a year in revenue. So think of it as revenue per satellite is about 6x higher. The bandwidth that we have is multiples of all the bandwidth that Intelsat has. And most of their revenue, almost all their revenue is really on the broadcast space. We have no broadcast space. So what we've done is basically we've created a space company that's totally broadband oriented. We've built assets that are unique for that. And if you look at the productivity of our assets, let me just put another chart on that in -- yesterday in our conference call. The productivity of those assets continues to grow. We're on a learning curve that we think is going to allow us to separate ourselves even further from all the other space companies. So that's -- and I think going back to this whole issue of consolidation and what are the assets of those -- the Intelsat, that ones in particular were leveraged up because people believe that the cash flow from those assets would go on indefinitely. And what's happened is as the broadcast market has shifted to more of an over-the-top market and contracted, that hasn't turned out to be true. The productivity has declined, demand for broadcast is going up, and their assets aren't really super well. They're just not very good in the broadband space, in our view. So what we think is, to some extent, the bankruptcies that you're seeing, and they're the ones that are most stressed, are ones that are either companies that are broadcast-centric, that don't have good broadband assets because they were trying to defend the transponder pricing that they had. Or they were resellers, Speedcast being an example, that were basically all using these broadcast satellites. And again, so their productivity can't be higher than the productivity of the assets that they're using. And then you've got others that are also built around those assets that are now highly stressed, and that might be examples like Gogo or Global Eagle. And it's not totally clear that there really is a lot of value to us in our business in those assets on a go-forward basis.

So I think that background, particularly your comments on broadcast versus unicast and the relevance of that transition to the provision of broadband services, links to something I've learned from you in our prior conversations, which is the importance of proprietary technology in the satellite business and the distinction between the role that proprietary tech plays in satellite and the role that it might or might not play in mobile cellular communications. So just to share with you a framework that you could pick apart, I think many of our clients look at satellite and think that the cost of buying a satellite is equal for anybody who wants to be in the business and the capacity that buys you as a commodity. In the absence of any type of compelling on-the-ground distribution advantages, your returns just can't be good. And you've explained to me that the satellite business does not rely on standards-based equipment, so there's a lot more room for proprietary technology to create different cost positions in the industry. And then you seek to differentiate on cost. And so I just wanted to sort of offer you that on ramp to share with the audience how different it is from the industry we've -- many of us have spent more time thinking about, which is mobile.

Yes. And that's -- it's a really good analogy, and the distinction is very important. So what's interesting is if you look in the mobile business, think of this issue about the importance of data capacity. Remember in the early '90s, the cellular industry is standardized on TDMA, right? And Qualcomm was this lone voice saying, well, CDMA's got a lot more capacity, thus, it's going to be important. And they got a couple of operators to bite on that, and that was where you had Sprint and AirTouch, which became Verizon, sort of doing CDMA and the TDMA and the wars between CDMA and TDMA. The thing that really cemented that capacity was important again was the iPhone. And really what unified the CDMA and TDMA was basically the third generation, the 3G, it became Wideband CDMA. It was kind of the unification of that. And it's obvious that data capacity is the dominant issue, and that's really what 5G is about as well. Well, in the satellite space, it turned out that all of the incumbents were going, "Hey, I don't really like this broadband thing. It devalues my existing bandwidth. I'm not sure the market's really there for satellite broadband." And so there was never that convergence on standards. The other thing that we observed as a technology provider was while the standardization was good, it was pretty good for the operators and really good for customers, it was really bad for the telecom technology providers. So if you look at all the companies like Motorola and Ericsson and Nokia and, who was it, Siemens and Alcatel and Lucent, I mean all those companies have been devastated by standardization because it commoditized the technology. So here we were sitting on this technology. We're going, well, the operators don't want it. The market is demanding it. We should use it and become a vertically integrated technology and operator company, and that's the best way to capitalize on the technology. So starting in 2008, that's what we did. And one of the first things that happened is our satellite manufacturer when we did ViaSat-1 copied our technology and sold it to one of our competitors. So we had -- one of the good things was that tested the value of our technology in courts. We went through a court case, and we won a jury trial and several hundred million dollars. And so that validated that we really did have proprietary technology. And then what we did with the second generation and on is we protected that technology more by making -- by getting into the payload business, which is really where the technology is. And we've actually -- one of the things that has suffered is because there's not technology standardization, and remember the satellite manufacturers would only build what the operators want and the operators -- the traditional operators didn't want this broadband stuff, basically, we kind of have seen our technology business recede, but our services business has grown tremendously. And we think that that's going to benefit in the future as the gap between our technology and the technology you can buy on the market, that is the stuff that's available to other operators, that the separation grows. And so what we think that people should be really focused on, you can see this in the other worlds as well, is think of it as there's 2 different things: there's a market value of the bandwidth that makes your company economically values -- valuable, what is the market value of your bandwidth; and then there's a cost of that bandwidth, which is the productivity of your assets. And if we can basically get big advantage in productivity, that's what creates economic market value for our shareholders. And that's what we're doing.

