Viasat Inc (VSAT) 2016 Q3 法說會逐字稿

Welcome to ViaSat's fiscal year 2016 third quarter earnings conference call. Your host for today's call is Mark Dankberg; Chairman and Chief Executive Officer. You may proceed, Mr. Dankberg.

Thanks. Good afternoon, everybody. This is Mark Dankberg; Chairman and CEO; and I have got with me Rick Baldridge, our President and Chief Operating Officer; Shawn Duffy our CFO; and Keven Lippert, our General Counsel. Before we start, Keven 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 10K and Form 10Q. Copies are available from the SEC or from our website. With that said, back to you, Mark.

Okay, thanks. So we will be referring to the slides that are available over the web, and I will start with some highlights and some news, and then Shawn will discuss our consolidated and second level financial results. Then I will give some additional details and some color on our strategic initiatives and thinking. Then we will summarize our outlook and take questions. So besides our earnings we have got news on three important new topics. They are an update to revise that two launch plan, starting construction on the ViaSat-3 global compilation, and a new strategic partnership in Europe with Eutelsat.. So I will start with the financial highlights and then come back to the other three.

In our results, show we are making good progress on increasing the value our ViaSat-1 band in the Satellite Services segment, and increasingly in our government segment too. Keep this in mind as we move later to the ViaSat-2 and -3 discussions. Our Satellite Services segment earned record revenues in the third quarter of $141 million. That is up 14% year-over-year, and grew adjusted EBITDA 33%, $63.5 million.

Government Systems also earned record revenues in the third quarter of $151 million, up 15% year-over-year; and grew adjusted EBITDA 16% to $37.5 million on a year-to-date basis, excluding the non-recurring benefit of last year's Loral litigation. Satellite Services adjusted EBITDA was up 55% year-over-year to $180 million; and on a year-to-date basis, government adjusted EBITDA was up almost 20% to about $100 million. Company-wide adjusted EBITDA for the third quarter was pretty much flat year-over-year on a 2% increase in revenues, and that is due to unfavorable comparisons to last year, as the NBN infrastructure project in Australia winds down, and also on increased R&D spending especially for ViaSat-3. Year-to-date company-wide revenues are up 5% year-over-year and adjusted EBITDA is up 16%, excluding the benefits of the Loral settlement on the same factors.

So earnings growth and margin expansion in the Satellite Services and Government segment are driven by higher value applications of our bandwidth. This includes steady growth in government and commercial air, mobile broadband, consumer ARPU gains, cost reductions due to scale, and improved network efficiencies. While a large majority of our bandwidth is already generating revenue, we have been quite successful in migrating those bandwidth usages to higher value applications and service plans. We still have quite a bit of room for sustained earnings growth and I will come back to that later.

We are just at the very early stages of applying ViaSat-1 to government applications, but it should be clear that producing the world's most cost-effective satellite bandwidth creates fabulous fuel for profitable growth. So these three new strategic announcements are good news for sustaining our long-standing historical EBITDA growth rates. First, while we still have great confidence in SpaceX, it is clear that a few recent challenges are stressing that Falcon manifest. So based on our ViaSat-1 actuals, we anticipate ViaSat-2 would generate about ten times the revenue of a typical commercial satellite. Figure around $40 million to $50 million a month in steady state. So we are extraordinarily sensitive to an on-schedule risk.

Given that, we and SpaceX have agreed to move our existing contract to a future ViaSat-3 satellite, and we have added an option with them for another launch as well. For ViaSat-2, we executed a contract with Rion for launch in the first quarter of calendar 2017, as well as for a launch of one of the ViaSat-3 satellites. So all things considered, our Rion launch builds confidence in our target mid-calendar 2017 service launch provides that too, and it also preserves our total ViaSat-3 capital budget -- I am sorry, ViaSat-2 capital budget.

Then as we suggested last quarter, the pieces have fallen into place to start construction on the first two ViaSat-3 satellites with Boeing Satellite Systems. The first one will be the follow-on to ViaSat-2, with about four times or better bandwidth productivity, and covering a third of the world including all the Americas. We are targeting ViaSat-3 in service about two to three years after ViaSat-2. The second ViaSat-3 will cover the Europe, Middle East, Africa region and launch in 2020. And that leads to our third important strategic announcement; we have reached agreement with Eutelsat to form a new retail consumer services entity in Europe, which will be 51% controlled by ViaSat and initially leveraging their existing KaSat Satellite which already uses our network infrastructure. Along with that, we will acquire 49% of the existing wholesale KaSat business for about EUR130 million.

So obviously this is a great way for us to translate the WildBlue device playbook that we used in the U.S., but this time with an enduring relationship with a very strong European partner, and an even bigger pay-off to the fabulous bandwidth productivity to the ViaSat-3 class satellites. We have a long and successful history with Eutelsat and we are extremely pleased to be making our first global expansion with them. We still have a lot of work ahead of us to complete the new ventures, but we have already accomplished a lot with the execution of this framework agreement. So with that I will turn it over to Shawn for more detail on the financials.

Thanks, Mark. Slide 5 shows revenue and adjusted EBITDA performance for the third quarter of fiscal 2016 compared to the same period last year. Revenues were up 2% year-over-year through strong growth in Satellite Services and Government Systems, offsetting declines in Commercial Networks. In Satellite Services, our segment revenues continue to grow, up 14% to record levels once again this quarter. In addition to strong revenue growth, we are seeing the continued benefits of the operating leverage of the service business, with adjusted EBITDA increasing over 33% year-over-year. Top line growth across -- occurred across all our market sectors with residential broadband offerings and in-flight conductivity being the primary drivers.

