When I read this review of the TerreStar Genus phone, not only did it confirm my own views about the limited prospects for the phone and the wider lack of interest in dual mode satellite phones, but it brought back quite a few memories from the late 1990s.
Most notably, likening a satellite phone to a “brick” is never a good sign (“It’s huge! It will scare people…If we had a campaign that featured our product, we’d lose“).
Also the suggestion that AT&T hasn’t identified the market correctly if it thinks people will use this as their “everyday mobile device”, along with criticism of the “hefty series of price tags” (“What it looks like now is a multibillion-dollar science project. There are fundamental problems: The handset is big, the service is expensive, and the customers haven’t really been identified“)
The FCC has just released its ruling on cost sharing rules for the 2GHz BAS relocation, which requires that the 2GHz MSS players will have to pay their pro-rata share of the costs incurred by Sprint Nextel in clearing the band. Back in 2009, Sprint Nextel estimated these expenses would be about $100M for each MSS operator.
The FCC ruled that MSS operators would have 30 days to pay these costs after Sprint Nextel presented them with a bill (which could happen very soon after the ruling becomes effective, sometime in November). If the costs were not paid, then the Commission could take enforcement action, although it would not automatically suspend an MSS operator’s license as Sprint requested. In addition, joint and several liability for the costs would continue in the event that a license was transferred to another party (although the Commission did not address how this liability would be impacted by a bankruptcy filing). With respect to ICO Global’s potential joint liability with DBSD for its relocation costs, the Commission outlined certain principles which would apply to this question, but indicated that Sprint Nextel would have to pursue litigation against ICO Global to resolve this claim.
Though this ruling presents certain issues for DBSD, related to its emergence from bankruptcy, it also has a definite impact for TerreStar, as it is now likely that Sprint Nextel will seek to claim $100M+ from TerreStar by the end of the year. Of course, TerreStar Networks might wish to file for bankruptcy to try and avoid this liability just like DBSD (perhaps after the claim is presented rather than before?) and it remains unclear whether TerreStar Corporation would also be subject to joint and several liability. However, if TerreStar is to monetize its 2GHz spectrum in the future (e.g. through an incentive auction) for which it will very likely need the FCC’s cooperation, it seems rather unlikely that the FCC would also allow it to escape this reimbursement obligation, reinforcing that the FCC has numerous levers to ensure that the 2GHz spectrum question is resolved in the way it wants.
Back in August, it looked very much like TerreStar was poised to file for bankruptcy, using a large DIP facility of $200M+ to fund the company for the next several years while it attempted to build a satellite roaming business. Harbinger indicated at the time that it believed a bankrupt TerreStar’s spectrum was worth $1.5B to $2B. This suggests that (as we speculated) the plan may well have been to cram-up the ~$1B of first lien debt that is outstanding at TerreStar Networks. However, it appears that Blackstone was unsuccessful in finding takers for that financing, and so the question arises as to what is now their Plan B for TerreStar.
In TerreStar’s 2010Q2 10-Q, filed in early August, the company stated that “there is substantial doubt that the available cash balance, investments and available borrowing capacity as of June 30, 2010 will be sufficient to satisfy the projected funding needs for third quarter of 2010″ and yet we are now at the end of the third quarter and there has been no bankruptcy filing. AT&T also announced the commercial release of the Genus phone last week. Some have therefore suggested that this means TerreStar is likely to avoid bankruptcy, because AT&T would not have gone to market with the phone immediately prior to a bankruptcy filing.
Of course, if TerreStar had actually secured further investments, these would have to have been disclosed in an 8-K filing, but it is certainly clear the company has managed to make its cash resources last longer than it previously expected. One possibility is that AT&T paid upfront for the Genus phones it is selling, which would have provided some cash for TerreStar without the need for a public filing. It would also explain why AT&T is keen to move ahead with the commercial launch if it now owns several tens of thousands of Genus phones.
