Sunday, November 18, 2012

The "Mobile Satellite to LTE" Bubble is Forming

Investing in U.S. mobile satellite ventures never has been too terribly profitable. You might argue Long Term Evolution is a different matter, so mobile satellite spectrum re-purposed for LTE support would make sense. But one has to wonder whether yet another phase of "mobile satellite" over-investment is about to begin.

Mobile satellite has for decades seemed like an eminently sensible idea to some, the theory being that mobile satellite service could reach many users out of range of the fixed mobile networks. But more money has been lost, than made, trying to do so, with some notable exceptions.

Not to worry, the new idea is to re-purpose such spectrum to build new Long Term Evolution mobile networks. So far, Lightwave has failed, Dish Network is waiting and Globalstar wants to join the party.

But one wonders whether the new efforts will be more financially satisfying than the former efforts, as there is a fairly consistent record of failure in the U.S. mobile satellite business. Some might remember Teledesic, for example.

But similar efforts to create new national wireless networks also have lost billions. Some will remember Teligent, for example.

In fact, the very fact that there exists mobile satellite spectrum to re-purpose illustrates the general failure of most mobile satellite ventures in the past. But LTE is not mobile satellite, it is fair to say. Neither, on the other hand, is LTE an automatic winner for every potential contestant. 


If you accept the notion that four national LTE providers is at least one too many, one will be skeptical about adding three more facilities-based providers. Nor is the effort to monetize mobile satellite spectrum especially encouraging.
LightSquared, for example, can trace its current business plan back to 1988, when it was known as American Mobile Satellite Corporation.

SkyTerra was born from American Mobile Satellite Corporation, renamed Motient Corporation, then changed to Mobile Satellite Ventures.

The point is that SkyTerra has experimented with a number of business models over the years. 


Spectrum Deficit By 2013
For its part, Dish Network has acquired DBSD North America, formerly called ICO North America. Reston, Va.-based DBSD, whose S-band satellite and planned terrestrial network foundered on a lack of financing, filed for bankruptcy in 2009.

Then there was Iridium, the Motorola-owned fleet of low-earth-orbit satellites intended to provide a range of data and voice communications. It went bankrupt, but not before being purchased by Craig McCaw's Eagle River investment entity.

The Globalstar project was launched in 1991 as a joint venture of Loral Corporation and Qualcomm. On March 24, 1994, the two sponsors announced formation of Globalstar LP, a limited partnership established in the U.S., with financial participation from eight other companies, including Alcatel, AirTouch, Deutsche Aerospace, Hyundai and Vodafone.

The point is that mobile satellite communications, at least in the U.S. market, has a long history, but precious little evidence of marketplace success. The latest move to repurpose most of that spectrum for use supporting Long Term Evolution networks, a sensible enough idea, especially if one assumes there is not much chance to gain spectrum from new auctions.

But perhaps nothing is more common in the early days of what investors believe will be “big new markets” than over-investment. Simply, too much money chases the opportunity. Historically, one fairly easy way to eventually lose lots of money is to invest in “mobile satellite” services.

So it is logical that owners of mobile satellite spectrum are attempting to repurpose that spectrum for terrestrial mobile networks, at the moment Long Term Evolution. To some extent, the companies simply are making a bet that the value of their spectrum will grow if available to support LTE, even if the actual networks aren’t built.

It’s a logical theory. So far, Lightwave has failed, Dish Network is waiting and Globalstar wants to join the party. Since, ultimately, spectrum is valuable, there is a logic to these efforts.

But there is probably a stronger argument that the spectrum ultimately gets used, than that three new networks get built. Clearwire, keep in mind, had precisely that same thought, believing it could build a big wholesale 4G business. But it hasn’t really worked.

Assume, for the sake of argument, that three new facilities-based Long Term Evolution networks were to launch, in a market where there are five facilities-based national LTE networks. What would happen?

The obvious and nearly immediate likely consequences is a destabilizing of retail pricing for mobile broadband service. And since mobile broadband is the current driver of mobile service revenue growth, that could have unpleasant consequences for incumbent providers.

The other medium-term consequence would be that the original business plans conceived by the new providers would begin to deviate from plan. The very market entry they represent would act to depress market revenues for mobile broadband, the lead application the new providers likely would emphasize.

It might take a few years, but the three new contestants likely would wind up being acquired. In fact, some might argue that is the whole point of launching the new businesses. There being a rather slim chance of actually rising to leadership of the mobile market, the better investment strategy is to create the network, build share, and then sell to one of the other contestants.

Assuming all that new capacity does wind up in the hands of the larger contestants, eventually, you might also ask whether a “spectrum glut” would result. Not necessarily. The incumbents are “rational” and would activate the new spectrum as needed, and not in a way that destroys the scarcity value of their networks.

It wouldn’t be so much that “supply creates its own demand,” as that supply would be activated in a controlled fashion, to support incremental demand, and not in a way that destabilizes retail pricing.

But there are other possibilities, such as one of the three proposed networks actually finding a new niche. Dish, with its combination of video entertainment and broadband access, would be a possible candidate.

Dish might not focus on “cheap broadband” only, but rather lead with video entertainment. Dish might merchandise broadband to sell video subscriptions, in other words.

Some might argue it is unlikely all three proposed new networks will be able to launch. More likely is an asset purchase of the spectrum by one of the other market-leading firms.


Dish Network, conceptually, at least has the "video entertainment" angle. Where all the other networks are based on voice and data, Dish presumably would want a video-driven business model.

Of course, mobile video has its own history of failed hopes.


No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...