ViaSat's commitment to delivering quality broadband service offerings since the onset resulted in another quarter for residential R2 growth, up 7% year-over-year to $56.74 per month. This level of ARPU represents a $12 or 27% increase from when we first launched ViaSat-1. Turning to Commercial Air, during our third quarter we had an average of 437 aircraft in service and saw an increase in average revenues per aircraft, which allowed our Commercial Air revenue to more than double from the same period last year. All of these factors, along with the continued operating scale economics and lower advertising and marketing costs, generated year-over-year increase in segment-adjusted EBITDA of $16 million.

Before I move to Government and Commercial segment results, I wanted to briefly highlight some of our key Q3 satellite service metrics. Residential subscriber counts were just over 687,000 by quarter end, just lightly up from the previous quarter and up over 12,000 year-over-year. Even though we are at or near full capacity on the vast majority of our ViaSat-1 beams. And to clarify, when we say we are at capacity can change, this is related to the capacity we have currently dedicated to our fixed consumer business, and not the capacity we have set aside for the future growth of our Commercial Air and Government Mobility services.

Our Q3 returns improved once again, sequentially down from Q2. And subscriber acquisition costs were down materially, compared to both the previous quarter and the year ago period, due to higher utilization of (inaudible). (Lost audio).

Ladies and gentlemen, please stand by. Your conference call will begin momentarily. Again, please remain on the line. Your conference call will begin momentarily. Thank you.

Again, please remain on the line. Your conference call will resume momentarily. Thank you.

Speakers, welcome back. Please continue.

We will pick up to where we were. In commercial network, our revenues were down year-over-year as we complete the NBN co-infrastructure project as well as declines in terminal sales, and our system products sales. However, with an expected service launch in April 2016, we began seeing new consumer terminal orders from NBN at the end of the third quarter. As Mark mentioned earlier, our ViaSat-3 activities are well underway, which drove our Q3 segment R&D expenses to double the levels incurred last year. I will provide a bit more context around these impacts later, but in the near term we do expect to see a pretty significant uptick in development activities supporting our ViaSat-3 satellites, as we close out FY2016, and continuing for the next several quarters. slide 6 shows revenue and adjusted EBITDA performance for fiscal 2016 year-to-date period compared to the same period last year.

Speakers, welcome back. Please continue. Comparisons with the prior period, however, are excused by the aggregate $40 million earnings impact of the Loral settlement realized in Q2 fiscal 2015. Just to recap, in the second quarter of last year, we booked a one-time benefit to product revenues of $21 million; and a one-time G&A credit of $19 million as a result of the settlement. I will not spend much time on this slide, as many of the revenue drivers are the same as those in the quarterly comparison on the previous slide, but I would like to highlight the 16% growth in adjusted EBITDA year-to-date to $250 million, excluding the Q2 FY 2015 Loral impact.

Compared to a relatively flat year-over-year Q3, where the rising impact of our ViaSat-3 R&D activity tempered our EBITDA results. For further context, if we look at the core business without the Q2 FY 2015 Loral settlement, and our year-over-year R&D acceleration, our existing business grew adjusted EBITDA on a year-to-date basis compared to last year by approximately 26%. Much of this continuing growth is driven by top-line R2 growth. and expanding leverage working in our Satellite Services, where our year-to-date adjusted EBITDA grew by more than 55%. Turning to the next slide, in both the quarter and year-to-date periods, operating income was lower due to higher depreciation and amortization expenses on essentially flat Q3 adjusted EBITDA and slightly lower year-to-date adjusted EBITDA, due to the recent increases in R&D expenses I mentioned earlier.

Turning to taxes, as you may know, the Protecting Americans from Tax Hikes Act was enacted in December 2015, which permanently ended the federal R&D tax credit retroactive to the beginning of our fiscal Q4 2015. The $6.4 million impact of this legislation from a net tax expense position for a net tax benefit for further quarter and year-to-date periods. When looking at the comparative Q3 period, it is important to keep in mind that our Q3 last year also included a fed R&D credit restatement touch-up benefit of approximately $8 million.

So year-over-year, our benefit is approximately $2 million less. As a result of the points we have just discussed, non-GAAP net income for the quarter and year-to-date was down $4.1 million and $12 million respectively. The earnings per share figures follow the shame relationship, with a slight variance for the change in shares outstanding. However, it is worth summarizing that the impact of the Loral settlement in the year-to-date figures for last fiscal year represented a benefit of $0.50 per share.

So excluding that, as well as the R&D uptick and fed R&D credit impact, our non-GAAP diluted EPS generated from our core business actually increased 100% year-over-year. Looking at our cash flows, you can see that our year-to-date cash flows from operation, just below last year's level by about $52 million, which was mostly due to large swings in working capital. The increase in working capital was driven primarily by prior inventory levels and our government product lines, a decrease in customer advances, and lower accrued interest due to the normal timing of those interest payments.

It is worth noting that our Q3 quarter cash collection activities were very strong, with operating cash flows generating over 50% of the year-to-date activities. And our trailing 12-month cash flow from operations is trending at approximately $300 million which is funding a significant portion of our next-generation satellite investments. Our year-to-date capital expenditures and investment for fiscal 2016 decreased by about $27 million, due to our acquisition of NetNearU last year for $56 million. Partially offset by capital and software investment and other Q3 investments (indiscernible).

So we ended the quarter with $200 million outstanding on our $500 million revolver; and $161 million outstanding on our $5 million and $25 million (indiscernible) loan commitment. Our net leverage increased slightly to 2.6 times trailing 12-month adjusted EBITDA due to the higher debt balances. Yet our Q3 liquidity position remains very strong. A bit later, I will spend some time recapping our satellite CapEx profile as we look outward as well as see how that shapes up into our overall debt and leverage outlet. So with that I will turn it back over to you, Mark.