UPDATE: According to documents filed in the TerreStar bankruptcy proceeding, Echostar and Harbinger allowed TerreStar to draw a further $10M from the Purchase Money Credit Facility, and agreed to waive any claims of potential default. However, given that the bankruptcy documents show only a few hundred thousand dollars of expected receipts for TerreStar from phone sales by the end of the year, it still appears possible that either AT&T may have paid upfront for the phones it expects to sell, or (perhaps more likely) Elektrobit has been left holding the bag, given the substantial losses it expects to book on its TerreStar receivables.
However, TerreStar obviously also needs to raise further funds to cover its ongoing operating expenses. It has a window of time until August 2011 when no cash interest is due on TerreStar Networks’ first lien debt, and so it appears plausible that raising say $50M+ at TerreStar Corporation (presumably against the security of the 1.4GHz spectrum) would be (barely) sufficient to see the company through to next August, when the outcome of the FCC’s MSS NPRM/NOI should be clearer.
This might well still require a bankruptcy filing by TerreStar Corporation (depending on what happens with the outstanding $408M of Preferred Stock, since the company has certain obligations to redeem the Preferred Stock if it has more than $10M in available funds), but we assume that Blackstone might conceivably look to keep the TerreStar Networks subsidiary out of bankruptcy (assuming this is permitted by the cross-default provisions on the first lien debt) and thereby enable TerreStar Corporation to retain its shareholding in TerreStar Networks.
Then, if it turned out that the FCC’s NPRM allowed the 2GHz satellite spectrum to be monetized for more than the value of the outstanding first lien debt (something we regard as unlikely), the proceeds could potentially flow to the owners of TerreStar Corporation (although it is implausible that an auction could occur by August 2011 so there would certainly still be significant arguments about the value of this spectrum). In addition, this outcome would ensure that TerreStar’s 2GHz spectrum is not on the market as an alternative to LightSquared, while Harbinger seeks to secure partners for its LTE network buildout over the next 9 months.
UPDATE: From the TerreStar bankruptcy filings, it does not appear that the approach we speculated about would have been feasible, presumably at least in part because it was not possible to raise money against the security of the 1.4GHz spectrum, and in the last few weeks the alternative plan to accepting a deal with Echostar was to try and prime the first lien debt which Echostar controlled at TerreStar Networks. However, in the end TerreStar has had to take a $75M DIP from Echostar and agree to the restructuring plan which Echostar proposed.
Its now been announced that JetBlue has signed an MOU with Viasat to install Ka-band connectivity on its fleet, starting in 2012. One of the primary reasons cited by JetBlue was that the satellite capacity was much cheaper than at Ku-band.
We analyzed the cost of providing service for Aircell and Row44 in one of our recent research reports, and concluded that (as JetBlue also asserted), Ku-band satellite capacity can rapidly become the dominant cost driver for aeronautical broadband even at moderate usage levels and take rates. For example, we estimated that at a 25% take-rate, the cost of Ku-band satellite capacity would be between $30K and $80K per plane per year, depending on the amount of bandwidth allocated to each customer. This compares to an amortized satellite equipment cost of perhaps $40K per plane per year. Viasat’s Ka-band satellite could reduce the capacity cost by a factor of up to about 5 times, bringing the cost of capacity down to say $6K to $16K per plane per year.
Thus the strategic question for JetBlue is whether it will use this capacity cost differential to make the service free to end users (or free for most applications other than say streaming video). As noted in past news articles, charging for in-flight broadband has a huge impact on take rates. However, Row 44 (with expensive Ku-band capacity) and Aircell (with a limited amount of terrestrial bandwidth) can’t afford to offer free usage, unless they constrain the service significantly (e.g. no streaming video and limited bandwidth). JetBlue has already offered free (albeit very limited) service on its Beta Blue plane, whereas Southwest (which will set pricing on its Row44-equipped planes) has indicated that it plans to charge for the service.
If JetBlue did offer free service, then this would certainly shake up the in-flight broadband business. Would airlines step-in to pay Aircell directly for their service instead of relying on passenger revenues? Will there be a return of the sponsorship model used on airlines like Virgin America for a period last year? More to the point, will the mere prospect of such disruption cause airlines thinking about installing Ku-band to consider waiting for Inmarsat’s new Ka-band Global Xpress service in 2014?