So that segues, I'm going to slip, implausibly nicely to what I was going to ask you next, which is I wanted to ask you what are the least well understood or most misunderstood aspects of Viasat today? And you talked a lot about the cost position of your company and your ability to provision bandwidth cheaper than your competitors and, importantly, your ability to share some of that efficiency with customers to lock in your competitive position. So I don't know if that qualifies as one of the most misunderstood aspects of your company, but it certainly intrigues me as something that many investors might recognize incorrectly.

Yes. I think -- so first thing is, I think we acknowledge that investors are looking for good value propositions. And so to the extent that they don't understand it, it's probably mostly us not communicating it effectively. And that's -- so one is we really appreciate the opportunity to do that. But I think that the idea here is that if you look in the market that -- and we think these are validated in other telecom markets like in the cable market, right? The thing that makes the cable companies such a good -- so such a good investment compared to the traditional telcos is that their bandwidth productivity, their assets have been much more productive. They can deliver higher speeds than the traditional telcos can. Fiber has not been as cost effective partly because the cable companies got a big lead, and there's this game theory aspects of overbuilding. But cable assets have been very productive. So what they do is, they basically offer more value to their customers. Cable ARPUs are quite a bit higher than telco ARPUs are, but the amount of bandwidth that the cable companies deliver to their customers is far higher than what you'd get with a DSL derivative. So you might see ARPUs being 2:1, I'm just picking an example, but the speed differential can be 5 or 10 or 20:1, right? So we don't look at that as weakness on the part of the cable companies. What's happened is they've created a value proposition that's really strong. The -- as they get productivity advantages, they give a lot of that to their customers, but they -- their margins continue to improve. And what we think a really -- another really good example of that would be Intel during the Wintel days when desktop computing or laptop computing dominated as opposed to mobile computing. They've had -- they've struggled in getting that same advantage in the mobile space but still have it in the desktop space. And what happened is Moore's law gave them productivity advantages. It would give some of those productivity gains to their customers to preempt competition, but their value continued to grow as they went down on the Moore's law learning curve. So we think we can create the same thing where the market value of our bandwidth is driven by customers' next best alternatives. In some places like the residential business, that's determined by terrestrial, could be terrestrial wireless or wired. But we tend to compete only in those markets that don't have really good fiber or really good cable. And that's -- there's 20 million-plus households in the U.S. still on DSL. That's indicative of the market that we can attack. But the most attractive markets for us are those where we're only competing with other satellite operators. And if our productivity continues to grow relative to them, we think that we can create a franchise that's very long-lasting and very high margin. And those markets are things like government markets where they can't rely on terrestrial because most of the time they're operating in foreign countries, so they can't rely on the terrestrial infrastructure there. They can't rely on the mobile infrastructure because that builds on the terrestrial infrastructure. Same thing for in-flight connectivity, for passenger ships, things like that. There's many markets that -- where the competition is going to be other satellite operators. And the dominant way that people measure value, the buyers measure by speed and bandwidth. And so those are the metrics that we look for productivity advantages in.

