Viasat Inc (VSAT) 2018 Q3 法說會逐字稿

Welcome to ViaSat's Fiscal Year 2018 Third Quarter Earnings Conference Call. (Operator Instructions) And as a reminder, this conference call is being recorded.

And I would like to introduce your host for today's call, Mark Dankberg, Chairman and CEO. You may proceed, Mr. Dankberg.

Good afternoon, everybody, and welcome to ViaSat's earnings conference call for our third fiscal quarter of 2018.

So I'm Mark Dankberg, Chairman and CEO, and I've got with me Rick Baldridge, our President and Chief Operating Officer; Shawn Duffy, our Chief Financial Officer; Robert Blair, General Counsel; Bruce Dirks, our Treasurer; and Paul Froelich in Corporate Development.

So, before we start, Robert will provide our Safe Harbor disclosure.

Thanks, Mark. As you know, this discussion will contain forward-looking statements. This is a reminder that factors could cause actual results to differ materially. Additional information concerning these factors is contained in our SEC filings, including our most recent reports on Form 10-K and Form 10-Q. Copies are available from the SEC or from our website. That said, back to you, Mark.

Okay. So we'll be referring to slides that are available over the web, and I'll start with some highlights, and then Shawn will discuss the consolidated and segment of our financial results. Then I'll come back and go into some depth on our residential, broadband, in-flight and government businesses. And we'll review our outlook and take questions.

So the first highlight is that ViaSat-2 is ready for service. We anticipate launching service region by region, starting as early as next week, with national coverage planned for the end of this month. We'll feature unlimited plans nationwide, with the highest speeds available in high-demand markets.

Test marketing of the unlimited plans has gone well, and I'll discuss that in some more detail later. We've been anticipating a significant ramp in, in-flight connectivity, and that's under way. In the third quarter, we entered production service with Qantas in Australia, using our second-generation platform, and we began testing service on both new and retrofit aircraft with American Airlines.

We've signed contracts for 92 additional aircraft from existing airline customers during the third quarter, and that brought our total of in-service and under-contract airplanes to just over 1,500. Then today, we announced a new direct contract with United Airlines for over 70 additional aircraft, including their new 737 MAX fleet. And just to be sure, those new aircraft are in addition to the 92 planes that were signed during the third quarter, so that puts us closer to 1,600 total now.

And that's also significant because United was the only other airline, besides JetBlue, where we were originally in a subcontract situation. Now we're really pleased to be working directly with both JetBlue and United.

And we had another very strong quarter in Government Systems. Revenue is up 9% and adjusted EBITDA was up 20% in the third quarter on a year-over-year basis. Orders were strong at about $200 million in the third quarter, and that also drove backlog to a record for Government Systems at just over $700 million. We see good growth prospects across all the government major product lines, which, as a reminder, were enabled by significant R&D investments that we made in prior periods.

And we'll give more information on that later as well. So, Shawn, just do that financial part.

Thanks, Mark. Overall, our third quarter results were in line with our expectations, based on the business drivers we discussed last call. Looking forward, we are well positioned on multiple fronts. Orders were up 23% year-over-year, with record backlog of $1.1 billion. We're about to launch ViaSat-2 services, and we're beginning to transition to lower payload R&D expenses on the ViaSat-3 program, which will further support P&L and cash flow improvements.

Turning to the quarterly P&L. Revenues grew year-over-year to (inaudible) despite the capacity constraints we've had to contend with, ahead of the ViaSat-2 commercial service launch. Plus recall that last year's Q3 also included $6.8 million of Loral settlement proceeds.

EBITDA reflected the impacts of the elevated expenditures associated with bringing ViaSat-2 into service. Gearing up for strong IoT growth, flow-through of the year-over-year Loral settlement impact and the continuing R&D expenditures on next-gen technologies,

However, I do want to point out that while our Q3 R&D levels were up year-over-year, as expected, we did see a decrease of $6 million on a sequential quarter basis as we began transitioning to the capital portion of the ViaSat-3 program.

So, looking at our segment results, in Government Systems, Q3 was marked with nice growth across broad product service portfolios. Revenues were up 9% year-over-year, reaching $182 million, and adjusted EBITDA grew at double that rate of 20% year-over-year to $48 million.

Product and service margins were up significantly, more than offsetting a $9 million combined increase in R&D and SG&A, which included elevated proposal efforts that helped secure the $199 million in new orders, bringing segment backlog to a record level of over $700 million. And remember, this excludes future awards anticipated under our IDIQ contract, such as the $350 million silicon agreement we recently announced.

End result: It was a very strong Q3 segment EBITDA quarter, with margins increasing over 200 basis points year-over-year. In Satellite Services, revenues were down year-over-year primarily due to a modest downtick in residential subscriber count, which is stabilizing ahead of our new service launch, and the $6.8 million Loral settlement impact I mentioned earlier. These factors were firstly offset by an 8% ARPU growth to a new record high of $68.23.

In IFC, we saw another quarter of revenue growth, with 34 more aircraft in service at quarter-end versus Q3 last year and increases in ARPA.

Turning to Q3 satellite segment earnings. We see the impact of the top line reductions plus the additional expenses we're incurring as we near ViaSat-2 service launch alongside our upcoming large-scale service ramp in commercial air.

As we look forward to the ViaSat-2 commercial service launch this month, we expect to see our segment revenues resume growth with these ramped-up expenses beginning to moderate as a percentage of revenue. Conversely, we also expect to see much heavier advertising activity over the next few quarters.

Plus, in Q4, higher commission expenses related to higher subscriber additions. So I want to take a moment here and quickly remind everyone that the new revenue standard will come into effect for us in April of 2018, and it has 2 primary impacts to our consumer broadband business.

Commissions incurred for new subs after April 1 will began to be capitalized and amortized over those subscribers' expected lives. And two, we'll record or recapture commission asset, net of amortization, associated with the subscribers already on the network at that time, which also will be amortized over the estimating remaining customer lives.

As part of our implementation efforts, we've completed review of many of our customer contracts. With this accounting difference being the single-largest known Q1 impact, and while our implementation review is still under way, I would expect this impact to result in approximately 80% to 85% of our SAC being capitalized in the future versus the 60% to 70% we've ranged historically.

