Back in September I tried to analyze LightSquared’s cashflows, in order to predict when it might run out of money. A Debtwire article published in the FT earlier this month gives some useful data which has enabled me to refine that analysis. In particular, it states that after paying Sprint its $290M and raising $265M as an unsecured term loan, at the end of the second quarter of 2011 LightSquared had about $614M in cash.
That amount is very close to my estimate, with one exception, namely that it appears LightSquared did not repay Boeing’s vendor financing loan of ~$120M in December 2010. That loan was secured against the SkyTerra-2 (formerly MSV-2) satellite, which LightSquared told the FCC in October 2010 would be launched in the first half of 2011 (to meet the deadline in LightSquared’s Canadian license to place the satellite in orbit of April 23, 2011). Of course that didn’t happen, and I assume that LightSquared cancelled the launch to save money.
However, as I noted last December, Boeing’s contract to build the MEXSAT system involves the construction of two copies of the LightSquared satellite, and it would therefore make sense for Boeing to re-use the SkyTerra-2 satellite for its MEXSAT contract. It therefore seems that Boeing has either kept the vendor financing in place (under the assumption it will pick up the SkyTerra-2 satellite in bankruptcy) or has already reached some agreement with LightSquared to purchase the SkyTerra-2 satellite in exchange for forgiving the vendor financing loan. Either way, it seems that LightSquared’s creditors may now have one less asset to monetize in bankruptcy.
Taking this incremental $120M in cash into account, I have updated my financial projections, and these are given below. LightSquared has stated that it is funded through Q1 of 2012, and that it can “extend the runway of cash” if it has to. This appears to be a reference to deferring the next significant payment to Sprint, which I understand is due at the end of Q1. Sprint may or may not choose to extend the payment schedule with LightSquared, though Sprint’s actions would likely be dependent on what best positions it to retain the $290M already paid by LightSquared in a bankruptcy. However, as shown in the analysis below, even if no more payments are made to Sprint, LightSquared would still run out of money by the summer of 2012.
Interestingly the DebtWire article claims (incorrectly in my opinion) that LightSquared’s “liquidity is sufficient through 2012″ and more astonishingly that “current trading levels [for the first lien debt at 50 cents on the dollar] imply investors are taking a conservative approach and only valuing the more dependable 20 MHz, valued at USD 0.13-USD 0.14 megahertz per population (MHz/POP)”. Of course, this is far from a conservative approach when comparing the valuation with TerreStar and DBSD, because LightSquared must make lease payments of $115M per year to Inmarsat for access to the L-band spectrum, equivalent to an additional 30 to 40 cents per MHzPOP (depending on the appropriate discount rate), when attributed solely to the “more dependable 20 MHz”. Put another way, if LightSquared has to keep paying for the Inmarsat lease, and can only make use of 20MHz of its L-band spectrum (even if that were usable in the near term, which it will not be), then if DBSD and TerreStar were the appropriate benchmark (20 to 25 cents per MHzPOP) the LightSquared spectrum would be worthless.
Over the last few months, LightSquared’s CEO has repeatedly asserted that he is “absolutely confident” that LightSquared will “build a system that works with” GPS and that “we will get the proper approvals from the FCC and other government agencies” when the decision comes out in September. Most recently, at the end of September, he told the New York Post that he is “absolutely confident the FCC will approve” LightSquared’s revised business plan later this year. I’m also told that earlier this year he personally “guaranteed” to investors there was no interference problem with GPS, which was how LightSquared was able to raise a further $586M in February.
News has emerged today which yet again proves Mr Ahuja’s confidence was completely misplaced, with Aviation News reporting that the next round of government tests of LightSquared’s “lower 10″ plan are getting underway this week and will run into the first part of November. Notably, as I indicated earlier, high precision receivers are not part of this testing and so the testing needed for the FCC to reach a decision will certainly not be completed this year (not to mention the fact that the NPEF doubts it will be possible even to complete its analysis of the current testing by November 30).
High-precision GPS receivers, which original tests showed would suffer unacceptable interference even from lower-band transmissions, are not being tested as LightSquared is still working with suppliers on development of antennas and filters to protect these devices, used in agriculture, construction and scientific sectors.
Indeed a new ex parte submission, documenting a meeting between the GPS industry, LightSquared and the FCC on October 25, brokered by a member of Congress, confirms that “LightSquared has proposed to develop new filters that could be employed in new high precision GPS receivers that would protect against interference from Low 10 MHz transmissions. Such filters have not yet been made available”. This is despite LightSquared’s assertions on September 21 that “Javad GNSS has completed the design, made prototypes and tested those prototypes. Preproduction units will be released for public tests in October, followed by mass production.”
