Description

Due to what seems like confusion among investors and perhaps a coordinated stock promotion effort, Santa Claus has delivered a Christmas shorting opportunity. I believe GSAT is a compelling and timely short and that the current inflated price may last only a brief time. Shorting GSAT now given what has recently transpired may be a better risk/reward than shorting immediately before Kerrisdale’s original short pitch when TLPS approval was up in the air.

Many people on the VIC are no doubt familiar with GSAT from Danconia17’s 9/9/2013 long writeup at $0.65 and the subsequent Kerrisdale short presentation from October 2014. Reading the recent long write-up on DISH also provides helpful background on mid-band spectrum. In a nutshell, GSAT is a failing satellite phone business that competes with Iridium and owes over $600M in debt. Its business did the equivalent of $34M in annualized EBITDA in Q3 2016 against $23M in annualized capex and $15.2M in annualized cash interest, so it is hard to see how GSAT stays solvent, much less ever repays its debt or raises financing to replace its aging satellite constellation. To make matters worse, it is out of compliance with its debt covenants, has been selling equity nearly constantly to “equity cure” the breaches and faces steeper principal repayments next year. According to GSAT’s most recent 10-Q by the end of 2016 it will have drawn substantially all of the equity line that has kept it afloat.

So what is a failing satellite company to do? In order for GSAT to operate its satellite phone service, the FCC granted it use of 11.5Mhz of spectrum at 2483.5-2495 MHz. For several years now, GSAT has attempted to generate some ancillary value for the spectrum rights. First, it attempted to use it for fixed wireless internet through WISP OpenRange, but that failed. Note that the FCC readily granted them approval to do this. It then turned to pairing its licensed spectrum with an adjacent 10.5 Mhz of unlicensed spectrum to offer a 4th 22 Mhz WiFi channel at 2.4 Ghz. While Kerrisdale correctly demonstrated that this “TLPS” proposal was likely an unviable business plan, in the end Kerrisdale was proven even more right when the FCC rejected the plan outright due to some combination of interference concerns and dislike of the precedent of giving a company control over use of unlicensed spectrum without having an auction. So the phone business is failing, the first attempt at monetizing the spectrum failed, the TLPS plan failed, why isn’t this thing trading for pennies on the way to BK?

In the past month, after finally throwing in the towel on the original TLPS proposal, GlobalStar finally agreed on a compromise spectrum usage proposal with the FCC. Essentially, they pared down their request to only operate an ancillary “low power” terrestrial network in its 11.5 Mhz of licensed spectrum. Somehow, this has driven the stock from a low of $0.63 post TLPS rejection to $1.81 as I type this. Each incremental step in the FCC approval process seems to be good for 20-30% in the stock, most recently news reports appeared that the votes are lined up to approve the most recent compromise proposal. Amazingly the stock is now up 26% YTD despite not getting the TLPS approval longs were so certain was around the corner at the beginning of the year. The stock is apparently being driven by ignorant, basic high level notions that spectrum is “scarce” and “valuable” and “they are getting approval by the FCC”. For a number of reasons, I believe these notions fall apart under scrutiny and GSAT is headed for eventual bankruptcy. While the original Kerrisdale presentation primarily focused on why TLPS was unviable, there is also good info there on why this backup spectrum proposal has no value (http://www.kerrisdalecap.com/wp-content/uploads/2014/10/Globalstar-GSAT-presentation.pdf).  

Below is why I believe the recent appreciation in GSAT makes no sense and the equity is likely worth zero:

People piling into the stock do not seem to understand exactly what will be approved. GSAT strangely notified investors that it was essentially pulling the TLPS proposal it had bet its solvency on through a tweeted link (http://www.globalstar.com/en/ir/docs/Statement_11_10_16.pdf). It never specified exactly what this new low power terrestrial service would be since to offer a WiFi channel a la TLPS requires 22Mhz. It is not clear that a 10Mhz LTE channel could even offered given the low power requirements and the out-of-band-emissions limitations from having Sprint’s ClearWire spectrum above and unlicensed WiFi below. Not to mention the fact that GSAT’s licensed spectrum is not banded as part of ITU’s mobile wireless standards and has no semiconductor ecosystem. Meanwhile, on Monday December 19th Cowen distributed a note promoting the FCC proposal on circulation and saying that “TLPS makes more sense than ever” and “more WiFi capacity would be in the public’s best interest”. This was followed by a table that valued the 11.5 MHz of licensed spectrum from $0.75 to $1.25 per Mhz/pop, which are perhaps reasonable valuations for unencumbered, full power, banded cellular spectrum. The analyst does not seem to realize that TLPS/WiFi is off the table and GSAT’s spectrum is not comparable to banded full power cellular spectrum. I believe this kind of uninformed speculation and simple spreadsheet math, along with a thin speculative end-of-year market, has driven GSAT’s recent stock price action.