So taking all these elements into account, we should see our Satellite Services EBITDA margins in the first full quarter of ViaSat-2 commercial service higher than where we were when we started on ViaSat-1 and more similar to where segment margins were in early fiscal 2014, which was just over 1 year into service on ViaSat-1. From there, we should see margins begin to increase each quarter, given that this is a high fixed-cost, low-variable-cost business.

Moving to Commercial Networks. Quarterly revenues were up $1 million, with increases in airborne terminal sales related to our IFC business more than offsetting lower fixed terminal sales. We also saw some exciting contract wins for next-generation antenna solutions generating a good order flow and positive book-to-bill ratio in Q3. Looking forward, we expect to see strong revenue growth from ramping Gen-2 ship sets installed as our airborne terminal sales are reporting commercial network segment with the subsequent service revenues being reported in Satellite Services.

Segment-adjusted EBITDA was negatively impacted by a modest increase in R&D expenditure coupled with higher support costs as we prepare for the heavy production ramping commercial air. And again, although segment R&D was up year-over-year, it was actually down sequentially, by approximately $6 million, as we begin transitioning to our ViaSat-3 payload construction phase.

Going forward, we expect R&D expense to continue stepping downward year-over-year to recurring levels that are similar to early fiscal 2017 as a percentage of our revenues. Overall, our Q3 performance reflected the drivers we anticipated and have been preparing for. ViaSat-2 commercial service launch is right around the corner, and we're seeing in-flight WiFi system deliveries ramping, and our ViaSat-3 payload program is now transitioning to the construction phase. These milestones, coupled with record backlog at the company and at our government segment level, have set the stage for another phase of strong growth company wide.

Turning to Slide 5. We see revenue and adjusted EBITDA performance for the Q3 year-to-date period. Top line revenues and EBITDA year-over-year trends across the segment generally reflect the same quarterly drivers we just reviewed, so I won't go through all the details here.

Additionally, I will note that while we've been waiting for ViaSat-2 service launch and have been heavily engaged in preparing for those activities, our business as a whole is still growing, with year-to-date revenues hitting at new record high of $1.2 billion.

Similar to what we've provided in the last 2 quarters, Slide 6 shows a reconciliation of adjusted EBITDA for both the quarter and the year-to-date periods. A reconciliation isolates the large transitional items we've been highlighting in order to show the fundamental EBITDA growth in the base business.

If you look at that Q3 year-to-date reconciliation on the right side of the page, you can see the increased R&D investments, the ViaSat-2 and commercial mobility ramping cost, plus the $20 million impact to the conclusion of the Loral settlement payments all-in together accounting for approximately $100 million of negative variance, which was then partially offset by organic growth from our existing businesses.

Turning to Slide 7. We see our year-to-date income, cash flows, debt and leverage trends. In earnings, we see the EBITDA impact I discussed earlier plus a $12.2 million impact related to the recent tax reform enacted in Q3.

The net result was Q3 year-to-date GAAP net loss of $47.4 million and non-GAAP net income of $5.3 million. Among other things, the new tax legislation lowered the corporate federal income tax rate from 35% to 21%.

As the company has a March 31 fiscal year-end, the lower corporate tax rate will be phased in, resulting in a U.S. federal rate of approximately 32% for fiscal year 2018 and 21% for subsequent years.

Although, the lower federal income tax rate will be positive for the company in the long term, along with other provisions of the revised tax code still under review, in Q3, we did revalue our net deferred tax assets, based on future utilization dates of those assets, to the new federal rate of 21%, which resulted in additional income tax expense of approximately $12 million for the 3 months and 9 months ended December 31, 2017. This noncash charge reduced our per-share figures on a non-GAAP and non-GAAP basis by $0.21 and $0.06, respectively, for both the third quarter and the year-to-date periods.

Turning to cash flows. Despite the ramped-up expenses and capacity constraints, our year-to-date cash generation from our operations was $283 million, and this funded over 75% of our entire capital spend year-to-date, including the final phases of the ViaSat-2 ground networks buildout and our ViaSat-3 Constellation.

Looking at the chart in the lower right. Net leverage at the end of Q3 was 3.4x EBITDA, which is well within our comfort range. As we look outward, this ratio will likely move higher over the next few quarters as we continue the ViaSat-3 Constellation build, alongside incurring SAC cost as we scale our ViaSat-2 subscriber base, with future delevering as the EBITDA begins to grow again.

And finally, our Q3 liquidity position remains very strong with $162 million of cash, nothing outstanding under our $800 million credit facility. So, when we take these 2 amounts into account, plus our [outstanding all fees], we have total liquidity of $933 million.

So, with that, I'll turn it back over to you, Mark.

Okay, Thanks, Shawn. So I'll start with our government segment.

Once again, we had excellent performance in Government Systems across multiple product and service offerings. The products and services are a steady and expanding portfolio that have Link 16 Tactical data links, information assurance in cyber appliances and software, and narrowband and broadband satellite terminal products and services.

We use this chart from the last call as it helps isolate the seasonality of the government business. You can see the year-over-year revenue growth on the left that's attributable to the products and services that I just mentioned. On the right, you can see EBITDA has been growing a little bit faster, even taking into account higher R&D and other selling costs we're incurring to continue positioning us for the future.

We ended the quarter with record backlog over $700 million after strong awards of about $200 million during the quarter. One of our growth strategies is to increase our addressable market by migrating products that have been very effective in the early adopter organizations, such as Special Forces, to the mainstream Army, Navy and Air Force.

During the third quarter, we executed a $350 million ID -- Indefinite-Delivery/Indefinite-Quantity, or IDIQ, contract with Special Forces that helps illustrate the demand for products and services that are generally ViaSat-unique nondevelopmental items as opposed to programs of record.

Just to be sure, we only include firm purchase orders when received under that contract in our new awards in backlog. As those products and services are adopted by other users, we're also aiming to execute similar types of contracts with those organizations.