Perhaps that is why LightSquared apparently now plans to commission its own “independent” tests by Alcatel-Lucent Bell Labs, though of course those will not carry any weight in the testing process mandated by the NTIA and FCC. However, it is hard to see how LightSquared can claim that “These solutions will undergo extensive National Telecommunications and Information Administration (NTIA) and Federal Communications Commission (FCC) testing in the coming weeks” unless “in the coming weeks” is re-defined as “sometime next year (if we’re still in business then)”.
Over the last month, we haven’t seen as many high profile appearances from Mr Ahuja, and apparently he has been visiting India to promote his Augere venture in Asia and Africa. However, he is scheduled to speak on Wed Nov 2 at the Open Mobile Summit in San Francisco, and it will be interesting to hear how “absolutely confident” he is in whatever LightSquared’s story changes to next week.
Last week we saw various filings from wireless companies and the CTIA on DISH’s transfer application/ATC waiver request for DBSD and TerreStar, which mostly just served to highlight the interests of the different wireless operators. MetroPCS asked the FCC to force DISH to provide a detailed business plan before considering the transfer application (in other words to do a deal with them rather than wait for a decision on the AT&T/TMO merger). Sprint asked the FCC to condition the transfer on reimbursement of Sprint’s relocation expenses, and to impose the same buildout conditions on DISH as the FCC imposed on LightSquared in exchange for the waiver (thereby potentially forcing DISH to pay Sprint to host their buildout). Meanwhile the CTIA asked the FCC to defer any decision on a waiver until a more general rulemaking proceeding has been completed (thus allowing the wireless operators to find out what will happen to the AWS-3 spectrum and whether the 1755-1780MHz block will be available for pairing). Ironically the CTIA cited potential interference with the G-block PCS spectrum as a reason for delay, when Sprint (who owns this spectrum) didn’t even mention it in their submission.
DISH’s formal response is due tomorrow (Oct 27), but an ex parte submission and an interview with Charlie Ergen both highlight that DISH is going to make sure that the FCC is under more pressure to conclude this proceeding (and finally take the first step towards its goal of making more spectrum available) than DISH is. Indeed DISH would clearly prefer to wait until a decision is reached on the AT&T/TMO merger, because if the proposed merger is blocked then more potential partnerships would open up. In my view, the most likely partnership would then be with AT&T, which has come close to a deal to buy DISH in the past, and would be able to use both DISH’s 2GHz MSS-ATC spectrum and its 700MHz E-block spectrum (in conjunction with the MediaFLO spectrum that AT&T plans to buy from Qualcomm).
However, the FCC still has to avoid giving DISH a similar spectrum windfall to LightSquared and if DISH is to avoid having to give up part of the spectrum for reauction, it will have to come up with some creative way to reimburse the Treasury financially for any windfall (e.g. offering to pay the difference between what it paid for DBSD and TerreStar and the amount the adjacent J-block 2020-25/2175-80MHz spectrum sells for in any future auction).
Today, it seems that Sprint has rowed back somewhat on its implied threat on October 7 to force Clearwire into bankruptcy, announcing a “non-binding memorandum of understanding to work together” on ensuring that the two companies’ LTE networks will be interoperable. While this is only a limited first step towards Sprint providing concrete backing to Clearwire in 2013 and beyond, it appears designed to allow Clearwire to go out and attempt both to raise new funding and to secure other partnerships ahead of the early December deadline for Clearwire’s next interest payment on its debt. In the short term the most plausible new partnership would be with MetroPCS, which recently said it is “uniquely positioned” to do something with Clearwire, though it is unclear how much of a financial commitment this would actually involve on MetroPCS’s part.
Yesterday Sprint also indicated that it plans to deploy a 2x10MHz LTE Advanced network in its 800MHz iDEN spectrum, which is the subject of today’s MoU on interoperability. By using this additional spectrum for LTE (rather than CDMA as had been previously stated), Sprint would be able to deliver LTE across a mix of low (800), medium (1900) and higher (2500) frequency spectrum bands depending on the population density in a given area, with the Clearwire network providing supplementary capacity in the densest urban areas. If other wireless operators adopt a similar model, then Clearwire could become a wholesaler to multiple major carriers (e.g. Verizon as a supplement to its LTE deployment which is currently at 700MHz and will later extend to AWS), as would be needed if Clearwire is to build a wholesale business on the back of covering at most 100M-150M POPs. Nevertheless, it will take a minimum of several years before a company like Verizon or AT&T would realistically need that additional capacity and so Clearwire still has a lot of work to do (and a lot more money to raise, now that its WiMAX cashflows are likely to diminish) to pull off its transition to a multi-network wholesale provider of urban capacity, not least in creating an ecosystem around TD-LTE in the 2.5GHz band.