So keying on that subject of your key customer verticals, I want to ask you about your communications -- satellite communications segment -- Satellite Services segment, which has 2 major components. One is the residential broadband business, with which I think many people are familiar and which isn't growing and which you've described as not being highly, highly profitable. And then there's the In-Flight Communications segment, which is growing more quickly and has grown extraordinarily over the last several years. Maybe differentiate for us what makes IFC the place where you are allocating your incremental bandwidth today rather than residential fixed?

Okay. So the one thing I want to talk about just on the residential, to start with, is, this is an asset-intense business, as all of telecom is. And so one of the things that you'd want to do is get the maximum return on those assets that you can. And so as an example, one of the ones we look at it in the telecom space is the mobile business, where they have a finite amount of bandwidth which is really driven by the amount of spectrum they have and then the densification. And so if you look at the mobile space, the yield that they get, let's say if I've got a customer' who's going to pay me $60 a month in the mobile space, they might have to deliver 15 or 20 gigabytes of bandwidth to that customer. Whereas in the residential, fixed residential space, a customer for $60 a month might expect 100 or 200 gigabytes, right? So the mobile companies have prioritized mobile business over fixed. So what we do in the satellite space is, we're prioritizing those markets that give us the best returns. And that's basically the government and in-flight, as we've described before. And that's because they get packaged with other services in order to make that viable. It's not just a pipe. We do a lot of other services for those as well, which can overall improve the margin. But on the residential space, the residential space hasn't grown in subscribers very much, but it's grown in revenue and margins substantially. And that's because the -- the simple analogy that we use is, it's way better for us to have one residential customer at $100 a month than to have 2 at $50 a month. It's because we only have one set of acquisition costs, we only have one amortization. We tend to have lower churn because those customers are getting more. So over the last 2 years, I think that our residential business has grown by 25% or 30% in revenue even though subscribers have been relatively flat because we've migrated our base more upstream. Because we have a finite amount of bandwidth, we've targeted higher-end customers, okay? Okay. Now -- but on the other side, the in-flight and the government businesses are growing fast. We've gone from -- if you were to go back just in our Satellite Services segment, which is about $800 million -- for last year, a little over $800 million, about 25% of that came from nonresidential. About 5 years ago, over -- it's more like 95% of our business was residential. So we've been growing our portfolio. That's mostly been In-Flight Connectivity. And we've been very successful there because our value proposition is so high because we do 2 things. One is, with just cost of our bandwidth is a lot lower, so people get more bandwidth per dollar. And if you fly on our airplanes and compare that to, say, Gogo or Global Eagle, you can just -- you can experience that because video streaming works better. The cost of a session on the airline is less expensive. And then the other thing that we do really well is, we have the geographic distribution of our bandwidth is way more favorable. So we can deliver a lot more bandwidth to hub airports like Dallas, Chicago, New York, Los Angeles, San Francisco for our major airline customers. So that makes our service a lot more attractive, and that's grown. We've grown market share in those. What's interesting, think about from a go-to-market perspective, and by the way, our government market has grown from tens of millions to a couple of hundred million annually. That's in our government segment. So the issue is on the enterprise or government side, you go through some form of acquisition where you go through an acquisition authority, you have to convince that one. Once you do that, then you get your business in lumps as you win an airline or you win an organization. On the consumer side, we can target that just basically by the amount of money we spend on marketing, by the discounting that we use for particular promotions. And so we have the ability to move bandwidth back and forth between them. And actually the fact that there is churn on the residential side, it's -- on the one hand, it's bad because, in some sense, we had acquisition costs that are wasted when a customer goes away; on the other hand, it's good because we can then repurpose that bandwidth for these higher-value markets as we penetrate those markets higher, right? And that allows us to increase the yield on our satellites over time even when we don't increase the bandwidth.