We're having success with our migration strategy. The Department of the Army recently filed a report with Congress under a tactical network modernization strategy. The report emphasizes the Army needs to move faster to develop and deploy new and commercial technologies, not just for cost-effectiveness but to close capability gaps. And that report specifically calls out 2 ViaSat products we initially developed for early adopter organizations, and it indirectly refers to others that we've done as well.

We continue to be optimistic about our growth prospects. Having an approved DoD budget will be very helpful, as the continuing resolution environment makes it more difficult to contract for new capabilities and/or with new organizations.

So, turning to in-flight connectivity. We've been anticipating a significant ramp in aircraft, starting in this current fourth quarter. And our objective in the third quarter was to pave the way for that. We began testing service on several American Airlines planes, including some of each, both factory line-fit new 737 MAX as well as retrofits to existing planes that had air-to-ground connectivity.

Passenger feedback for our service has been positive, and there is a sample on the chart. We're also in production service with Qantas in Australia. Qantas has been very forward looking in leverage free Internet service to every passenger. And one of their innovative applications is an agreement for live-streaming cricket matches. They also have agreements for other media and entertainment, including music, live-streaming TV and video on demand. We're also making good progress in bringing our European airline customers online with our European KA-SAT joint venture and forthcoming transatlantic coverage on ViaSat-2.

As I mentioned earlier, we added 92 additional aircraft in the third quarter, which brought our total of in-service and contracted planes to just over 1,500 as of the end of that quarter. And just today, we announced a new direct contract with United Airlines for 70-plus additional planes, mostly new 737 MAXs. And that puts us close to 1,600.

That contract also includes our wireless, in-flight entertainment product. We think there's a clear trend towards airlines capitalizing on in-flight connectivity as a cost-effective way to both differentiate themselves and to drive passenger satisfaction. And that will lead to higher levels of passenger engagement, which is our unique value proposition.

Just to reinforce that point, in the third quarter alone, we delivered over 4 million personal electronic device WiFi Internet sessions for our in-flight connectivity customers. We're optimistic that as we expand our geographic coverage with ViaSat-2, ViaSat-3 and possibly additional regional partnerships that we can grow in other regions as we have in the U.S.

Okay, so now let's turn to consumer broadband. Pretty much everything we did in the third quarter was aimed at launching service on ViaSat-2. We've long said that we want to grow the addressable market with higher-speed plans with more bandwidth and more video streaming. So in the second half of the third quarter, we began test marketing new unlimited plans on ViaSat-1 that were coupled with video stream management to extend viewing hours, something we expected would be a good trade for many customers. We were still limited by the amount of bandwidth available on ViaSat-1 in the speeds that we could offer, but the test marketing results appeared very favorable.

At the end of the third quarter, we already had over 40,000 subscribers on these new plans. About 20,000 of those are subscribers who migrated from other existing plans. So we've learned a lot about the attractiveness of these new unlimited plans relative to each of our older existing ones as well as gained insight into price elasticity and the effects of combining different plan types in the coverage areas.

Overall, even given the bandwidth limits of ViaSat-1 and an appropriately limited marketing budget, we've seen churn come down, gross adds increase due to the attractiveness of the plans, and net subscriber losses moderate. Higher speeds and more bandwidth also support a higher ARPU with the combination of the new plans and retail wholesale mix leading to an 8% year-over-year growth to a record of over $68.

Just as an aside, given the macro environment of growing share of over-the-top Internet video, in January, we've seen stand-alone cable broadband pricing from multiple suppliers rise to $75 a month for a 25 megabit per second service, which just helps illustrate the value of high-speed, high-bandwidth Internet access.

So with that as a backdrop, we're really looking forward to the launch of service on ViaSat-2 starting regionally as early as next week. And then the next couple of slides will go into more depth on that.

So first, a quick review of what our objectives have been for the ViaSat-2 satellite and network. We think we're set for a good start on most of these points. We wanted to grow our addressable market with faster speeds and more bandwidth and, thus, that meant speeds up to 100 megabits per second in high-demand areas and offering virtually all our new plans in an unlimited-type format.

We want to expand into new geographic markets. Our in-flight connectivity business will take advantage of the ocean coverage on transatlantic and Caribbean and will likely include Mexico and south of there, too we've also been pretty rapidly expanding our WiFi services in Mexico, but I'll defer more color on that for another day.

One of the most important new capabilities with ViaSat-2 is flexible bandwidth allocation. The idea there is to mitigate obstacles to revenue and earnings growth, where we've got bandwidth available on the satellite, but it's not in the right place at the right time or it's not in the right direction in terms of downstream versus upstream.

ViaSat-2 has a lot of flexibility in each of those dimensions, so we can respond to long-term trends, such as more demand in one geographic area versus another, or short-term imbalances, such as due to different peak busy hour demand in different time zones or due to geographic variances due to airline schedules, for instance.

So with ViaSat-1, we found that downstream bandwidth was a bottleneck. We'd have traded back some of the upstream that we had left over for more downstream if we could. We also had some markets with very high demand that sold out quickly and others with little demand that required more promotions or discounting, or service plans with lower economic yield.

Also, we can use the flexibility to allocate bandwidth to different vertical markets that have different peak demand profiles, such as businesses or mobility or government applications compared to residential.

We know based on backlog, for instance, that our in-flight connectivity business could triple in the next 3 years and represent a significantly larger portion of our services revenue. We also have opportunities to scale other businesses with complementary demand profiles.

So we've got opportunities to drive growth and returns using ViaSat-2 across multiple dimensions in addition to subscriber growth, such as higher ARPU, expanding into geographies and verticals we serve, and more effectively allocating bandwidth in a thoughtful way to opportunities based on the demand environment and the long-term growth those opportunities represent.

So it's a different dynamic than when we launched ViaSat-1. On our current fleet, including ViaSat-1, 90% to 95% of the bandwidth is generally allocated to residential broadband, and the geographic allocation of capacity is fixed. That mix and those constraints should change with the addition of ViaSat-2. And beyond the geographic allocations, we also want to manage our growth trajectory in terms of loading our satellites and optimizing long-term value.

In 2.5 to 3 years, we're planning to have ViaSat-3 available, which is designed to have an even greater productivity improvement, more coverage and greater flexibility so we can further scale existing businesses and enter new markets and geographies. Just makes sense for us to consider these opportunities thoughtfully and strategically and not simply just focus on residential subscribers while still understanding that that's the largest part of our Satellite Services business.