Okay. Thanks a lot, Shawn. I think we may not have our slides up at this point. They should be back up. If not, we will posting them and you will be able to reconcile them. So they are online, so you can get them, and I will try to describe what is up there as well. At this point I will try to put some color around our strategy and our outlook, given all that has happened leading up to today's announcements. We have got a chart on our capacity expansion schedule; and our launch window with Arion for ViaSat-2 is first quarter 2017. Satellite construction is proceeding well, given the overall progress date.

We have more visibility into the post-launch activities leading to the beginning of service. While there is still some blips, we anticipate the ViaSat-2 in-service date to be mid-calendar 2017, which is about where we have been for the last year or so. So net-net, a little slippage in the actual launch schedule for more certainty, but without materially impacting the timing of revenue generation for the satellite. And remember ViaSat-2 has about double the bandwidth economics to ViaSat-1, so that creates a lot of opportunity for growth. When you look at the ViaSat-3 platform, the economic gains of our prior generations are even more compelling. It is also very important to note that we are also gaining efficiencies in delivery time. The gap between ViaSat-3 Americas in-service, and ViaSat-2 in-service, will be around two years or so; that compares to over five years between ViaSat-1 and ViaSat-2.

So obviously we would have benefited from launching ViaSat-2 quite a bit earlier, but the technology was not there yet. We have been investing steadily for years in the successor to ViaSat-2, so this time we are aiming to cut that gap between satellites roughly in half while still getting an even bigger boost in productivity. Then we are planning to launch ViaSat-3 Americas in mid-calendar 2019, and ViaSat-3 in the Europe, Middle East, Africa region planned for early 2020.

We and Eutelsat are still studying and evaluating options for interim capacity earlier than that. The coverage and flexibility of the ViaSat-3 class satellites enable much greater opportunities than merely (inaudible) markets, and we have got another slide to help illustrate that. Okay, So they are back. Good.

So this slide shows the planned initial ViaSat-3 constellation, and the checkmarks show the first two are now underway. That column chart is just a reminder of how incredibly productive these satellites are in terms of simple bandwidth economics, and that basically means how many gigabits you get per million dollars. With capital investment, the representative bean maps show how extensive the coverage is. Essentially two-thirds of the world with the first two satellites. And it illustrates that we have the reach and economics to serve vast emerging markets in South America, Africa, Middle East, and western Asia. And we also have the ability to dynamically manage exposure to those markets based on demand and business case.

Each ViaSat-3 satellite is anticipated to have as much bandwidth as all the rest of the satellites in the world combined. And that includes all of the (indiscernible) satellites that are now under construction. So we think we will have the economics to profitably address the most attractive market anywhere in those footprints. That purple and white map helps illustrate the importance of the dynamic coverage flexibility in the ViaSat-3 class satellites. About 95% of the world's population lives on the land area that is shown in white. Only 5% of the population lives in the purple area. And then for comparison, 5% of the population also lives in the land area that is shown in tiny little red dots. It has been shown over and over that while there is proportionally a higher penetration of satellite services in the lower-poplutaion density regions, by far the highest amount of absolute demand is in the highest density population regions.

So one of our key design criteria for ViaSat-3 was to have the flexibility to allocate bandwidth resources efficiently over very broad geographic areas while still being able to focus it on very small ones. It is a very unique capability in the satellite industry. So this next slide -- pretty simple but effective illustration of how we use advances in bandwidth productivity as enabled by our very unique vertically-integrated technology to both compete in the market and generate compelling financial returns. The chart shows time on the horizontal axis and EBITDA yield in dollars per satellite. For ViaSat-1, the useful bandwidth on the two WildBlue satellites were running about $200 million in annual revenue, and about $80 million in annual adjusted EBITDA combined.

So that was a pretty nicely profitable business for us. But compared to those, ViaSat-1 offered almost a 20x gain in bandwidth economics. So we used some of that productivity to give better service to our customers, and we turned a lot of it into earnings. Our current run-rate is almost $250 million a year in Satellite Services adjusted EBITDA, and it is still rising, and about 90% of that is attributable to ViaSat-1. So that is almost six times the yield of each of WildBlue's first two satellites in earnings for service that subscribers are much happier with, and gives us a bigger addressable market. Looking ahead, ViaSat-2 is about another factor of 2 gain in that productivity. And in theory, using the actual bandwidth yields that we are getting on ViaSat-1 which are again still rising, we could target EBITDA per satellite at as much as $500 million a year.

But as with ViaSat-1, we expect to give a lot of that productivity gain to our customers, and use some to improve our margins and shareholder returns, and finance even better satellites. The productivity growth on ViaSat-3 is even more striking, and we believe will create substantial separation from satellite competitors and enable us to compete more effectively in broader markets. The table at the right lists some of the key factors for ultimately determining how productivity gains flow into operating margins. Some of it is eaten up by bandwidth deflation. That is customers' expectations of getting more bandwidth each year for about the same price. We will also trade more bandwidth to get higher growth rates.

For instance, we can offer virtually unlimited plans that use more bandwidth but are much more attractive to customers and let us add subscribers faster. In that case we are competing, to some extent, with bandwidth productivity gains on the edges of terrestrial markets. But other factors work to improve our margins. As our business scales, we get more and more cost efficiencies. Some of our higher value customers are on variable service plans where they get more value by consuming more bandwidth, and they pay higher subscription fees. For example, that is true for Aeromobile and Government. Increasing our speeds even for the same volume of bandwidth also increases value, and we are seeing that effect now with our SpeedBoost offering.

