Globalstar Inc (GSAT)

[236T568]

"Don't Be Fooled By GlobalStar's New Spectrum: Downside Over 90%"

http://seekingalpha.com/article/4035410-fooled-globalstars-new-spectrum-downside-90-percent




Don't Be Fooled By GlobalStar's New Spectrum: Downside Over 90%

Jan. 9, 2017 9:06 PM ET
|
23 comments
|
About: Globalstar, Inc. (GSAT)




Ruerd Heeg


Ruerd Heeg Premium Research »
Follow (1,010 followers)

Long/short equity, deep value, special situations, book author


Send Message
|
Ruerd Heeg's Book on Value Investing (Dutch)


Summary

•
Last year, the FCC approved a small band of GlobalStar's spectrum for terrestrial use: from 2483.5 to 2495 MHz.
•
Some people have concluded the value of this spectrum is worth several dollars per MHz pop. That implies a multi-billion dollar value for GlobalStar.
•
Assigning a value to spectrum is tricky. In this article, I will discuss technical details for valuing the company's new spectrum that haven't been discussed before.


GlobalStar (NASDAQ:GSAT) is a much-discussed stock. Last year, I wrote 2 articles on GlobalStar (see here and here for PRO subscribers). Another useful source is Kerrisdale's original short thesis. I recommend reading my earlier public article, and if you have time, also Kerrisdale's original short thesis, since this is good background information. A side note for people who have read my PRO article on GlobalStar: I again heard cognitive stress when the company answered the question about monetization of the spectrum during the conference call discussing the new spectrum.

Last year, the FCC finally approved GlobalStar's so-called TLPS spectrum, or at least part of it: 11.5 MHz, from 2483.5 to 2495 MHz. Many investors now have several questions:
1.How is GlobalStar going to use the spectrum?
2.What are the conditions for using the spectrum?
3.How much is this spectrum worth?

How GlobalStar is going to use the spectrum

When listening to the conference call, it becomes clear the company has no firm plans for using the spectrum. I expected that: if GlobalStar had concrete plans, it would have been more appropriate to announce them in a press release before the conference call. The company said there are possibilities, though: selling the spectrum, leasing the spectrum or partnering with a telco to build a network around it. During the conference call, the company said it won't build a network alone. So what it will do depends on what opportunity comes up. Some bulls think Sprint (NYSE:S) would like to use GlobalStar's spectrum, since Sprint has spectrum next to it. They should read this comment.


GlobalStar mentions the spectrum is going to be used for so-called small-cell wireless communications. That means the company will need many base stations, making its network more expensive to build. From the conference call and approval document, I understand GlobalStar still wants to use in-house modems as base stations. In that case, it has to partner with an internet provider.

Such a network can work fine, but I do not have great confidence in it. In the Netherlands, we already have such a network. Maybe 20-30% of houses share their Wifi with strangers. In return, they can access shared Wifi from others on their phone. In practice, the latter does not work very well. Outside I can pick up several signals from other modems, but these signals are usually too weak for my phone. In the Netherlands, Wifi transmitters are restricted to sending at most 0.1 Watt. This is to prevent interference between signals of modems in neighboring houses and apartments.

Conditions for using the spectrum

An important restriction is the 1 Watt sending power restriction. In the conference call, GlobalStar mentioned 4 Watt, but that refers to peak power. The reason for the low power restriction is to prevent interference with other frequencies. The low power restriction limits the use of the spectrum: clients cannot be too far away from the access point. Hence the suggested small-cell use.

I have done some research on the distance 1 Watt can travel at this frequency. Specifically, I have read reports from users of (illegal) high-power modems operating at similar frequencies. In cities, 1 Watt should be sufficient for access points to communicate with client devices from 100 meters away. Within buildings, the maximum distance is much smaller, and also "misplaced" buildings can block the signal outside. On the country side, the distance is larger - maybe 200 or even 300 meters. I think there the maximum transmitting power of the client device limits the maximum distance.

Another condition for using the spectrum is operating a Network Operating System, or NOS, 24x7. In case of interference, people should be able to contact an operator. This operator can then deny authorization to access points and client devices. Among others, this is for TV broadcasters. They sometimes use GlobalStar's frequencies. If they experience problems, they can call GlobalStar's NOS. The NOS operator can "switch off" specific access points and clients.


How much is GlobalStar's newly approved spectrum worth?

Spectrum can be worth much. In the beginning of 2015, spectrum was auctioned for $44 billion. Then, similar frequencies were purchased for several dollars per MHz per head of population, or MHz pop. But is this spectrum really comparable or is it very different?

No, there are several important differences. Two differences are obvious: the low power restriction and especially the slightly higher frequency. I think this knocks off about 2/3rd of the potential value of the spectrum. These features cause signals to penetrate less far. That this matters is confirmed by experts: "Empirical evidence from studies of other spectrum auctions shows that relationship between frequency and auction value tends to follow an inverse exponential function". But with hundreds of millions of inhabitants (pops) and a bandwidth of 11.5 MHz, a value of $1 per MHz pop already corresponds with a value of several billions of dollars.

Unpaired versus paired spectrum

Unfortunately for GlobalStar longs, there is another difference: the company's spectrum is unpaired. GlobalStar has to separate downloads and uploads by using different time intervals instead of different frequencies. The technology behind unpaired communication poses many problems. One of these problems is synchronizing uploads and downloads between the client device and the access point. For example, this synchronization depends on the distance between the client and the access point. Another problem is interference between adjacent cells. Because of these technical problems, unpaired spectrum has always been much cheaper than paired spectrum. In the AWS-3 auction of only 2 years ago, 15 MHz of unpaired spectrum was purchased for as little as $0.52 per MHz pop.