So on the next slide, we'll give some color how we think about turning our additional bandwidth and productivity gains into economic value. The top-left chart is about productivity gain allocation. Our services business sales bandwidth packaged in different ways, and there must be a market value of the bandwidth that you can find by dividing the services revenue by the bandwidth sold.

ViaSat-2 was designed to get twice the productivity of ViaSat-1. We do have an antenna deployment issue that may degrade capacity compared to the ground tests. So we and the Boeing team are working to characterize the on-orbit performance to estimate the maximum capacity of the satellite. We currently estimate the maximum capacity could be around 260 gigabits when optimized for capacity given the antenna system as we understand it.

Still, we believe ViaSat-2 will have at least double the useful bandwidth as ViaSat-1, taking into account it's flexibility, geographic demand models, service mix, available spectrum and our current understanding of the satellite network performance potential.

The really important question is, how should we turn the bandwidth into revenue and earnings? The upper-left chart shows 3 thought experiments in some market -- test market data. The first column would be if we got all the productivity gain. That is the market value in terms of dollars per gigabit and revenue just stays the same for the life of the satellite or for the next 3 years, say. So if we were to triple bandwidth, which would be 1 unit on ViaSat-1 and 2 on ViaSat-2, then our revenue would triple. And that's probably too optimistic.

The next column shows what happens in a hypothetical scenario where the market requires we give all that productivity gain to customers so that ViaSat-2 respectively yields the same revenue ViaSat-1 did. Since ViaSat-1 has about half the useful bandwidth of ViaSat-2, then it revenue would halve. Combined, the revenue would be about 50% higher than with ViaSat-1 alone. And that likely seems too pessimistic.

The third column reflects a bare case, arguing that ViaSat-2 is purely "maintenance CapEx." And we'll, in total, get no more revenue than we did before with just ViaSat-1. So based on just 2 of the points we've shared today, the test marketing results for our higher-value consumer plans and our rapidly growing in-flight connectivity backlog, it should be pretty clear that the market opportunity doesn't seem to reflect that view, which is even more pessimistic than giving all the productivity gain to our customers.

The fourth column is the most interesting, because it reflects data from the over 40,000 ViaSat-1 subscribers using the ViaSat-2-like plans as of the end of the third quarter. And it supports what we expected and planned, that we would share the productivity gain between customers and ViaSat and that we'd see a yield in between the first 2 cases.

So just to be clear, we don't intend this as guidance. It's just a data point, but it's an exciting one in terms of the growth opportunity. Some other factors that could make things better or worse are also shown on that chart next to that fourth column.

Then the lower right-hand chart is about the trajectory over time for achieving the economic yield that's possible and represented in the upper-left cases. When we start, we only have the revenue from ViaSat-1. So as I noted a moment ago, we'd like to optimize that revenue on the combination of our fleet, and we'd like to do that by the time ViaSat-3 goes into service. That's estimated as the second half of calendar year 2020. So roughly 2.5 to 3 years from now.

ViaSat-3 is designed to improve productivity again. So we'd like to have filled our available bandwidth by the time it launches service. So if we sold bandwidth too fast, we may be undervaluing it relative to its market potential, so we might want to slow down a little and increase our yield. If we're building revenue too slowly, we may have overestimated its market value; then we should probably try to speed up.

Somewhere in there is the "Goldilocks, not too hot, not too cold" rate. And that's what we're focused on finding, based on facts from the market and having nobs and levers to adapt. So just to be clear, we're really focused on the revenue and earnings trajectory, and that's not exactly the same as focusing on subscriber accounts.

So this chart on the service plan approaches helps illustrate how we look at subscriber counts. Residential revenue is simply the product of net subscribers times ARPU. Simply stated, fewer subs with higher ARPU is more profitable than more subs with lower ARPU. Subscriber acquisition cost expenses and investments are lower. Service levels are better, so churn can be lower given the right subscriber base. We've designed plans that are consistent with this approach and still should help grow our addressable market.

Based on our third quarter test marketing, the economic yield of the plans is good, demand currently seems sufficient to achieve the revenue trajectory we want, and churn can be better than with lower-priced lower-performing plans.

Also, more attractive plans don't require as much marketing and promotional spend. The upshot of all this is that we'll measure and adapt our consumer business based on the revenue and earnings trajectory we can obtain, driven by economic yield of the service plans and find the number of subs that delivers those results.

Finally, the other point that affects subscriber counts will be the total amount of bandwidth that we allocate to residential service. Based on our growth rate in in-flight connectivity, WiFi and government services, and taking into account opportunities in enterprise and maritime, we anticipate that a significantly smaller portion of our total bandwidth of the fleet, including ViaSat-2, would be allocated to residential than the 90% to 95% on our current fleet. So that could limit subscriber count until ViaSat-3 is in service, but it would likely improve revenue and earnings, and it would enable additional markets that would increase our overall growth outlook.

Okay, so to wrap up, I'll quickly summarize the main drivers we see in each of our business segments. Government Systems has been executing really well. Overall, we see good opportunities to continue that. We continue to be successful with our unique nondevelopmental items as compared to programs of record. Although, we've also been successful on a number of programs of record or turning our NDI products into programs of record, too.

The Defense Department has become more receptive to incorporating commercial technology to close critical capability gaps. We've had a number of unique capabilities that we can leverage in tactical data links, cybersecurity and satellite broadband products and services.

In the short term, the lack of budget authority and reliance on continuing resolutions makes it kind of challenging to continue to introduce new products and services or to work with new organizations because these continuing resolutions are intended to merely maintain the status quo. But so far, we've been successful, and our Special Forces IDIQ is a good example of that. But short-term quarterly variations are always a possibility.

Our Commercial Networks business has several leading indicators for revenue growth and lower R&D expenses. We've seen positive book-to-bills, anticipate a significant ramp in in-flight connectivity equipment sales which show up in that segment, and we've seen sequential reductions in ViaSat-3 quarterly R&D expense.