The bandwidth in the flexibility of ViaSat-2 and -3 will allow us to offer peak speeds of 100 megabits per second; and with ViaSat-3, bursts in the gigabits-per-second range, creating more opportunities for higher value. We can also use the coverage and the bandwidth flexibility of ViaSat-2 and -3 to move bandwidth around to new geographic markets and new applications within those markets. For instance, we can move bandwidth across geographic or market applications that are different in peak busy hours, increase win-win situations with very attractive pricing in (inaudible) markets while still growing earnings. Aeromobile usage coordinated with residential is a good example of that effect, but there are many more. Of course as we give productivity gains on new satellites to our customers over time we reduce the EBITDA yield on our older satellites. The overall return on each satellite is based on the integration of the area under each curve, adjusted for the time, value of money.

We believe we have provided enough financial data to allow investors to model these effects and find that the returns on capital are very attractive. So our success in the commercial in-flight connectivity market has attracted a lot of attention, and we think it is worth focusing on for a few reasons. Existing conventional and lower-yield high capacity satellites provide a pretty poor in-flight connectivity experience based on a very visible and powerful contrast with what is possible with our R band with economics. Others just do not have the bandwidth productivity to enable a satisfying experience to a large number of passengers simultaneously on hundreds or thousands of planes at an affordable price. And streaming video is a great stress test that everybody understands and effectively separates out weak bandwidth. Nobody can fake a satisfactory streaming experience at scale.

Failure mode is just too obvious. You get suffering and frustration. Ultimately we believe we can achieve similar effects and contrasts with existing satellites in pretty much every other market we enter. Pretty much all those markets will compete with incumbent business models that are generally based on very limited usage of expensive bandwidth. That is true in commercial, in-flight connectivity, general aviation, maritime, oil and gas, government, you name it. Our bandwidth enables fundamentally different business models that just can not be replicated with small amounts of expensive bandwidth.

For instance, the current in-flight connectivity market is oriented around charging very few passengers very high prices for very limited service. The airlines have been okay with that until competitors such as JetBlue and Virgin America have divided passengers with free Wi-Fi, including free Netflix and Amazon Prime. That is just an astonishing difference to a passenger compared to paying $50 per flight for Wi-Fi that does not even support YouTube. At CES, the keynote speaker, Netflix CEO Reed Hastings, even included a reference to Netflix on Virgin America at 30,000 feet. JetBlue, who will offer free Wi-Fi, including Amazon Prime video on all its flights, has been the industry leader in growing and keeping passenger revenue. That is not surprising when survey after survey identifies in-flight connectivity as the single most desired passenger amenity.

Of course, JetBlue is already known for its focus on passenger experience. We think in-flight connectivity is a key part of that. And Vice President Jamie Perry captured that perfectly when he points out, why would anybody want to have Wi-Fi on a plane that the passengers do not use. So the key to changing the business model with high-quality, high engagement in-flight connectivity is that it also changes the flow of money. The initial model was for a Wi-Fi provider to charge passengers as much as possible and then share that revenue with the airline.

Not only did that make for grumpy passengers, but it almost completely forecloses the potential for internet and media companies to use a positive connectivity experience as a way to engage with valuable passengers. Clearly the more passengers that use Wi-Fi, the more valuable the opportunity. And there is nothing that gets the passenger's attention more than tie-ins with high profile internet and media entertainment, social media, and other apps that are part of daily life. So last quarter, we began to see the power of that effect when JetBlue's Amazon Prime promotion was introduced.

We will present more data on this soon with JetBlue and Amazon. But for now we will just say that Amazon has gotten striking gains in the popularity of Prime video compared to other forms of entertainment. That makes to a win-win/win-win situation for the airline, the passengers, the sponsoring internet companies, and ViaSat. It is a powerfully different business model that simply can not be produced without our bandwidth productivity and the sheer amount of bandwidth, no matter what the price is. Last year we said that 2015 would be the year that airlines began to understand the implications of our in-flight model. We think that definitely happened. And we think this will be the year that airlines begin to take action. So stay tuned for that.

So this next chart shows where we have achieved strong growth in adjusted EBITDA this year in Satellite Services and Government Business segments, even with close constraints on bandwidth. We think we can continue that in the next two quarters leading up to the ViaSat-2 launch. This chart shows the main ingredients. We are aiming for sustained consumer ARPU growth through higher value service plans including our recent 25 megabit speed booth, as well as new plans for businesses, customers, and additional ancillary services.

We are seeing strong commercial aerogrowth as we add more airplanes, serve more people per plane, and engage those passengers in higher value activities like streaming video. In our government segment, one key element of recent growth has been in selling and installing Ka-band and KA/KU band terminals on more and more government aircraft. And we will begin to see more and more service revenue this calendar year plus aircraft service. That revenue will show up in our government segment. We are also seeing good demand in growth opportunities in our (inaudible) information assurance businesses.

The next slide gives a little more color on our Eutelsat strategic partnership. As I mentioned before, the purpose is to combined ViaSat's expertise with the retail services in satellite technology with Eutelsat's leading position in broadband services in Europe; and better leverage the synergies between satellite TV and satellite broadband in Europe as has occurred in the U.S. And it is also a good example of our strategy to partner with leading companies in each target global market, versus trying to invade like just a purely American entrant. So our partnership will start by leveraging Eutelsat's existing KaSat and the coverage map is a good indicator of how we can both benefit from the integration of KaSat with our current Ka-Bands. And with ViaSat-2. Whole leverage stays as existing wholesale business space in residential and roaming agreements for mobility.

We will also start a new retail entity that will be led by ViaSat to more effectively monetize the existing KaSat bandwidth. We intend to use the retail business on KA-SAT to generate a running start for ViaSat-3, and to help fund a portion of the new satellite. Our Eutelsat partnership is only for the European market and we still have to more work to do with Eutelsat to turn our existing framework agreement into the final operating agreements; but it should create a very cost-effective path for us to build on the existing space and ground infrastructure, our existing wholesale business and augment that with a higher value retail service. We are both pleased with the joint progress and the basic idea is to recreate what we did in the U.S. with the legacy WildBlue business, and then turn that into an enduring profitable growth business.