Low power licensed approval should not be a surprise. To anyone following GSAT over the years the fact that the company could get approval for a lower power ancillary terrestrial use of its spectrum should be no surprise. After all, the FCC granted ancillary use once with OpenRange in the past. Observers believe that GSAT has rejected this compromise in the past (http://tmfassociates.com/blog/2016/11/10/back-to-the-future-part-2/) because it feared there would be no value to the proposal, accurately in my opinion. This was always a fallback option and the stock never should have traded below $0.70 either in 2013 or last month if this approval really made the stock worth $1.80- because it has been obvious all along the FCC would accept this. It is my belief that GSAT only accepted this compromise because it had no other option but simply hoped to generate enough hype to get its stock up enough to be able to raise financing to keep this sham going.

The last FCC-related excitement was a golden shorting opportunity. This is not the first time that uniformed speculation on FCC actions has produced an extreme short term stock price reaction. When news of the original TLPS proposal being placed on the FCC’s Items on Circulation broke in May, the stock jumped from $1.80 to $2.75. Of course the market was way ahead of itself and the stock reverted to below $1.00 in the next few months. Amazingly, GSAT’s spectrum is now arguably valued higher on a per MHz/pop basis today since back then people thought GSAT was getting 22Mhz of spectrum and now it is only 11.5 Mhz.

The mobile world is awash in unused mid-band spectrum. Vague notions that spectrum is “scarce” and “valuable” seem to drive the investment case for GSAT. In reality, there are enormous amounts of un-deployed mid-band spectrum that Sprint and Dish have neither found it worthwhile to deploy nor been able to license to the other wireless operators. In contrast to GSAT’s measly 11.5Mhz, Sprint controls between 120 MHz and 160 MHz of adjacent 2.5 GHz spectrum (depending on the geographic market) which has much less restrictive power requirements and is banded for LTE. Sprint uses less than 1/3 of this spectrum and despite its desperate need for cash has been unable to sell or license this spectrum to other mobile operators on acceptable terms. Similarly, Dish controls 88 MHz of lower frequency mid band spectrum and has also not been able to monetize it to date. Next to the swaths controlled by Sprint and Dish, GSAT’s spectrum is insignificant and unlikely to be worth the trouble it would take a carrier to deploy given its lack of standardization.

The relevant precedent spectrum transactions imply a much lower stock price. At $1.81, factoring in 1.1B shares O/S (constantly increasing), around $640M of net debt and $250M in value for the failing satellite business EBITDA, GSAT’s spectrum is being valued at $0.66 per MHz/pop. Sprint bought ClearWire’s large swath of 2.5 Ghz spectrum at around $0.30 per Mhz/pop. Moffett Nathanson recently estimated mid-band spectrum was worth $0.41 per MHz/pop (http://www.fiercewireless.com/wireless/analyst-sprint-selling-some-its-2-5-ghz-spectrum-won-t-be-its-financial-salvation). If you apply $0.41 per MHz/pop to GSAT’s 11.5 MHz, you get around $1.00 per share of equity value.