UPDATE: Now Southwest has agreed to buy AirTran, which already has fleetwide in-flight connectivity through Aircell, will Southwest have yet another reason to reconsider its Ku-band plans with Row44?
Last week it was reported that LightSquared had raised another $750M loan from UBS, to add to the $400M loan it secured from UBS back in July. Today it is also being reported that SK Telecom is considering an investment of up to $100M in LightSquared.
This is undoubtedly good news for LightSquared, although it appears that the new $750M loan is likely to be a refinancing of SkyTerra’s outstanding $750M in first lien debt, which would otherwise have become cash pay on October 1. Even so, Harbinger held the majority of the original first lien loan, and so if the $750M is all new money from third parties, this would enable Harbinger to inject as much as $400M to $500M of additional funding into the LightSquared venture, simply by rolling over its original first lien holdings into subordinated debt or equity. Given that Harbinger stated back in July that it was raising up to $1.75B, then this would appear to match with that target ($400M July loan + $750M September loan + $400M-$500M of new Harbinger equity + $100M of third party funding), although in reality there would only be $1B of additional funding for the network buildout (plus potential vendor financing from Nokia Siemens Networks, which has not yet been announced).
However, last week also brought more ominous news for Harbinger in the form of the FCC’s denial of Globalstar’s ATC waiver request. As we noted at the time, Harbinger faces aggressive buildout milestones and has a pending request for a waiver of the SkyTerra-1 satellite launch deadline. If the FCC is sending a signal to Harbinger that it will not tolerate missed deadlines due to funding problems, which was the principal rationale for the Globalstar ruling, then investors in LightSquared will have to worry about how valuable the spectrum assets would be if LightSquared failed or otherwise could not meet the FCC’s deadlines. Indeed Harbinger’s own “voluntary commitments” included the condition that the LightSquared “authorizations” (its ATC license and perhaps even its MSS license depending on the interpretation of this phrase) will automatically be “null and void” without any need for further action by the Commission if LightSquared fails to meet the buildout milestones.
In particular, the question arises of whether the FCC would impose costly buildout conditions on a potential purchaser of the spectrum assets, or possibly even veto a purchase by AT&T/Verizon on competition grounds. In our view, the Globalstar ruling introduces rather more uncertainty about whether it will be possible to guarantee active bidding in the event that the LightSquared assets are sold in the future (especially in a bankruptcy auction where approval of the transfer would have to be sought afterwards from the FCC), and thereby makes it harder to put a floor under the value of LightSquared’s spectrum.
One of the most interesting questions about Inmarsat’s new ISatPhone Pro is how well it will work at low elevation angles, including for example whether the phone antenna needs to be pointed towards the satellite. This is going to be particularly relevant in Alaska, much of which lies very close to the nominal edge of coverage, and well outside the 20 degree elevation angle contour (where Inmarsat suggests that “more user cooperation is required”), as shown below.
However, I’ve been told by Inmarsat that the phone is performing better than expected, even at relatively low elevation angles, so it will be interesting to see what this means in practice. Given that the beams used for registering the phone on the Inmarsat satellite are lower power than the beams used for a call, it appears probable that either the phone will register successfully and then calls can be made OK, or the phone won’t register and then no calls can be made at all.
Its surprising that we haven’t yet seen any published real world tests of the Inmarsat phone in comparison to Iridium, similar to the Frost & Sullivan reports which compared Iridium and Globalstar in 2008 and 2002. However, I’m sure similar analyses will be undertaken by both Iridium and Inmarsat at least for their own internal purposes, and possibly even for external publication if they believe the results are favorable. If you’ve tried out the phone in “fringe” coverage areas then feel free to let us know about your experience in the comments section below.