Before Mark provides some final comments, I wanted to touch in on some points surrounding our ViaSat-3 program and our related CapEx profile. A couple of items to note are; one, the upper right chart includes not only our ViaSat-3 program activities, but also completion of our ViaSat-2 satellite project. Additionally, the satellite CapEx included in the chart reflects the full network build-out. In other words, in addition to the satellite construction launch and launch insurance costs, it includes all capital costs for the construction of our ground and infrastructure network. As we look out over the next three to four years, you can see the satellite CapEx across all programs averages around $275 million per year and we have quite a few levers around pacing our satellite build and to even a greater extent the build-out of the ground infrastructure. When we think about our project spending profile we have quite a bit of flexibility.

We have approximately $570 million of liquidity under existing debt facilities, and our core business continues to grow and generate significant levels of cash. My earlier comments surrounding our cash from operation trends, remember we generated about $300 million over the past 12 months which effectively entirely funds this level of satellite CapEx. Additionally, as we look to our leverage levels reflected in the chart on the bottom right, we see strong de-levering trends over the past several years and we are currently in a very low leverage position, which gives us a lot of flexibility for additional funding sources over the next few years.

Looking out, we expect to hit our peak leverage level in around fiscal 2018 or 2019, which includes the incremental fueling of ViaSat-2's service launch and related subscriber acquisition cost. So all-in, we expect to stay within a comfortable leverage level throughout our ViaSat-3 satellite projects with some peaks in 2017 and 2019, and additional levers to pace our network activity based on affordability. With that I will turn it back to you, Mark.

Okay, thanks. So here I would like to kind of just deal with what we have talked about so far down to the effects that we expect on our financial outlook, the last quarter of fiscal 2016 and our upcoming fiscal 2017. We see good growth prospects in Satellite Services and Government Systems. Clearly our future is in services more than in selling technology and products in our commercial segment. NBN infrastructure is winding down and there will be some unfavorable comparison to net segment year-over-year. And also year-to-date government orders -- as Shawn mentioned, have been stretched out a little bit due to the extended defense budget continuing resolution. That is kind of a sector-wide issue. So far, it seems like our own government business has been resilient enough to overcome that, but we always like to point out that quarter-to-quarter timing risks are always there for new orders in our government business. Our adjusted EBITDA for fiscal 2017 may be affected -- or will be affected by the fact that we are building the ViaSat-3 payloads ourselves. Overall, the cash costs for the two ViaSat-3 satellites are planned to be quite favorable.

On average the same or less than ViaSat-2 ,even with targets better than four times the bandwidth. But as we have been working through the accounting rules on the development and construction project, we see we are going to end up expensing more of the ViaSat-3 projects and capitalizing less of it. From a management perspective it does not change the cash profile; but from an adjusted EBITDA perspective, it will result in a significant growth in our R&D expenses in our fiscal 2017. The incremental R&D expenses for ViaSat-3 for fiscal year 2017 compared to fiscal year 2016 are estimated at an increase of about $30 million.

And that R&D outlook is baked into the capital spending outlook Shawn just described. The upshot is that we are estimating adjusted EBITDA for fiscal 2016 at about $345 million to $350 million, based on a solid fourth quarter. Adjusted EBITDA for fiscal 2017 is estimated in the $380 million to $390 million range. For reference, those estimates are about the same as the current consensus of eight analysts for fiscal 2016, and about $10 million to $20 million less than their consensus for fiscal 2017. Actual results could be affected up or down by timing and specifics of new contract awards along the way, but the takeaway is that the improving economics in our Satellite Services businesses and our strong position in the Government segment, are support solid financial results; and we are reinvesting that in what we believe is a truly transformational opportunity for the Company.

I also wanted to point out we are working with additional potential strategic partners for both the Americas, and the European, Middle East, Africa for ViaSat-3. In the event we conclude those agreements, we anticipate they will benefit our earnings outlook or our capital outlays for both. We are aiming for ViaSat-2 to be in service around the end of the first quarter of fiscal 2018, so we anticipate good sustained growth in fiscal 2018 as well.

That is it for our prepared remarks and we will open it up for questions.

Thank you. (Operator Instructions). Our first question comes from Mike Crawford with B. Riley and Company. Your line is now open. Please go ahead.

Thank you. Mark, could you just try to illustrate a little bit about capacity issue, how much capacity of a satellite you would need to provide service of say a certain speed through a certain number of devices on, say, planes in an area?

Okay. So yes, it is -- what you need to do when you do that, is you have to take into account the route structure of the airline, because most of the airlines tend to have focal points in New York, Chicago, Boston -- New York, Chicago, Dallas, Denver, Los Angeles. As an example, Houston. It turns out that you need in each of those markets, at peak times, several hundreds of megabits per second to serve even less than 100 people per aircraft in those markets. So that, when you look at it, even lots of these high (indiscernible) satellites have less than 100 megabits per second in those markets. So it is only a fraction of what you would need even to support, say, less than 1,000 aircraft. So it is a lot.

Yes. Okay. That is helpful. Maybe take it another way. So -- let us say ViaSat-2 has 300 megabits per second capacity. With a satellite like that, what level of service you envision to provide, is there a cap on the amount of commercial aircraft you think you could provide in a given area with the ViaSat-2 architected service?