Furthermore, GlobalStar's band of unpaired spectrum is only 11.5 MHz. Bands of less than 15 MHz are said to be less valuable because of certain scale limitations of narrow bands.

So if I am generous, unencumbered unpaired spectrum could be worth $0.6 per MHz pop today. But here, the low power restriction knocks off more than half of the value. Let's assume we arrive at $0.3 per MHz pop because of the power restriction. I think the slightly higher frequency and the narrow band knock off another $0.15 from the value of the spectrum.


Then there is still the issue that GlobalStar's spectrum won't work on existing mobile devices, as Gerst Capital made clear. Last year, this author published several articles on the subject. See, for example, here. Another expert confirmed these findings to me. A recent commenter explained this is probably a show stopper. It is very unlikely new phones will be made supporting GlobalStar's new spectrum. But let's pretend we are bulls and assume this issue can be fixed on new phones.

So I think a valuation of $0.1 per MHz pop is still very generous. There is still the possibility GlobalStar monetizes its new frequency outside the US. However, this is unlikely to happen soon, since there the spectrum has not been approved yet. Furthermore, unpaired spectrum has been auctioned for much lower prices abroad than in the AWS-3 auction in the US.

Staying optimistic, a value of $0.1-0.15 per MHz pop corresponds with a total value of between $368 million and $552 million for GlobalStar's spectrum.

How this could work out in practice

In the recent conference call, GlobalStar confirmed it will not build a network of its own. Instead, the company has to partner with (probably) a telco. Together, I imagine it will distribute new modems on which the spectrum can be used. I suppose it will work as in the Netherlands: customers of Ziggo (part of Liberty Global (LBTYA, LBTYK)) can opt to share their internet connection, and in return, other customers share their internet connection with them. In the Netherlands, it does not work outside because of the weak modem signal of only 0.1 Watt, but with 1 Watt of power, it will probably work for GlobalStar's spectrum.

There is one big "IF", however. In the Netherlands, Wifi operates at a low signal to reduce interference. If GlobalStar's modems are going to transmit at 10 times higher power, interference could become a huge problem. So, instead of building a network using people's home modems, GlobalStar might have to build a network with outdoor access points every 100-200 meters. That way, it could reduce interference by cleverly choosing the locations of the access points. However, building many access points outside will be too expensive to make the network a profitable asset ever.


In a big country like the US, it will be much more difficult to get enough participants for a meaningful scale than in the Netherlands. Initially, there won't be a benefit for participants, since nobody else shares its internet connection. Furthermore, using the new frequency requires a new phone or tablet (if this is possible at all). I suppose the only way to get around this is giving the new modems away for free for a while. Once there is sufficient scale, GlobalStar and its future partner can start trying to charge subscription fees.

How much would people pay for such a connection? I suppose at most a couple of dollars per month. I suppose the partner bringing in the clients will get at least half of that revenue.

Now suppose 50 million households take a subscription for $5 per month, or $60 per year. Not now but within 3 years, since GlobalStar does not have a partner and there are no concrete plans for monetizing the spectrum anyway. If we assume GlobalStar gets $30 per year, its share in upfront capex is $30 per household (how is the company going to finance that?), and money is discounted at 10% per year, the company will be doing fine. In that case, the spectrum will indeed be worth about $4.5 billion.

But many assumptions here are extremely optimistic. For instance, the assumed subscription price of $60 per year is very optimistic. A more realistic price is $24 per year. Also, telcos could do something similar with fully owned unpaired spectrum acquired at a low price. So the signing up of millions of clients creates the value here, not the spectrum. Getting 50 million households subscribed within 3 years is almost impossible anyway. Don't forget, clients will most likely need new phones and tablets to get the benefits of the new service.

Conclusion

Without partners and concrete plans for monetizing the spectrum, assigning any value to GlobalStar's spectrum is a dangerous game. And don't forget, the new spectrum is unlikely to work on existing phones and tablets. Nevertheless, extremely optimistic and risk-seeking investors can come up with a value of about $550 million. If we assign $300 million to its satellite operations, the company's total value is $850 million. This $300 million is slightly above book value, which is not bad for a risky and consistently money-losing business. More details on the satellite business can be found in Kerrisdale's short thesis (link above).


At $1.66, the market cap is about $1.8 billion. With about $750 million debt and derivative liabilities, net of cash, the enterprise value is about $2.6 billion. So, GlobalStar stock is really not worth more than the difference between the total value of $850 million and the $750 million of net debt and derivative liabilities: $100 million, or about $0.09 per share.

Investors here make a common mistake - they overestimate the probability of multiple events depending on each other. Now that the spectrum has been approved, many investors will reassess the probability on the other events: that the company finds a good partner, that the combination will be successful at monetizing the spectrum and that GlobalStar's share in the results is good enough.

At major positive events, investors become euphoric and bid up the stock. As time goes by, they realize how much has to be delivered to justify the price increase. Especially as soon as the stock has gone down 20% or more, many investors realize their mistake. So often, an event like this one creates a perfect shorting opportunity. After the event, the stock price usually adjusts to a more reasonable level.

Disclosure: I am/we are short GSAT.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.