But in speaking of moving capacity around dynamically, one of your priority markets, IFC, has recently faced a huge blow from COVID, and so we seem like we'll have excess plane capacity for quarters, if not years. And I wonder how that is affecting your business? And then more importantly, if you could just share a multiyear outlook for whether you can use this period of dislocation to actually accelerate your market share gains. And by the way, that's -- I should say, just to keep people focused on it, that's a question from the audience.

Okay. Good. Yes. So one thing right now, I think one of the companies, I think, is one of the best companies in the world that's really suffering is Disney, right? And that's because they have these theme parks that go out and to leverage the value of their other media and entertainment assets. But right now, those theme parks are a huge drag because they can't -- people can't go to them. But I don't think there's another media company in the world that wouldn't instantly take over all those theme parks if they could, right? If Disney said, "Okay, we don't want these anymore," I think there'd be huge demand for them. And that's how we look at our airline business because, right now, the airline business is very stressed, and we have a bunch of our business tied up in that. But we think that business is coming back, and we wouldn't trade that for other businesses with others. So that is impacting our business in the short term. Now the telecom is a high fixed cost, low variable cost business. So a little bit under 10% hit to revenue is much more than that from a margins perspective, so we're dealing with that. We've had to make some cost cuts in order to be able to preserve our cash flow, and that cash flow is funding our new assets. But on the other hand, if you look at the point that you're making is, hey, now we can repurpose that bandwidth, and so we are seeing increased demand on the residential side, as are our pretty much all terrestrial telecom providers because of work at home, school from home. And so we can repurpose some of that bandwidth. We're very [ careful ] not to repurpose it in a way that's going to inhibit our ability to service our airline customers as that comes back. So we're going to be -- I would say we're going to be conservative on how we repurpose the bandwidth, but we [ can't ] . And one of the things that you -- I think I've mentioned this before is whereas there's conflict in the supply chains of others where the distribution is tiered between value-added suppliers and the route satellite operators, for us, we don't have those constraints. We can move our bandwidth around as most effective. And that's helping us. We are getting good growth in the residential, and we think we're going to see good growth on the coverage as well in the near term.

Several of our guests have asked about your reception outside the United States, particularly in light of the pandemic, whether that's impacting demand and what your emphasis and strategy will be outside the United States in the next year or 2?

So in -- outside of the U.S., our residential business is smaller. We have been growing -- it is growing. We've been doing what we call shared services, so shared services, for instance, shared WiFi rolled out into rural towns in Mexico and Brazil and offer people connectivity via WiFi, and basically, people will pay cash for that or they could pay digitally for time. And those are nascent businesses for us. Right now, I would say, growth is a little constrained because we haven't really built up the distribution as much as we have in the U.S. And so it's a little bit harder to do in the current environment. Brazil is very affected as one of our main markets. And Mexico is also pretty affected. But we will be entering other markets. They won't show up as much in our aggregate results, but those markets will be growing. And we're still working in Europe where there also is increased demand, but we don't necessarily have the distribution yet to capitalize on it the way we can in the U.S. Internationally, we are growing in In-Flight Connectivity. And the -- although we've gotten orders, the airlines don't yet want to announce them. So we -- they haven't wanted to make announcements yet. And we are getting almost all of our [ growth ] is international as well. Did that answer your question there?

No, it does. It seems like generally, bandwidth demand is accelerating faster than the 5-year planning cycles that associate with satellite launches. And I think your ViaSat-3 project looks extremely capacious absolutely and relative to anything you've done in the past, and yet the planning for that began years ago. And so I wonder how do you plan a business and how do you have an expected return on capital when the number of customers you can serve and ultimately the economic demand of theirs that you can serve continues to exceed prior expectations. I think that happened in your residential broadband business years ago.