If you look at it -- one of the things we have said -- and remember, always what is really important to any users of these applications is peak, peak busy hour demand, not just the average demand. The peak demand is what drives it. And the peaks generally occur multiple times per day as you have waves of aircraft coming in and out of these major airline hubs. So we have said several times that -- let us say we have allocated on the order of 10ish% of the ViaSat-1 bandwidth to support hundreds of airplanes that we have now -- so call that 10ish gigabits, we expect to be able to support well over 1,000 aircraft,1,000 aircraft for ViaSat-2. But that means tens of gigabits per second, especially as we are seeing more usage with these streaming applications, Netflix and Amazon. So that what you are talking about, is tens of gigabits. But again, even if we applied only 10ish% of ViaSat-2 that is 30 gigabits of availability, which kind of dwarfs that amount. That would be available on these markets on any of these other satellites.

Okay. Thank you. And then just because the IOIC calculation is a little tough on the outside given that some of the revenue you generate from ViaSat-1 gets blown up in the Government Systems. And then you have this continued CPE and subscriber acquisition investment, could you give a rough break down -- would you estimate your ROIC on ViaSat-1 has been or will be, and the same for ViaSat-2 and ViaSat-3.

Yes, well, you know, what I say is we have done a -- how you would normally go back -- this is Rick -- and look at the investment thesis that you had to the IRR, which you can convert to ROIC payout. And we think that the ViaSat-1 has turned out as good or better than what we had anticipated to start with. But I think more than that, if you look at it like a project -- well, it is like that. If you look at it like a business, like you said, it is pretty hard. Until you get into a steady-state mode it is kind of hard to calculate an ROIC for business that you are reinvesting in customer acquisition and in the R&D for the next generation product.

So if we calculate it from an IRR standpoint, we are achieving really high IRR -- when I say "really high" kind of Benford Capital-ish IRR returns on these investments. So we expect to be somewhere between the famous chart that shows the DBS operators, and the cable guys and the wireless guys; we expect to be somewhere between the DBS guys and the cable guys on ROIC on our business, our ongoing business.

And that is between $20 million and $40 million, that range.

Yes.

Okay. Thank you.

Thank you. Our next question comes from the line of Andrew DeGasperi with Macquarie Capital. Your line is now open. Please go ahead.

Thank you. I wanted to get, maybe some additional color from either Mark or Shawn on why did you guys form a JV as opposed to maybe offering the service directly in Europe. Can you highlight why is it important to have a wholesale partner there? Thanks.

Okay. So, one, there are a number of benefits that we get by working with Eutelsat. One is Eutelsat has a very strong position in Europe. They have an inventory of orbital spots that are a good resource for expansion. They have strong regulatory and landing right positions throughout Europe. And as we have seen in the U.S., there is a very high congruence between the people who have satellite internet and people who have satellite TV. So working with the leading satellite internet -- satellite TV provider is a strong position. The other thing that worked really well for us in the U.S. was we acquired the WildBlue business, and it gave us a running start in our retail business, which has turned out to be a big driver of the financial success of ViaSat-1. We have such a higher proportion of retail customers, our ARPU has grown substantially.

So the big things that we get here with Eutelsat are the ability to start, creating a retail network around the existing KA-SAT satellite, to have that in place by the time we launch the new satellite, build on good orbital position and to work with them on better leveraging the synergies between satellite TV and satellite internet. So could we do those things without the joint venture? Yes, we could. We just feel like this is a much more expedient -- it is a good low-risk return trade-off, and with like working with Eutelsat. They are a good partner for us.

Great. Thank you.

Thank you. Our next question comes from the line of Chris Quilty with Raymond James. Your line is now open. Please go ahead.

Thanks. Just to follow up on the prior question, what specifically did you invest in, I think you mentioned $103 million in terms of assets or investment capital that you own in that joint venture?

Okay. So the number we put out is EUR130 million, and just to be clear, we have a framework agreement with Eutelsat. We will be creating two new entities. Essentially a retail entity which will be new, which we will have 51% of, and then a wholesale entity which will build around the existing KA-SAT, they will have 51% of that. So the EUR130 million essentially gets us 49% of the wholesale business, which is the existing KA-SAT business and infrastructure; and that includes satellite itself, all the ground network infrastructure, all of the networking equipment that goes into making that, as well as 49% of the cash flows associated with that business.

Got you. And for Shawn, how should we expect that to flow through the P&L and pop up on the balance sheet?

I would say that we are still in the early work, as Mark mentioned, of completing up the joint venture and closing the deal. So I would say we are going to update you guys as we get closer to all of the financial details. The best way to think about it is we are effectively going to get 50% economic benefits of the combined group.

Okay. And, Shawn, actually when the transcript cut out, you were just saying on the script about the Zack. So I think we missed the whole piece that followed that, which I think probably included the government piece. Could you just run through that quickly? Or circle back to it?

Sure. I can tell you from a Zack perspective, the Zack was down this quarter from last quarter's results from the higher period. We had a higher balanced of refurbished equipment that we used within our new phase. Then trending down, it went down again. As far as -- I am not sure exactly on the government systems specifically what you are looking for. We had very strong top line growth, very strong EBITDA performance growth. And a lot of that is driven across government, mobile broadband, tactical dataing. But we also had an uptick in the service base there, 17% year-over-year.

Service up 17%?

Yes, service revenue.

Okay. Also a question, Mark, you are building the payloads for ViaSat-3, if I am not incorrect. This is the first time you have ever built payloads. And so are you --

No.

No?

One of the things we have talked about over about the last two years is we have had a subcontract with Telus to do the Ka-band payloads and that is for cross links and feeder links on the Iridium next constellation. So that is pretty interesting, because there is close to 80 of those satellites. So it is really given us very good experience working with a satellite prime on getting space qualified. And we shipped already -- shipped the first launch worth of ship sets for that. And we are about a quarter of the way through that manufacturing process. And it is really relevant because the way that we structured our contract with Boeing, is essentially they are giving us the payload framework for us to attach our modules to. So it is actually kind of a direct -- think of it as a direct predecessor to what we are doing on ViaSat-3.