However, at least Clearwire appears to have a plausible plan for how to move forward, unlike LightSquared, which has now basically been written off by MetroPCS and (after today’s announcement) implicitly by Sprint as well. On that basis, it looks like Harbinger’s write-down of its LightSquared investment at the end of September will end up being only the first of several such actions. Its also very hard to see how investors in LightSquared’s first lien debt believe that they will be able to realize even 50 cents on the dollar in bankruptcy, when the most likely outcome is that there will be no purchasers for LightSquared’s spectrum at any price, because the regulatory risks will not be resolved and the ongoing Inmarsat lease obligation potentially outweighes the residual value of LightSquared’s spectrum assets.
‘If seven maids with seven mops
Swept it for half a year,
Do you suppose,’ the Walrus said,
‘That they could get it clear?’
‘l doubt it,’ said the Carpenter,
And shed a bitter tear.
In what seems like a reprise of LightSquared’s assertions in July that the FCC would reach a favorable decision in September (and indeed that they weren’t even “planning for the possibility that the FCC denies LightSquared the waiver when the decision comes out in September”), LightSquared is now asserting that the company “expects guidance from U.S. regulators by year’s end”. However, last week I spoke to several knowledgeable people about the testing and regulatory process and all were unanimous in their view that an FCC decision by December was utterly implausible.
The reasons for this become very clear once the current testing process is better understood. Those tests are being conducted by government personnel at an Air Force base (most likely White Sands or Holloman, where the previous round of government live sky testing was undertaken), beginning late next week and running through November 4. That testing is confined to “cellular and personal/general navigation” receivers as specified in the NTIA letter of September 9. As of last week, I understand that there was no intention to test the Javad precision receiver with the new prototype filter, and although LightSquared is trying hard to have that included in the testing process, it is certainly not part of the cellular test plan.
However, even if LightSquared succeeds in its efforts to get some testing done in the next three weeks, it would be essentially irrelevant, because the TWG tested 33 high precision and network GPS receivers in live sky conditions (and a total of 44 high precision and 13 timing receivers in an anechoic chamber) and these would all need to be retested before conclusions can be drawn about how to move forward. Indeed it is extraordinarily difficult just to come up with a test plan on this issue, given that circuit board level integration would likely be needed to incorporate the proposed filter into quite a number of these devices, which will take months of work.
After the current round of tests completes on November 4, it will take some time for the NPEF to write-up the results of that work, and I understand that no further testing is planned by the government in the rest of November. Indeed, it may only be possible to complete a summary of the cellular and personal/general navigation testing by November 30, and a complete report may be delayed into December. Once this is understood, LightSquared’s assertions simply become unsustainable.
Another, more worrying conclusion that emerged from my discussions last week is that the effects of the LightSquared debacle may now range much wider in terms of regulatory impact for the satellite industry. Comparisons were drawn with the ITAR backlash in the late 1990s, which has scarred the satellite industry for more than a decade, as another example of Congress running amok with a technical issue that has evolved into a political soundbite. In this context, DBSD and TerreStar may find it hard to gain a waiver without giving up something significant (such as half their spectrum) and the eagerness of the FCC to defend other satellite spectrum rights may also come under question.
The MSS industry in particular looks relatively vulnerable, as it has seen a significant decline in underlying service revenue growth this year, from an average of 7% growth in 2007-10 to less than 3% growth in 2011. As a result, this cannot be good news for a sector which has seen far more capex than can be justified by future MSS revenue prospects, and may now be faced with an even more challenging path to exploit the spectrum assets which have supported several of those investments.
In October last year I wrote about the supposed spectrum “crisis” and how the FCC’s Spectrum Summit likely marked the point at which the spectrum bubble began to unwind. A year later, many of those predictions appear to be coming true, with most people now writing off LightSquared (and ridiculing Sprint for selecting them as their “preferred” provider of additional capacity).