Okay. No, I wouldn't put it that way. But the basic point that you're getting at, I would say, it's a foundational pillar of all information technology. If you look at, let's say, 3 pillars of information technology: compute power, storage and transmission, right? And so storage and compute power are basically directly Moore's law businesses, right? You can -- what's happened over time is you can look at the trends of productivity improvements for memory, disk drives, solid-state disk drives, compute power, and you can sort of project into the future. And then you can also project the demand. Now the things that drive supply and demand are different, right? The things that drive demand on the compute side and on the storage side and actually on the transmission side are all video related. It's video processing on the compute side, so higher resolution, more compression; on the storage side, it's video files; on the transmission side, it's live or over-the-top video. So basically you can deduce the trends, and kind of the long-term trends have been in the 20% to 30% annual growth in per capita bandwidth consumption from a transmission perspective, right? And that's been offset a little bit by ARPU growth. So like if you look on the cable side, you've had ARPU growth of probably 6% to 7% over the last several years, and that's projected out pretty well. And then if you look at rural cable providers like Cable One, their ARPU growth has been in excess of what has been for Comcast and Charter, right? So think of that ARPU growth as offsetting to some extent the deflation in bandwidth. So what we do is we say, well, look, hey, we've got the same effects, and we've got -- and they're for the same reasons. It's primarily video growth. What's happened with us is we've gotten ARPU growth that's higher because we're in markets that are higher demand, and we're all catering to the higher ability to pay segments of those. That's true all the way across our portfolios. So actually, even though you can look at per capita bandwidth growth, it's been offset by ARPU growth for us to a greater extent. So what we've projected over the life of our satellites, which drives return on capital, is bandwidth deflation in the 10-ish percent range. And we're actually running a little better than that. Now you can't tell until you're done, but we have satellites that are -- like our oldest satellite is 15 years old. The next one is 13 years old. The first when we built ourselves ViaSat-1 is coming up on 9 years from launch. And our empirical history has been pretty good. It's been consistent with kind of that 10-year deescalation, which is consistent with 20%-ish return on capital over the life of these assets.

Speaking of return on capital, I wanted to make sure that I asked you about contribution margins, profitability and returns on your first- and second-generation satellites. And if you'd venture a guess about ViaSat-3, said differently, what are the EBITDA margins of those if they were evaluated as a stand-alone business before allocation of corporate? What is the contribution margin of those legacy satellites? And what might -3 be?

Okay. So you've touched on a couple of really important points there. Number one is when we look at return on capital, we really want to do is look at the return on capital of the assets that are in use, right, and subtract out the -- don't count the assets that are under construction, right, that are not deployed yet, right? Okay. Then the other thing that you really have to understand there is that the new assets that we're bringing to market are growth assets. They're not purely maintenance assets. And that's because that we have this relatively predictable rate of depreciation of those assets, right, from a value of the bandwidth perspective, okay? So basically, then now -- then the next step, to answering your question is, to realize that this is a high fixed cost, low variable cost business. So what you've seen with our assets, and basically we're overlaying, like first we had a return curve for ViaSat-1, then we overlaid ViaSat-2 on top of that. With ViaSat-1 before we did ViaSat-2, what you saw was that the contribution margin, that is the -- think of it as delta EBITDA divided by delta revenue, right? So one was the contribution margin of individual revenue, that increased over time as the satellite became more and more full because our fixed costs didn't change and our variable costs were far lower than the incremental revenue that we got. So by the time we were in like that fourth-ish year of Viasat -- fourth and fifth years of ViaSat-1, we were seeing delta EBITDA of like 70% for delta revenue, okay? When we started, it was in the 25%, went to 30%, 35%, we kept growing. With ViaSat-2, and then this is -- you noticed this in our results, like in the first year of service, our EBITDA margins were in the 30s range. And they went up like to the 50s range. And they would be even higher except that we're investing right now in some of these new distribution channels, especially international, okay?