-- (inaudible) that is the people that are doing that are people that have done this before, when they were at Motorola. And we have continued to bring people with space experience into that facility.

So you are doing it in Carlsbad, or with where?

In Arizona.

Okay. And if you can just elaborate a little bit, with ViaSat-3, is that going to require separate antennas, either for the mobility applications or fixed site to deal with the high number of beams? I think you mentioned previously a thousand beams on those satellites?

No. So -- We talked about a progression in sort of beam counts from dozens, hundreds, to thousands. So think of it that way. But I think we also said that one of the cool things is that the terminals that we put out on ViaSat-2 will be forward-compatible for ViaSat-3. So the terminals themselves are basically indifferent to the next generation of satellites.

Okay. And with -- I will ask one final question and jump in the queue. Once you have this capacity online obviously, you have an existing base of business and consumer. You are also established in the aero portion. As you look out to some of the other traditional satcom applications out there, what else do you view as attractive? And do you intend to go after all these markets directly? Or does it make sense to co-opt some of the existing players and sign distribution agreements for certain applications, or even within the markets like aviation where you already have a foothold.

Yes. So the two -- I can think of three that we are pretty well-established in already. One is the commercial in-flight connectivity market. Number two is the government market. I mean, that one has been fabulously successful for us. We have a very, very high market share and we are pretty much creating that market. Then the third one is the general aviation market, especially in hiring the aircraft. One of the things that we have found in going after those markets -- and this is sort of what I was trying to point out in the in-flight connectivity market, is that the business models of the incumbent echo systems really do not -- did not want there to be lots of bandwidth at very low cost.

So that made it -- it has made it, I would say, a little more difficult for us to get into the markets. Taken longer. Ultimately we think it is going to be far more rewarding, because we have a prime position in those markets. But the fundamental issue is sort of changing the business case that a lot of it, we should have tons of bandwidth and it should be very low cost. So we expect to apply that to the Maritimes business. Maritimes business is a great opportunity for us, especially when we have that enormous amounts of ocean bandwidth, far more than has ever existed in the world before, and great flexibility to apply it to the right markets. We will see if we can work with existing distribution to accelerate that under the terms, or not.

Same thing is true for oil and gas. Lots of demand in the oil and gas market. Lots of demand in enterprise applications, in backhaul for Wi-Fi or cellular. Just think of any place where people think bandwidth is too expensive, that is a great target market for us. And that is almost every place.

Got you. So basically distribution strategy evolves over time?

Yes. We just do not want to find ourselves in a place where our distributors do not want to drive down air time prices. Think about it as guys who want to buy the same thing for less money. Probably not a bright place for us to go.

Great. Thank you.

Thank you. Our next question comes from the line of Andrew Spinola with Wells Fargo. Your line is now open. Please go ahead.

Thanks. I would like to ask specifically about your guidance for Q4 and fiscal 2017. It is pretty significant growth in EBITDA, both sequentially and year-over-year, considering the $30 million headwind you referenced. And I am just wondering, when I look across your segments, it looks like the biggest driver of that is got to be in ARPU and potentially, I guess, maybe planes coming online. Could you talk about the backlog of planes coming online, and your view on how ARPU can grow in Q4 and next year at these levels, continuing at these levels?

Okay. On the planes -- so think of it as multiple dimensions. One is adding planes. We have been doing that pretty steadily. And I think -- I am not going to comment right now on the backlog because we think -- stay tuned. We think it is going to grow pretty significantly. I think, though the other really important elements that we are seeing are more people per airplane. So that is steadily rising. It is just like adding more airplanes -- if you are not growing the penetration, the only way you are going to grow is by airlines.

But if you grow in penetration, that is a really good way to grow. And the other one is more bandwidth usage per person. And that is especially true with the in-flight streaming which is a very, very popular option. So that is a very good growth opportunity for us. I think people tend to underestimate that. On the ARPU side, the way we are getting growth there, is we are offering new plans. We have had, for instance, this 25 megabit per second SpeedBoost offering has been really popular. We just introduced that.

And then we have also been offering plans, if you look on our website, plans that let people use more bandwidth in off-hours. They are generally pretty attractive. They are a little higher priced but people get a lot more utility out of them. And then we are offering some enhanced home services care opportunities as well. Those are the things that will drive ARPU. And both of those -- all those things will be manifested in Satellite Services EBITDA. And then on the government side, probably the biggest driver for us will be as the aircraft that we have been outfitting with Ka-band terminals start using Ka-band bandwidth, that is going to be an important factor as well.

And you know, just like we did at the end of WildBlue, we actually were turning down customers and growing EBITDA and earnings just from a cost-efficiency standpoint, we are doing that over this next year to 18 months.

Understood. Could you possibly give us a general magnitude of how much your ARPU is in the aviation business maybe on a year-over-year comparison?

Well, the one thing we would say -- and we have used this comparison about a year and a half ago where we said, think of each airplane as like 100 subscribers, 100 consumers and it is actually -- even though consumer ARPU has risen since then the airplane usage has grown even faster. So we are not there yet but it is getting to be where you could think of it as more double. Double that. Think of it as closer to 150 to 200 passengers --

Got it.

-- 150 to 200 consumers, sorry.

Understood. Just a question on your CapEx outlook table -- or chart, rather in the presentation. The CapEx for ViaSat-2 that goes into fiscal 2018/2019/2020, is that the ground network investments? I guess the points of my question is, does this table include your terminal CapEx?