We've continued long-term opportunities in advanced ground networks, including flat-panel phase arrays. And our success in capturing new airline customers though does create ongoing needs for retrofit and new line-fit supplemental type certifications or STCs, for additional aircraft, and that does involve additional R&D expense.

On Satellite Services, the biggest factor is the launch of services on ViaSat-2. We've got just one more quarter of year-over-year comparables that include the SS/L settlement payment. So that'll go away. ViaSat-2 enables us to return to residential subscriber growth. But remember that we're most focused on revenue growth and associated earnings, driven by higher value, higher speed and higher bandwidth service plans. Short-term revenue gains will be somewhat tempered by initial promotional pricing. Near-term earnings will also be affected by SAC expenses.

And as with ViaSat-1, though, we expect to overcome increased fixed costs and create a flywheel effect that progressively allows a higher proportion of revenues to fall to the bottom line as our satellite fill. We also anticipate a higher proportion of Satellite Services revenue to be generated from enterprise and government applications, such as in-flight connectivity. That may constrain total residential subscriber growth until ViaSat-3, but it will improve monetization through higher-value applications with complementary peak geographic and temporal demand.

As a reminder, government applications of broadband satellite bandwidth are reported in our Government Systems segment. So that may have some impact on Satellite Services segment results.

So -- and with that, let's open it up for questions.

(Operator Instructions) Our first question comes from the line of Mike Crawford with B. Riley.

In Airborne mobile, what percent of United's fleet will you now have? And how is that carrier going to manage -- do you believe that carrier will manage passenger in-flight experience expectations?

Okay, okay. I think we had around 260 to 280, maybe? 280 in...?

Number of planes? Yes.

Yes. 280 on United before roughly. Out of close to, say, 400 to 450, I think, in their domestic mainline fleet.

Could be a little bit higher than that, closer to 300 of United planes.

Okay, okay. So this will add to our proportion of that. I think United is still -- we don't want to speak for United in terms of what their fleetwide planning is. I think they'll speak for themselves in terms of how they're going to deliver the services through the course of the fleet -- throughout the fleet. I think our understanding is, in general, they're excited about the capabilities that we offer, and I think you'll see that they are probably more likely to try to figure out how to leverage them than to subvert them.

Okay. In Europe, you have a dispute with the regulator there, Ofcom, regarding Inmarsat's use of spectrum that was supposed to be for satellite mobile phones instead for a European aviation network, where they had this 300 tower buildout. And so, a, what do you think are your prospects in regards of that site? And then b, if that purpose -- if that spectrum is able to be used for that purpose, I mean, what -- would you be able to use your Ka-band spectrum for 5G service in Europe?

Okay. Multiple parts to that question. So basically, our position is that the application -- that the EAN application will deliver something like 99% of all the bandwidth that's delivered to passengers on airplanes through ground network as opposed to satellite, that the intent was that the ground network augment the satellite not that it basically be the network and with the satellite providing a tiny part. That's our position. I think it'll be resolved on the merits of the argument, probably on a jurisdiction by jurisdiction basis. I think their -- our understanding is that there are already at least a couple of important countries that haven't approved EAN and that were -- we are contesting it in others. And I think it will be resolved on the merits. But our argument's pretty straightforward. I think that the -- that there's a difference in taking Ka-band spectrum and applying it to 5G networks because Ka-band spectrum is generally applied -- there's a difference between what's called fixed satellite services and mobile satellite services. Fixed satellite service bandwidth is only allocated to operators at specific orbital locations, whereas MSS spectrum, like S-band and L-Band, is more omni-directional. And so, it's not really practical currently to allow each of the FSS operators to use MSS spectrum on the ground. There's really no protocol for doing that.

Okay. And last question, also in Europe, is any update on negotiations with Eutelsat?

Yes, I think both we and Eutelsat are working -- we're working away on the agreements that would result in the ViaSat-3 JV. We're not done. I think we're continuing to make progress on it, though.

And the next question comes from the line of Rich Valera with Needham & Company.

Mark, was hoping you'd talk about the -- it sounds like you're expecting at this point with what you know lower capacity on ViaSat-2 260, I guess, versus the original, like, 300. So I know you before said you thought you could kind of bridge to your original financial projections even if the capacity was a bit lower. Can you talk about maybe how you would bridge to your original financial projections if 260 turns out be the number? And then, what do you think the odds are that you can still come up with a fix that would get that back up closer to the original capacity?

Okay. So we'll start with the second part first. There's still -- there's maneuvers that we can do with the satellite that may resolve the problem. We're working with a satellite manufacturer at Boeing and with the insurers to make sure that everybody understand the situation and that when we take those actions, they're appropriate. And I would say it's not off in the distant future. It's not something we're going to do immediately, either. But I think that there's still -- I have -- we can't really put a probability around what the outcome will be. And the outcome's not binary. It could end up improving the situation. It could fix it completely. It's not inconceivable that it could make things a little bit worse. That's what we're trying to do is to make sure that anything we do is prudent. In terms of how we can work with the capacity that we have and still achieve our original financial results, the little things that we've said multiple times is, basically, when we intend to bring this capacity into service, that will put it in a configuration that maximizes its revenue potential. And that's not exactly the same as the situation that maximizes the capacity. So you think about how you'd maximize the capacity, you'd sort of have to think of it as sort of spread that capacity fairly evenly over the whole coverage area. And part of the point of having flexibility is to use capacity in the places where it has the most demand. So we've basically -- and the satellite has a great deal of flexibility. So what we intended to do was to allocate capacity to those places that have the most demand. That makes it a little more concentrated. And so the tradeoff would be that, at least for the beginning and this may change when we get ViaSat-3, but at the beginning, we look -- just like I said with ViaSat-1, we might be happy to have less total capacity in order to have that capacity in the right places and for the right uses. So, we have demand models. We're putting limited capacity in that configuration. The other simple reason for why we still think we can meet our plans is that we didn't, based on our business plans on perfection, that we had margin in there for problems. And then also, we've had things in other parts of the network work better than we planned. These are things that are on the ground that are not attributes of the satellite. So, I just want to make one example would be -- and this a -- I think this is a really good way to think about it is imagine that you had a satellite that had a solar panel that didn't deploy 100%, right? So the satellite has 10% less power than it used to. The satellite's impaired. There's a partial loss. It can't do exactly what it could have done otherwise, but if you come up with modems that have better performance than what you originally anticipated or you have video codecs that have better performance than originally anticipated, you could still achieve your targeted performance even though the satellite itself is somewhat impaired. And that's pretty much the situation that we have when you consider all things put together.