And you're talking about your consolidated or segment-level margin?

Yes. Yes, segment-level margins. And so there's no reason to think that the delta revenue -- the delta EBITDA or delta revenue for a given business like residential isn't up in those 70% ranges as we fill the ViaSat-2 as well.

It's exciting prospect. One topic which we'll both be in trouble if we don't touch upon is low Earth orbit satellites. Yes, there's a big hype-versus-reality angle to pursue. You're also focused on LEO as a business for yourself.

Yes.

I mean the way I would like to frame the question -- actually or the way the audience framed their question is, do they create new markets? Or do they take share of existing markets? And curious which direction you would take that question?

Okay. So absent subsidies, okay, so that -- so basically the reason that we -- so the big thing yesterday is we said, hey, we're going to build a LEO, which I think surprised a lot of people. And one of the questions is, is that because we think there's something better on the demand side or supply side than we used to be, that is LEOs are better than we thought they were, or is there something on the demand side. Basically it's the demand side that we're after. And the only thing that's really changed on the demand side are government subsidies. So that is the FCC has said, hey -- through the auction rules and said, "We're going to subsidize low latency through this RDOF subsidy." We want people to build low latency transmission systems. And that could -- that it is not a given that a LEO is a low latency within their definition, but it certainly could be, whereas a GEO on its own can't be. So that's what we're addressing. To go to the -- we don't think that the LEOs expand the market very much, okay? And I think that one of the good analogies again in the telecom world is, if you look at 5G, what people are saying that mobile 5G is going to do is, "Hey, we've got more bandwidth, we're going to offer higher speeds, and we're going to have lower latency." What I think is you're not going to see people say, "Look, we've kept the speed and bandwidth cap is the same, but we've reduced latencies, so pay us $20 a month more." I don't think you're going to see that. I don't think that the latency itself is going to expand the market. I think what's going to really expand the market, if it does, is going to be the speed and the value, right? So that's what we think are the dominant determinants of the broadband space market. We think that offering higher speeds that -- for instance, with ViaSat-2, we're the only ones that offer 100-megabit or even 50-megabit residential service. With ViaSat-3, we'll be able to offer that everywhere in the U.S. and in other markets as well. And we might decide to offer even higher speeds. We will offer more bandwidth, better value according to the types of growth, escalate learning curves that we've described before.

But then -- yes.

No, I was going to say that this issue, what we think is that the LEO market is -- the success or failure of the LEO market is really going to be driven by bandwidth productivity. We think that's part of the reason that OneWeb is in bankruptcy now, is if you look at it, they spent $2.5 billion to $3 billion. They needed another $2.5 billion. And at the end, they were going to have 1 terabit of sellable capacity which is going to be distributed all around the world. For ViaSat-3, we'll get a terabit for 6 or 7 or 100-ish million dollars in the best markets, right? So that's -- that basically is an indication of the productivity gap. And what's interesting is, they talked about latency, they talked about how inexpensive their satellites are. That's what most LEO operators talk about. They don't talk about bandwidth productivity, and that's the thing that we think is really going to set -- that's -- it's going to be hard to have pricing power if you don't have strong productivity. Our distribution is -- we're vertically integrated. People aren't going to have lower distribution costs than us. And then the other really big thing about low Earth orbit satellites is the life of those assets is short. So one of the problems is that, okay, you might be able to -- for instance, you might drive down market price because of the oversupply, but if that only happens for 2 or 3 years because the lifetime of the asset is short, I think we can deal with that. That doesn't make for an enduring value proposition.

So ground network technology, our audience would like to know more about your assets in ground network technology and your opportunity to compete in ground network technology as a service as a third-party provider.