That is just the ground infrastructure and everything to bring the network live. We do have some additional CapEx for expansion of the network over that time for ViaSat-2. The CPE is kind of a separate value, so we kind of think of that as success-based CapEx, so that is not reflected in that chart.

Got it. Just one last quick one for me. The press release with the joint ventures with Eutelsat talks about trying to potentially procure incremental capacity in advance of ViaSat-3. I am not aware of any Ka-band capacity over in Europe. Are you considering using Ku-band capacity? Or how could you go about that?

We are just thinking about some -- think of it as quick reaction Ka-band capacity, which would be along the lines of -- although not necessarily this, but along the lines of the flexible broadband system that we are doing with Boeing using their 702 SPs. The construction schedule on those are pretty short. So that would be an example of the way we could bring in earned capacity.

It could be that. It could be a Ka-band payload on another satellite. It could be any one of those things.

Got it. Thank you very much.

Thank you. We have a follow-up question from Chris Quilty with Raymond James. Your line is now open. Please go ahead.

Thanks. Just a follow-up. You had mentioned that the switching from the Falcon Heavy to the Arion 5, your capital budget is still in place but generally an Arion 5 is going to cost you $100 million to $120 million more than what you were paying for the Falcon Heavy. Are there other puts and takes in the prams program where you were able to produce savings to maintain the budget?

I think Arion had a hole in their manifest so we were able to grab it. We think we got a good price. And the insurance cost; our estimates for the insurance between the two, given that there is such an early flight on Falcon Heavy was much higher, so we take into consideration the two factors were within the estimate that we had.

Got you. Also the press release you had out today mentioned something about integrating your Ku-band service with Rockwell Collins. Can you elaborate on that? Is that merging or selling the yonder business? Or what was that referencing?

It is just operational spike. It is just the way -- think of really what really has led the way on our in-flight connectivity for general aviation and commercial aero has been -- think of it as passenger cabin usage. So this is just an example of ways that we are looking into flight operational data, cockpit-type data. So that would just be an application on those same aircraft. We have been working with Rockwell in the past on outfitting business jets.

Got you. So is it a data-sharing-type agreement?

It is really just more on integrating the connectivity with cockpit and operational avionics data, as opposed to passenger cabin data.

Got you. Shawn, can you give the specific CFO and CapEx in the quarter?

The specific CapEx --

Cash flow from operations and CapEx?

Yes, absolutely. The cash flow from operations was about $112 million. And then if you are just looking at the all-in investing activity, that was about $165 million. And that includes all of the satellite, the start work on our -- our ViaSat-3 projects, the CPE, including capitalized interest.

But not obviously the Eutelsat agreement yet?

Right. We have not funded that yet.

Okay. And on the ViaSat-3 program, when do you begin to capitalize? Because normally -- you have already elaborated, you are not capitalizing your R&D costs, which normally would get folded into the program. What is the trigger where you initiate normal GAAP accounting for building a satellite?

It is pretty similar to other assets that you construct. And it really starts to tee in as you start to construct the asset and do some of the early quality and certain other works around integrating them into the satellite. So it just happens a little bit later in the process, than if you were to potential particularly do it with a third party.

I think so it is fair to say that We are following GAAP, so --

Oh, no, I know. But the way that you are separately manufacturing the payloads and the bus, apparently you are not able to capitalize the R&D like you normally would in a program like this.

One way to think about it is the payments to Boeing will certainly be capitalized. It is just a capital purchase. The parts that will be expensed are really more prototyping-type things that are in advance of the actual flight hardware. If you want to think of a simple demarcation line, think of it as engineering models and test. Which normally the prime contractor would do as part of their own payload construction. When we buy it from them, if they do engineering models that ends up being capitalized. But when we are building the payload we are really going to end up capitalizing the flight hardware itself.

It is possible, Chris, that we have been a little conservative in these estimates. But that is -- we wanted to give an outlook that --

That is why we are going through on a case-by-case basis, looking at each part of the program and making a decision whether it should be capitalized or expended.

Exactly.

Think of it as the flight, in some sense, the flight hardware is sort of dividing line, at a simplistic level.

Okay. Just to clarify on the guidance for fiscal 2017, the $380 million to $390 million, that includes the assumption of spending $30 million on R&D-related activities that will not get capitalized?

That is probably closer to $40 million. That is an incremental $30 million over what we spent this year.

Ah. Got you. And for the aero business, aviation business; the number of net adds have been kind of tailing down. When will you complete your roll-outs with your existing customers, and assuming no new customers are brought on?

It is a little more complicated than that because all of our airlines are adding to their fleet or they are rolling -- a lot of this stuff is a little bit in flux because of fuel prices. But in general they are adding new aircraft to their fleet. That has been, announced each new order as it is coming. But that has been growing our backlogs. We probably had added over 100 aircraft without announcing those aircraft on the airlines that we already have. And that is going to continue, as you look at their fleet modernization programs. Then there will be another -- we are pretty excited about our prospects and we think we will see additional airlines coming on as well.

We actually hope to get more of the current customers' fleet as well.

Yes.

Okay. And ViaSat-3, is it going to have the traditional latency of a geosatellite? Or is there anything different about it?

The round trip time to the satellite is still the same. We have said in the past that we are doing things to help mitigate that for a lot of applications, but there will still be geosynchronous satellites and that aspect will not change.

Got you. Great. That is it for my questions. Thank you.

Thanks, Chris.

I think that is probably it for the time we have.

Okay. Good. Thanks. A lot of news today. A little bit longer call. We appreciate everybody's time and attention a lot. So thanks a lot. And we will talk to you next quarter.

Ladies and gentlemen, this concludes today's program and you may all disconnect. Everybody have a wonderful day.