Got it. And is potential insurance claims any part of that as well?

Yes, there are -- so basically, one thing we've notified the insurers that there's an anomaly on the satellite that we'll try to -- we're working with them to quantify what that anomaly is. We'd be entitled to coverage as a partial loss for the satellite. And we also think that, in the end, that the insurers like us would want to try to fix it. That will evaluate those risks and then we'll make a decision.

Got it. That's helpful. And then just with respect to the in-flight, obviously, very nice backlog now. And you've talked before about having a pretty aggressive ramp trajectory in terms of quarterly installs. Can you give us an update on how you think that might progress over the next 2, 3 quarters?

I don't know if we want to give any numbers. Rick, would you like to give any numbers or -- any color on that?

There's 2 different things going on because we're delivering to Boeing for line fit. And then we're delivering -- we're doing retrofits, and so our delivery schedule is ahead of the in-service schedules. But pretty much they start, on a quarter-over-quarter basis, I think you'll see Q1 -- or Q4 kind of double what it was. And our Q3, I think, those are the numbers, Shawn? And then, you'll see Q1 double that. So it steps up pretty steeply until we get into the summer. And then it -- the rate of growth slows down.

Right. I guess it would still take a few quarters. I thought sort of your peak was going to be pretty well over 100 per quarter. Is that still your thought at your sort of peak run rate?

No, it'll be higher than that. It will be quite a bit higher than that.

Okay. Like 150, is that sort of the number, or ballpark?

Yes. I think probably closer to 170, Shawn, isn't it?

Yes, I think between 150 and 200. So 170 is a good number.

Yes, in that range. And as we are adding more obviously, that number is growing.

Great. And so -- and not to beat this one, but you think you might hit that in -- within the next 4 quarters, is that fair?

Yes, yes.

And our next question comes from the line of Ric Prentiss with Raymond James.

Obviously, exciting times as ViaSat-2 moves from launch to test to market. Did a good job laying out a lot of your thoughts on market testing. Help us understand maybe a little bit about the distribution channels you would use. You've talked about maybe reducing the amount of bandwidth that goes to residential. How do you get at government, mobility -- obviously, inflight, you've got that handled. But just trying to think through how you're going to go to market to these different areas that are going to get bandwidth assigned to them.

Okay. So yes, just -- that's the one thing we didn't talk about. Just in the first case, I mean, residential distribution is still really important. We've mentioned in the past that we've been working on improving and expanding our residential distribution. And I think we've set some good things up there. Also, one of the things that's important is that, for about the last 9 months, we really have had very little direct-to-home TV distribution. And there's a good chance that, that will reengage with the new satellites we've added. So those are things on the residential side. On the government side, essentially, our distribution approach has been working directly with -- to configure those platforms, that is -- we've talked about, for instance, C130s or gulf streams. These are in government markets or specific types of ISR platforms. So one is to work with a lead on each platform in order to get qualified. It's equivalent of -- it's not exactly, but it's the equivalent of getting qualified on that platform, like an STC would be with a sponsor. And then we tend to go to other organizations that use those same platforms. So, for instance, Special Forces are early adopters in C130s or C5s -- or C17s, excuse me. C130s and C17s. And then we've worked with other organizations on new platforms, and one of them for instance, would be the Osprey V-22s and to go to different organizations that are users for those platforms. So, it's pretty much -- think of it as a direct sales approach. These customers use our terminals, and then they'll buy services from us. We also have been working with ground organizations, and as an example, one -- of, again, going back to this Army tactical networks modernization program, there's a lot of issues with what's called the Win-T program, Win-T program is primarily a satellite program, has very large terminals, not a lot of bandwidth, that's a really good example application for us to go into on the defense side. Another area I don't want to talk about too much, because we will talk about in more detail in a few weeks, is WiFi in emerging markets. And there we have couple of different distribution approaches that we're working. One is a partner that provides kind of information -- think of it as information services in remote villages, and those services are greatly enhanced, and their market's greatly expanded by having connectivity along with that -- those types of services. In terms of some other new markets, we have partners in enterprise, who, I think, in the past, we've mentioned you're working with Comcast as an example for a large quick-serve chain where we used satellite to fill in sites that aren't available to be served by their ground network, as an example. We'll expand other applications like that.

Okay. Obviously, not all subs are created equal in this kind of new world, as you mentioned higher-ARPU ones. How are you thinking about reporting or educating the market then on kind of the type of subs you're getting and then what we should think about from a modeling standpoint?

Okay, so on the residential side, we'll continue to report ARPU. I think on the -- we're still thinking about it when we have the applications that are more shared or -- and we probably won't do that until those obligations become material. So the thing, for instance, with in-flight connectivity, at the current levels, it's just not been a material part of our business, but as that backlog gets built out, it will be until we come up with an appropriate mechanism for reporting it and giving investors and analysts a way to kind of value these relative approaches.

Okay. Also in the U.S., there's been a lot of discussions about the CAF-II auction coming up in July? What are your views on that auction and what it might mean for competition or any interest from you guys as far as looking at what's going on with the CAF-II auctions?

So we have paid attention to the CAF auctions. There was at one point a movement to make them more technology neutral. That was somewhat undermined by the bidding rules that allocated value to latency, as an example, relative to speed. We think that we're really interested in it. We'll do -- we'll compete in the best way that we can given what the rules are. In the meantime, through the adoption of our service plans and through studies like this one they came out from University of Indiana over the summer, we're trying to demonstrate that a lot of customers really prefer higher speeds and more bandwidth to lower latency and -- when given a choice. But 2 of our CAF rules really fairly reflect that, and we're constrained in what we can do.