UPDATE: So now Frost & Sullivan has released its comparison of the Iridium and Inmarsat phones, which was commissioned by Iridium. It is notable that in Anchorage, Alaska, Frost & Sullivan “was unable to make or receive a call despite dozens of attempts and was only able to briefly find a satellite”. This points to difficulties with registration, as we suspected. However, Inmarsat sources tell us that it is perfectly possible to register on the satellite in Alaska, and make calls there. We haven’t yet got an independent view, but it would seem likely that the actual answer may lie somewhere between these two opposing views. We would speculate that you will probably have to have a pretty good idea where the Inmarsat satellite is so you can point the phone antenna at it during registration (maybe using a compass would be helpful?).
We’re told that TerreStar is planning to announce (tomorrow?) that the Genus phone will be released by AT&T next week. However, surprisingly enough, AT&T’s filing yesterday in response to the FCC’s NPRM/NOI on MSS spectrum, didn’t mention the Genus phone once. Not only that, but AT&T actually suggested that “rationalizing the MSS bands for terrestrial wireless use is a
good first step to implementing a comprehensive broadband spectrum strategy”, and supported the concept suggested in the NOI, that “there may be opportunities to ‘meet future [MSS] needs with less allocated spectrum in some or all of the bands.’”
Therefore the obvious question is whether AT&T cares about the success of the Genus phone, or instead would actually benefit from it failing, because it believes that the “2 GHz MSS band is a good target for the creation of new terrestrial mobile services”. Of course AT&T is a large company, and what is in the interests of AT&T at a corporate level may differ from the priorities of the staff working on the Genus launch. However, given the challenges that the Genus phone already faces, it is noteworthy that the project does not appear to enjoy much recognition or support when AT&T is setting out its strategic interests in the wireless business.
The FCC today released its ruling denying Globalstar’s request for a postponement of the deadlines in its ATC license, which required launch of its second generation satellites by July 1, 2010 and provision of two-way service to its ATC terminals by July 1, 2011. The FCC has granted Open Range a temporary waiver, which basically gives it 60 days to make other spectrum arrangements or its network will be shut down.
This ruling comes as quite a shock to most observers, because it was assumed that the FCC was contemplating providing more flexibility to MSS-ATC licensees after release of its recent NPRM/NOI. However, as we argued at the time, the contents of the NPRM/NOI were actually something of a disappointment to those expecting such liberalization, because the emphasis was on reallocation of the 2GHz spectrum for terrestrial use, with incentive auctions or other mechanisms used to ensure that the government receives “appropriate compensation for the step up in value” that would occur if the existing ATC restrictions were removed in that band. In that context, as we suggested, it would be hard for the FCC to provide further flexibility to ATC licensees in other bands (i.e. LightSquared and Globalstar) with no offsetting “compensation”. Nevertheless, we had still expected that Globalstar might be granted its requested waivers, because LightSquared had achieved the ATC license modifications it desired back in March.
However, now that the FCC has taken a hard line with Globalstar, it raises the question not only of what Open Range will do next for spectrum, but whether any future inability to meet license conditions will place other ATC licenses (and the associated spectrum assets) at risk. Notably, observers will presumably begin to wonder what will result from the proceeding relating to the reimbursement claimed by Sprint from DBSD and TerreStar for clearing the 2GHz spectrum band (estimated by Sprint at $100M+ per operator), compliance with the outcome of which was a condition of TerreStar’s ATC license grant back in January. Though DBSD has sought to avoid these costs through its bankruptcy filing, it is less certain that TerreStar would be able to do likewise. TerreStar has also recently requested certain waivers of the ATC base station and terminal requirements from the FCC. In addition, it is quite possible that there may be requests by Harbinger to extend the very aggressive terrestrial deployment deadlines associated with the LightSquared network at some point in the future, and in the near term, LightSquared recently delayed the launch of its first next generation satellite to December 2010, which will also require a waiver from the FCC, and questions are sure to be raised about whether this delay was solely attributable to technical problems.
With comments due in response to the July 2010 NPRM/NOI tomorrow, it is likely that a lot of last minute redrafting of submissions is going on tonight, so it will be interesting to see whether any of these issues are raised either by the satellite companies themselves, or by terrestrial wireless interests encouraging the FCC to continue to take a hard line on ATC.