Okay. So 2 -- so okay, 2 -- a couple of things here. One is, one of the potential problems with vertically integrated companies is that their vertical integration becomes noncompetitive, that they've got this internal customer, and they don't really compete well in the open market. And one of the things I'd like to point out is, in the LEO market, we've competed successfully providing payload technology on Iridium, but we didn't disclose it before. But now that they're bankrupt, they owe us a little bit of money because we were one of the main ground technology providers, one for their LEO system. We've also been a ground technology provider for MEOs as well in both user terminals and [ key place ]. So we think we're very competitive. I think we have more staying power than the other ones. So just in competing in technology, I think we did well. For ground-as-a-service, there's not a big market there. But there is -- in communications. But there is a potential market in Earth observation. So that's another market that we're entering. That's kind of a small market for us but growing. That one's -- I'm not quite sure if that's what you're getting at with ground-as-a-service. The other thing I would say, in order to make LEOs or MEOs competitive, ground technology is really important. That is phased array antennas through the user side and also effective gateway antennas which also have to be full-motion antennas or phased array antennas. And the other thing is a lot of these NGSO systems end up doing, I'd say, extraordinary things in space in order to minimize their ground cost. But because we -- basically no satellite can have higher throughput than the fiber network that feeds, right? You can't have a terabit per second satellite unless you've got a terabit per second ground network to heat it. So we have -- not only do we have the largest space networks, but we have the largest ground networks. And it's very extensive in terms of geographic location. We can leverage that when we go after non-GEO as well. So we think -- one of the points we made in our filing is that because we have a very extensive ground network, we can put more of our LEO capacity, had it be based on a more pervasive ground network which has longer life, can be modernized more quickly, and we think that's going to make us very competitive in the non-GEO space. But right now, we think that the LEO demand is really mostly subsidy. That's the dominant thing. And if the subsidies go away or aren't there, it's more a question of how valuable that is? And that's the purpose of subsidies, right? Is to make things that otherwise are not economical, economical, because the government thinks they're important. So the point is we're trying to completely comply with the purpose of the subsidies, but we're just sort of pointing out that the subsidies are sort of fundamental to make a new thing economically viable.

So Ka and Ku-band capacity, your question is, over 5 or 10 years, could you imagine either or both of those bands becoming capacity constrained? Do you think connectivity customers might make long-term decisions with that in mind, if so?

Okay. So Ka -- there's been a long debate between Ka and Ku. I think it's being decided -- people are realizing that Ka is the winner, and the reason is because there's more spectrum at Ka than there is at Ku. It's exactly like -- I mean the propagation characteristics of the 2 are very similar, but there's more spectrum at Ka than Ku. So that's what makes Ka more valuable as a spectrum. And so we've focused on that. Also, it's more available at the times that we did it. But the other big thing is just like in the terrestrial space, once you have a finite amount of spectrum, the way you get more capacity is through densification, right? That is more cells per geographic area, we'll deliver more capacity. So the way you do that in space is either more orbital spots or more capacity per orbital spot. So that's how you can grow capacity. The more orbital spots doesn't increase productivity, right, because your productivity doesn't change just because you have more spots or capacity does. But more productivity per satellite increases both capacity and productivity. And the way you get that is more beams per satellite. And you can't get -- it's very difficult to get more bandwidth per beam because there are regulatory limits on the amount of power you can use per unit area on the ground. So what you need is more beams. And that's the beam-forming stuff that we focused on. I think that we're -- I think we're far in the lead in that dimension of tough competition. Does that answer your question there?

It does. Yes. One last question on ViaSat-3. It's obviously so important to your results in the next few years. How do you think about what type of returns that, that investment ought to yield for the company? Should the contribution margins be any better or worse than the ones you described on ViaSat-1 and -2 as they mature?

