Does Viasat have a future?

Although I do not write about finance, by tracking news regarding satellite broadband I have been increasingly interested regarding Viasat, as one of fierce competitors of Starlink. 

Although not located in Silicon Valley, Viasat has some similarities. It is a technology driven company founded in Southern California. It was founded and led by engineers first and foremost. But since 2000, it has gradually expanded from providers of the ground segment solutions into various segments of the whole satellite communications value chain. Through series of mergers and acquisitions it showed steady growth for over thirty years. It became more and more typical business driven company. But it retained its technology driven aspect, resulting in its crown jewel - integrated satellite broadband solution based on its Viasat series of satellites, which were on the forefront of phased array, frequency reuse and high throughput satellites development. All this enabled them to plan and execute Viasat-3 constellation of three GEO birds covering all of the globe with 3 Tbps of combined throughput. The scale of the investment for the company was huge - almost two billion dollars. That almost matches current market capitalization of the company (currently at 2.4 billion) and total annual revenue (at 2.1 billion in 2019). This crown jewel increased company debt (already at 1.8 billion, but that includes the cost of the first Viasat-3 satellite). 

Then Eutelsat bailed out of Viasat-3 joint venture for EMEA region. But this was just the beginning of its trouble.

Although its main nemesis (Starlink) was announced at the beginning of 2015, it seems it was not taken seriously. After all, what competencies and know-how SpaceX had? Almost none. They did design some space hardware (Dragon-1) and Crew Dragon was still in early development stage. But high throughput communications? Frequency reuse? Ku-band/Ka-band experience? Phased arrays? Mass terminal production? Gateway design? Network management? SpaceX had no experience in areas where Viasat had strong technological edge. So it was not taken seriously. I mean, this guy, Elon Musk. Ok, he is successful. But then again, at that time, reusable boosters were not a thing. It happened in 2017. Just three years ago. Just three years later, most of the Falcon-9 flights are on reusable boosters. Essentially, in just three years, none of SpaceX customers requires new boosters for new contracts.  In just three years, SpaceX managed to cut its internal cost of F9 launch (without additional services) from some 55 million to just 15 million (estimated marginal cost of Starlink launch). This guy launched other successful business, Tesla, which produced just 35000 cars a year before (2014). Just 0.03% of total world production. 300 times smaller production capacity of Toyota or VW.  After 17 years in operation. And seven years of vehicle production. And he is going to take on Viasat? SES? Hughes Networks? Eutelsat? Intelsat? Which were, at the time, also large customers of SpaceX. And at the same time, SpaceX would also build low cost satellites. Much cheaper than Boeing, LM, SSL, Airbus and others, who have decades of experience. This fantasy of Starlink required revolution in so many areas at the same time. It is not just batteries as for electric vehicles. It is not just in reentry and landing precision as for rockets. No. Starlink required revolution in satellite manufacturing,  phased arrays, network management, traffic routing, booster reuse, launch vehicle payload packaging converging at the same time. How long will it take? And how much it will cost? Probably much more than two billion that Viasat planned. Or SES. Even OneWeb LEO constellation, with 5 Tbps total capacity spread around the world, required 3.4 billion investment before bankruptcy. And additional one billion investment after bankruptcy. It is still unclear how much more will be required to complete this constellation. So Viasat management, predictably, thought that they cannot beat Starlink in latency, but definitely can compete in lower price per bit and time to market.

But here comes 2021. Starlink is already signing up initial customers. Not only in USA. In UK. Australia. New Zealand. And Germany, South Africa, Greece just waiting for regulatory approval. COVID-19 has hurt aviation business, an important vertical for Viasat. Its original launch plan slipped from 2019 (at the time of announcement) to 2021. But booster reuse is real. And Starlink works (although still in Better Than Nothing Beta). Its space segment cost is much lower. Viasat can launch 1Tbps satellite for estimated cost of 650 million. Single Starlink launch (of 60 satellites) has similar capacity (1Tbps) for less than 60 million. Order of magnitude reduction of cost. Speed of Starlink obviously surprised all competition.