GSAT spectrum is inferior to the precedent transactions and deserves a discounted price. The above analysis assumes that GSAT’s spectrum is the same as Dish and Sprint’s mid-band spectrum but it is clearly not. GSAT’s spectrum is subject to severe power and OOBE requirements such that it is not even clear it can be used for LTE at all. Low power requirements are restrictive because they require the base stations to be placed much closer together to be usable, increasing the operator’s infrastructure costs. Even if a carrier decided to use GSAT’s spectrum, it would be only after a long process of getting the ITU to put the spectrum in standards, convince semiconductor and phone companies to produce products that utilize the spectrum etc. and even then the spectrum would only be useful in dense metro areas utilizing small cells. Impaired spectrum deserves a discount valuation and industry experts I have spoken with believe this discount would be at least 50%. Dish built most of its spectrum holdings by purchasing the debt of bankrupt satellite companies ICO/DBSD and TerreStar for between $0.13 and $0.15 per MHz/pop, a situation comparable to GSAT today because at the time the spectrum was unbanded although it did not face the same power and OOBE restrictions as GSAT. At the $0.13 per MHz/pop paid for TerreStar out of bankruptcy by Dish GSAT is worth less than $0.15 per share.

There has been a general de-valuing of spectrum since the AWS-3 auction. Due to dubious bidding techniques employed by Dish and its designated entities, the AWS-3 auction for mid-band spectrum cleared for Dish at an amazing $2.79 per MHz/pop. Some believe that Dish purposely drove up the price to re-value the rest of its spectrum acquired at much lower prices, some believe that other participants would not have followed along if the spectrum was not worth that much to them. I am not going to wade into the Dish debate but since then there have been no spectrum transactions anywhere near that value, and Dish’s current equity price implies only an approximately $1.35 per MHz/POP valuation for spectrum. More importantly, it looks like the currently ongoing broadcaster incentive auction for more valuable low band spectrum will finish with disappointing totals, below the $1.28 per MHz/POP paid in 2008 and perhaps as low as $0.90 per MHz/POP (http://tmfassociates.com/blog/2016/12/). This implies the apparent large inflation of general spectrum value of the last few years is over and going the other direction, a poor indicator for a company attempting to monetize a small piece of extremely compromised non-standard spectrum.

Several alternatives for mid-band capacity spectrum have emerged, so the spectrum may be worthless.  I believe one of the reasons for the apparent recent de-valuing of mid-band spectrum is the emergence of unlicensed (i.e. free) spectrum alternatives for the kind of dense metro small cell deployments which GSAT’s spectrum might theoretically be useful for at some point in the future. The first is LTE-U which contemplates deploying LTE over 5Ghz unlicensed spectrum (http://www.lteuforum.org/index.html). The second is 3.5 Ghz shared spectrum technology championed by Google (http://www.fiercewireless.com/wireless/fcc-puts-final-rules-place-for-spectrum-sharing-3-5-ghz-band). Along with these two alternatives for offloading traffic in dense metro areas carriers can also use WiFi offload. Hard to see why any carriers would pay billions for GSAT’s nonstandard spectrum when there are several alternatives for utilizing unlicensed spectrum. If GSAT cannot monetize its spectrum, it faces inevitable bankruptcy and the equity is worth zero.

GSAT is out of compliance with covenants, needs to continue selling stock and faces significant financing requirements next year. It is hard to know from the outside whether GSAT believes its own BS or is simply orchestrating a promotion to keep its market cap up high enough so that it can continue to issue stock, but a simple reading of the SEC filings indicates that GSAT is nearly out of capacity on its Terrapin line and faces over $100M of financing requirements next year. Despite the fact that the revised low-power spectrum usage proposal appeared on the FCC website as of Friday December 16th (indicating likely approval), on Monday December 19th the company sold 15M shares at $0.85 per share even though its shares were already trading well above $1.00. GSAT will likely sell another big chunk of stock before the end of the year to exhaust its Terrapin line. More equity will have to be sold to meet its stepped-up debt repayment requirements next year.

Out of the 1.1B and growing shares outstanding only about 50M are short although 60% of the shares are owned by controlling shareholder Jay Monroe.The last positive catalyst I can think of is formal FCC approval, which should be priced in and then some. After that, it will be a long cold wait for someone to actually pay them for the spectrum. More likely, you will see an attempted sale of a large chunk of equity to keep this thing afloat, perhaps before the disappointing results of the broadcaster incentive auction are finalized. Maybe they will concoct some new BS story around an “innovative broadband service” for “internet of things”, who knows.

Disclosure: The fund that I work for is short shares of GSAT.

Catalyst

• Continued equity sales
• Fade of hype from FCC approval ("sell on the news")
• Disappointing results from broadcast incentive auction
• Lack of spectrum monetization