Look, I would say there's potential for them to be better. We don't expect them to be worse. And there's good potential to be better. And part of the reason in the single largest market that we're in right now is the residential market in the U.S. And there, we had a competitor which has technology that's very similar to us. As a matter of fact, the satellite, the manufacturer they buy them from basically used our technology. That's what that trial was about. So as we get more separation with them, we think we have an opportunity to expand into our margins in the U.S. That is, we can keep more of productivity gains for ourselves. And certainly on a global basis, there's nobody that we can see building technology with the productivity we have. So we think there's opportunities for margin expansion there. But we don't -- and we don't see any reason that the margins wouldn't be at least as good. I think that the -- I think the most fundamental issue is return on capital, which gets played out over a long period of time. For ViaSat-1, we're 8 years into a 17-year life. And it's on the trajectory that we would like to think. For the orbit satellites, they're 15 years old, those have been on that trajectory. So I think we have a good reason to project that we'll get good return on capital over the life of these assets and that they can be better in the ViaSat-3 era.

That's exciting. One last question. Anything that Amazon does? People want to know about? Do you have any comment on Kuiper? And if the opportunity is real, why wouldn't they just try to buy your consumer business?

So I'll tell you, okay, that's a really good question. Number one is, look, I think the Kuiper system is actually probably the best of all the filings that are out there in terms of productivity, but we don't think it's as good as what we can do with GEO. I can tell you one thing that Amazon says publicly, and so do others that are in their position, their dominant business is basically becoming cloud services, as they do -- whether it's retail services for their own or for others or music or media or gaming, they're really -- and search, those are all cloud services. And one of the things that cloud operators will tell you is, they don't like latency, okay? That they got -- latency, even latency that is imperceptible has a measurable decrease on customer engagement. So -- and we heard this from Google, and we'll hear this from every service provider, is that from a service provider's perspective, people that use high-latency channels will do fewer searches, and they'll do -- they'll watch less videos than somebody who have a low-latency channel, even though that the end customer can't perceive the difference. And on the -- you can't perceive a set of GEO versus LEO, but they're talking about 20 millisecond differences in latency, which are imperceptible. So from their perspective, they want to promote low latency. And that's a good reason for that. The issue is the buyer values bandwidth more than latency, right, the guy that's paying for it. So I think that's kind of a paradox in the -- in this cloud services space. What I would say is, as an example, we work with Facebook and Apple, and they're going, "Hey, people can have a better experience through your service on, say, an airplane than they could otherwise." And that's a great opportunity for us, whether it's airplane or remote area. So I think that what you'll see are -- the cloud service providers will embrace satellite and that if the bandwidth economics, that is the customer-facing value, the guy who's buying the bandwidth is going to value speed and bandwidth more than latency, I think that, that's what's going to make them more successful. But yes, we spend a lot of time looking at new entrants. I think clearly Amazon, we could feel that -- I think Amazon can fund whatever they want in this, no question about that. They look for long-term returns. On the other hand, they also tend to be pretty rational. They don't waste capital. So I don't think they will underprice bandwidth that they bring to market. And on the other hand, they can capture value through cloud services, and that's a big part of what we're trying to do is to capture more of the value of our services than the transmission ones. So I think that one will be interesting. But from the Amazon perspective, they're going to feel that the space safety thing is still a constraint on the whole notion of LEOs and productivity. Right now, the way that LEO operators are trying to drive down productivity is through proliferating satellites even at the cost of lower reliability, for example, to make the satellites low cost. And from a regulatory perspective, it's not clear that's a viable solution. That's one of the things we'd like to address and I would like to -- and we'd encourage people to look at that and ask us questions about it. Peter, it's complex business.

Mark, thank you so much. This has been fun. It would be easy for me to keep going for another 0.5 hour. Our guests have to hustle to the next meeting room. So thank you so much. And for those on the webcast, please remember to complete the Procensus poll as we wrap up, which will be on the lower left-hand side of your screen. Again, Mark Dankberg, thank you so much for your contribution to our event. Best of luck, and we'll look forward to the next conversation.

Great. Thanks very much for having us. I really enjoyed it as well.

Me too. Bye-bye.