Got it. And last one from me is both Sprint and SoftBank on their earnings calls recently were talking up OneWeb a little bit as possibly ultimately becoming a rural broadband solution. Can you just update us on your thoughts about OneWeb and how much bandwidth it actually brings and the ability to put that product into the market?

So, I think OneWeb and other LEO systems are definitely interesting from a technology perspective. I think that we've been pretty consistent and haven't really seen any facts to the contrary that one of the challenges with a lower orbit system is the geographic distribution of bandwidth. And just from the amount of time we've spent on how we expect to use it with our satellites, we think that geographic distribution's really important. The -- I mean, it's just like -- imagine that you're a cellular operator and you can't really control the geographic distribution of your bandwidth and you have to fairly uniformly spread it over -- around the world. I mean, that -- that's definitely a disadvantage relative to people that can employ their assets in a more targeted way. So what I think at the end is that, while lower-orbit systems will have lower latency then we would, then we'll be able to use our assets to deliver higher speeds and more bandwidth for a comparable cost than that ultimately. That's going to be more valuable for a lot of the markets that we're serving that -- we do know that some fraction of bandwidth, generally less than 10% is more latency sensitive, and one of the other things that we're doing to try to serve that part of the market is to partner with other terrestrial services, where we can use hybrid approaches to deliver some combination of really high speeds and high volume over satellite, coupled with lower latency for those services that need it over terrestrial. I think we're going to compete well. I don't think it's a zero-sum game so I'm not predicting the failure of OneWeb. I just think that it's not going to bring an amount of bandwidth to the markets that we serve that's going to disrupt our ability to compete in those markets.

And our next question comes from the line of Andrew Spinola with Wells Fargo.

Shawn, I wanted to ask you about the accounting for these new ViaSat-2 plans. With the 3-month promotion at a lower price, are you recognizing the revenue from those subscribers at the full rate from day 1. And I guess the -- or as you -- and then somehow amortizing that discount over time? Or what sort of impact can we expect after the 3-month period from these new subscribers and did we see the ARPU boost from device at 2 subs in the current quarter?

Okay. So, a couple dynamics happening there. So, one, I would say that in general what you tend to see is kind of a blended effect of those early price points are a little bit skewed lower based on some of this promotional, and they -- we will see over time it's skewed upward. I would say that's pretty -- you should expect to see that. And then when we look...

So, Shawn -- so this is Rick. So, Shawn, in other words, we don't amortize that discount. We let it show up in the period that it's -- that the transaction occurs.

Pretty much. That's pretty kind of pretty much the end result that happens subject to some other considerations that we look at with the customer on a contract [or non-contract]. But at the end of the day, that's a pretty good approximation.

Okay. So, to just to sort of break it down, those 40-plus thousand subs, you're not seeing the full impact of the price points that they will be stepped up to after the 3 months in the December quarter? Is that fair?

Yes. I would say that's fair.

Okay. And the second question from me is, if you're willing, could you maybe talk from a high level about the EBITDA trajectory of the Commercial Networks business from fiscal '18 to fiscal '19, i.e., what your expecting from the step-down of ViaSat-3 expenses and then maybe particularly also what uptick could you potentially expect from the aircraft terminals?

So, I mean, I guess I would say without giving a straight number is that when you look at the R&D tick-down, we're going to see that tick down kind of quarter-over-quarter all the way through to the end of the year. So I think that if you look at a baseline of R&D spend closer to the percentage of revenues that we had early '17, you'll see a pretty significant savings coming into '19. And then you did hit on the other driver to EBITDA next year is really around the commercial air terminal. And that's definitely going to contribute some additional increases. So I think the pace of those we've talked about a little bit should be significant to next year. So I think you could see a pretty significant improvement in EBITDA. And we can give you guys some additional color to that as we head into next year -- or into next call.

And just to finish up on that point, the hardware that you ship is sold at a profit? Or is it breakeven?

Oh, no, it's certainly sold at a profit.

And our next question comes from the line of Simon Flannery with Morgan Stanley.

Mark, can you talk a little bit more about the broadband rollout with ViaSat-2? How are you thinking about your addressable market, given the greater speeds and capacity here? Are you going to really address the same sort of target markets that you went to with ViaSat-1? Or is there going to be an expansion in that footprint that you hit? And any color you could give us around the shaping of the broadband adds. When does it turn positive? And when does it really start ramping in your view as you offset the churn on your -- on ViaSat-1 versus the adds on ViaSat-2 as you scale that up. And then, Shawn, just a clarification. You talked about the margins in the initial quarter looking more like ViaSat-1 after a year. And just to check the math, are we in that kind of 19%, 20% EBITDA margin range? Is that the right ballpark?

Okay. Okay. So let me start with the first two. Okay, first in terms of target market. I think -- so our approach to marketing, I'd say, is we're working to become more granular, right. And I think that the best places for us to sell are for people who want to rely on the Internet more and more for video entertainment, so that's what we're targeting are people who want higher speed so that they can do video streaming and they can do enough of it to -- because they're using Netflix or some other similar service. So we'd like to target people who are doing that, but we'd also mostly like to target the people who are doing that and don't have access to, say, 25 or 50 megabit per second cable. And so that means really trying to better understand the intersections of those 2. And that would be, I think, just as an example -- and I just use this because I use -- did a bunch of it, but for instance, Craig Moffett when he looks at each of the cable companies, will look at what fraction of each cable company's footprint overlaps different competing services. For instance, how much of it is a DSL? Or how much of it is U-verse fiber-to-the-node? Or how much is fiber-to-the-home, right? So what we're trying to do is to market more to those intersections, those other places where we're trying to initially match all our supply of bandwidth to where we think there'll be demand. Now, I think when you look at from a macro perspective, the things that we think about are where do the speeds that we're marketing fit in the sort of the overall speeds that are available in the U.S., and one of the things we've shown in the past is 25 megabits is kind of in the -- right around the median speed in the U.S. maybe just a little bit worse than median, but when we started with ViaSat-1 12 megabits was kind of median speed. And now we'll have 50 and 100 megabit per second plans. So overall, from a speed perspective, we think we're -- we've increased our addressable market, but what we’re trying to do is be a little more precise in the way we do our actual advertising and promotions and in some of the channels that we're trying to engage to better target those intersections, okay. On the shape of the ramp, I'm sorry...