And then comes the ground segment. User terminals. One area where Viasat is a technology leader. Their Internet broadband terminal costs 800USD. With classic parabolic antenna. Starlink terminal is 500USD with phased array antenna.  They are actually cheaper than Viasat hardware. This is huge and (I believe) completely unexpected by Viasat and others in the industry. It is a huge achievement. Imagine if first Tesla car that was produced in 2008 was Model-2 with price of 25000USD and over 300 mile range. The very first model. Cheaper than comparable gas powered car even without government subsidies. With the same or larger range. And with cost of energy per mile ten times less than all other cars on the market. This is a level of revolution that Starlink brings in its first iteration. Something Tesla has not achieved yet, after 12 years of vehicle production.

So from financial standpoint, I don't see how Viasat can compete in satellite broadband. Even in its initial form, Starlink is so much better and cost effective. So how come Viasat-3 is not scrapped? It is too far in the investment cycle. Most of the money has first two satellites was already contracted. But it is clear that Viasat-3 will become a huge failure down the line. Existing Exede customers will likely flee to Starlink once it becomes available. Some will remain with huge price cuts. Existing broadband satellites will loose its network traffic. It will not happen overnight. But this transition will hurt Viasat bottom line badly. While debt will remain on the books. 

So interesting things are happening to Viasat on the investment front. Its founder and long term CEO, Mark Dankberg resigned from his role of CEO, but he retained his Chairman position. New CEO is another long-timer and COO, Rick Baldridge. It is a logical choice. No rocking of the boat. Business as usual. No reason to worry. But looking at the reported insider trading, a different picture emerges. Viasat executives have regularly sold stock that they received as part of compensation since 2011, which indicates they do not build stock positions. Both Mark Dankberg and Rick Baldridge were especially aggressive in selling their stock. Former CEO still retains around 2.2% of the stock (roughly 1.5 million out of 67 million shares). And the largest shareholder is lead by "Oracle of Boston", Seth Klarman, a billionaire that shares similar investing style as Warren Buffet, owning almost a quarter of the company. So what are investors seeing that I don't? 

Well, Viasat has strong corporate governance spanning over three decades. It is technology company that seemed to have an edge in satellite broadband. Until Better Than Nothing Beta came along. Suddenly, Viasat seems worse both technologically and commercially on every possible metric. When first Viasat-3 launches, initial Starlink constellation will be complete and out of beta period. It will have hundreds of thousands of subscribers. They might even overtake Viasat in number of subscribers by that time (around 600000 residential subscribers), and definitely some of them will be churned away from Viasat.

So where this supposed investor confidence comes from? Well, Eutelsat bailout from Viasat-3 EMEA joint venture required filling an investment hole of 168 million dollars (140 million euros). Which closely matches 175 million investment into by two major shareholders done a few months before. And beside residential broadband, Viasat has other verticals (government, in-flight connectivity, terminal design and production etc.), while consumer broadband is roughly 30% of the revenue. Other 70% of the revenue is not directly targeted by Starlink. Yet. But how long will it take Starlink to move into in-flight connectivity market? And maritime? And government? Long term Starlink will hurt all business areas that Viasat operates except its ground segment products (terminals and radios). At the same time, pilling debt of Viasat-3 constellation will remain to be paid off. 

Switch to phased array antennas and LEO constellations represents major technology shift which Viasat management and investors don't seem to get. Starlink seem to be their Kodak moment. Their latest actions against Starlink affirm that they are aware of this strategic threat, to which they are exposed more than other GEO satellite operators precisely because of their bet on Viasat-3 constellation. They have bet to build the best supercomputer at the same time PC was born. Their strong point will remain ground segment communication equipment. They will likely try to partner with other providers (SES mPOWER seems to most promising) in order to leverage their technology into building low-cost ground terminals.