Yes. (inaudible)

So on the shape of the ramp, I think the main thing that we tried to get across and our thinking is that we're really looking at the revenue growth trajectory as opposed to the subscriber growth trajectory, and part of that is we're trying to understand what the desirability is under different plans. And there's definitely -- it's not exactly linear, but there's certainly a proportional or fairly linear relationship between the speeds that we offer, the price of the plans and the amount of bandwidth that's consumed by them. So think about it as we have somewhat different yields, but if we can get a comparable yield that is dollars per gigabit by selling 100 megabit plans as opposed to 25 megabit plans, hey, we would like to do that, right. So that -- but because the price point's different, that we just can't sell as many of the 100 megabit ones as 25. And so we're going to look at what that mix is. I think we learned a pretty fair amount from the test marketing, but we couldn't test market 1,500 meg plans because the satellite didn't support those. So I think we're going to defer a little bit the actual subscriber numbers until we get some experience with the -- with actual marketing, which will start by the time we report next quarter, we'll have probably a full month of activity with all the plans in all the markets.

Okay, and to your last question on Satellite Services EBITDA, so you were -- that range of 19% to 20% looking at kind of the first full quarter of ViaSat-2 in service, that's a good range to start from, and then we'll increase sequentially each quarter from there.

And our next question comes from the line of Phil Cusick with JPMorgan.

A little bit following up on Simon's question. For ViaSat-2, I like the unlimited plans. Do you, Mark, expect to sell only unlimited going forward? How do you think about that?

Yes, that's what the plan is. Clearly, there's a lot of demand for those. We can see that in the way people trade old plans, even ones with the soft caps or the very high caps for unlimited. So that's a big part of what we're trying to understand is to be able to get a sense on the amount of bandwidth it takes to fulfill those on the price points and yields, but overall, well, when you couple the unlimited with the video streaming management, it looks like a good way to go, and that's our plan going forward.

And then, can you give us an update on ViaSat-3? What are you thinking at this point in terms of timing and capacity? Anything changing?

No, so, it's -- we're -- the main thing that -- and it's reflected in the financials is that we're on the transition from testing, it's basically engineering model testing and prototype testing to transitioning into building the flight hardware. And we're getting more and more of the subsystems and modules into that. I think you've done a lot of testing on the preflight hardware. But the -- basically, in general, for these satellite programs, the largest source of uncertainty is in the integration phase, and so that's still ahead of us. And so that's where we'll really figure out what the deployment schedule is. We're also looking at different launch campaign options with the prospect of reducing the orbit-raising time, which was a big factor in ViaSat-2's in-service date. So right now, I think that the second half of 2020 is a reasonable range to think about for when that would come into service. And in terms of the technical capabilities, nothing's really changed. I mean, we've done a lot of -- lots and lots of risk-reduction and integration testing, and so far, things still look really promising in terms of its technical and performance capabilities.

And if you could -- this is -- I know you're just testing the unlimited plans now on ViaSat-2. But if you could push yourself forward 3 years and think about what those ViaSat-3 plans might look like, how do you imagine those evolving?

I think that's clearly what people want. What we would expect is that we would be able to deliver those with more of them tending towards the higher speeds and probably being able to push up -- assuming market values this, that these plans are in some way more valuable, or give us a bigger market, which everyone needs to increase the take rates, but we may end up with higher streaming speeds to go along with them. You could think about moving the streaming speeds up a notch, as an example, as we do that as well. I think those are kind of the general trends.

And we do have time for one last caller. And our next question comes from the line of Chris Quilty with Quilty Analytics.

Shawn, thank you for that very detailed breakdown on the selling cost and whatnot, except I'm bad at math, so can you guide me a little more specifically where you think the service margins will go once ViaSat-2 subs kick in? Is it kind of like the 2014 level or better or a little less?

So what I would say is, is we wanted to kind of give you guys where we thought kind of the first full quarter or our starting point was. Because it's not going to be as low as it was when we started ViaSat-1. And then I would say you should see us continue to increase quarter-over-quarter sequentially, and I'd expect us to be able to do as good or better than we did on ViaSat-1, ultimately.

Okay, got it. So it looks like we're poised to get some double-digit Pentagon budget increases. Mark, is there any particular program or general area where you think you might be able to just -- to benefit disproportionately either because there was unavailable budget or things that might be money related?

I think that I would say that, rather than just looking at the budget, I think that the things that are -- should be most impactful for us is to go back and look at, for instance, this Army tactical modernization report that the Army just filed, I think about a week ago. And that was really more focused on capability gaps and I think that, rather than thinking about it as, okay, people have more money to buy the things that they already buy from us, I think what we're really looking at is taking the capabilities we have and curing those gaps. And if we -- for instance, if the mainstream Army finds out we closed some of those gaps, that means our market's going to be much bigger. It's going to grow. Things will be better than just the growth in the budget. Likewise, if we could do similar things with certain Navy applications or Air Force applications. So I think that, that sort of gating step is to deal with the demand model and the capability gap model. I think that what we've seen is often the arguments when you feel those capability gaps are so compelling that it makes for -- that people find the budget one way or another. I think this makes it a little easier to find budget, but I think the big issue for us is really getting people to test and understand the functional value and performance value that we bring.

Okay. And one final quick question on ViaSat-2 major component of that, I think, is the transatlantic for in-flight connectivity. Do you feel like you've got orders in the book already where that reflects an element of the customers' plans or source selection on ViaSat-2? Or do you think any orders related to transatlantic are yet to come?

On the transatlantic, I think we've got some of them. I think that there are more opportunities to come, but I think we've got some that already we intend to use it. I think you'll see first transatlantic flights from some of our customers immediately once we bring up the aero service, which should be just shortly after we bring up the residential service. And I will let those carriers talk about it because I think they'd like to make kind of a big deal about it, but we've already got customers that are counting on that already.

And that concludes our Q&A session for today. I'd like to return the call to management for any closing remarks.

Okay. So, I think that concludes our call for this afternoon. Thanks a lot everybody for your time and interest. And look forward to talking to you again next quarter.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a great day.