Indeed I’ve been surprised how investors’ focus is now much more on Clearwire and DISH, whose spectrum holdings are perceived to have some long term value, and its almost become received wisdom that LightSquared’s spectrum holdings may not provide any recovery in bankruptcy, because the FCC is unlikely to grant approval for terrestrial use of the L-band spectrum (at least ahead of the election, and even then, at best, with a long transition period to update or replace precision devices).
Clearwire’s bonds are also trading at levels that assume they will have to file for bankruptcy, and that their spectrum assets will be sold at a significant discount. I was told that the $2.95B of senior secured notes were trading as low as 50 cents on the dollar, with the $500M second lien at around 30 cents and the $730M of exchangeable notes at around 20 cents, which would put the implied value of Clearwire at $1.5B-$1.8B, despite holding more than 40B MHzPOPs of spectrum.
UPDATE: Prices for Clearwire’s first lien debt are now in the mid-70s, putting a value on the company of closer to $2.5B. While the first billion dollars or so of spectrum sales in a liquidation seems fairly assured, the second billion would be less straightfoward and a third billion even more challenging. Clearwire also faces a large upcoming interest payment in December which will likely represent a key decision point for how it should move forward.
Of course, about 60% of Clearwire’s spectrum is leased, and so may not be in much demand from potential purchasers, while the 17.5B MHzPOPs of owned spectrum is not all contiguous. More importantly, the majority of Clearwire’s owned spectrum comes within the FCC’s spectrum screen, and so it is hard to see why a carrier like Verizon would choose to buy Clearwire’s spectrum and thereby potentially impair its ability to acquire more lower frequency spectrum like SpectrumCo’s AWS-1 block in the future.
However, other than Sprint, virtually the only obvious alternative buyer for Clearwire’s spectrum is MetroPCS, which would probably not want to buy more than 2-4B MHzPOPs of spectrum (i.e. 20-40MHz across its existing 100M POP footprint) and would therefore potentially spend no more than about $1B in total. This puts the pressure on Clearwire to adopt the @Home strategy, which would mean filing for bankruptcy well before Sprint has a credible alternative network in place and threatening to cut off Sprint’s customers unless Sprint pays them a great deal more for capacity in the near term. Sprint would then either have to make a much better offer for Clearwire’s spectrum, or pay for transition services until it could transfer the EVO 4G customers to its own LTE network (allowing Clearwire to make money for its debtholders both from offering service to Sprint and from later selling at least part of its spectrum holdings to another player like MetroPCS).
So Sprint’s management team did a great job of living up to Dan Hesse’s introductory admonition in their Network Vision Strategy Update today, when they failed completely to explain either their near term cash requirements (other than to suggest they will be out raising capital “opportunistically”) and how this will be impacted by the iPhone, or how they are going to deploy a competitive LTE network with only 2x5MHz of PCS G-block spectrum (which they admit will run out of capacity by the time the network deployment is complete in 2014). Unsurprisingly, the audience applauded when this plan was described as “ridiculous” by one questioner.
In their presentation, Sprint management took every opportunity to tout their intention to use hosted spectrum from LightSquared over the wholesale access currently available on Clearwire’s network, because of the “better economics” this provides for Sprint shareholders. When pressed on whether they would provide further support to Clearwire, Sprint basically confirmed the hypothesis advanced by Pardus Capital earlier this year, that Sprint intends to drive Clearwire into bankruptcy, in the expectation that Sprint will be able to acquire Clearwire’s spectrum on the cheap. Sprint also refused to address the question of LightSquared’s GPS interference issues, suggesting in one slide that LightSquared was the primary source for incremental capacity beyond the PCS G-block and yet later insisting that LightSquared was “all upside” and they were not assuming that they would get any more money from them.
Of course, other than to indicate that it would not be providing funding to LightSquared (which is hardly surprising since Sprint currently does not have enough money to fund its own network buildout), Sprint had no answer to one perceptive questioner, who asked what would happen if LightSquared failed to raise enough money to pay for the hosting agreement, and highlighted that it is hard to see why anyone would want to invest in LightSquared when they can see what has happened (and by implication what Sprint has done) to Clearwire investors. Indeed one might very well conclude that if LightSquared does get approval from the FCC to move ahead, Sprint expects to do the same to LightSquared and pick up its assets in bankruptcy. Of course, at this point in time such plotting is virtually irrelevant, because LightSquared’s spectrum is not likely to be usable for years to come, if ever.
After this debacle, it is hardly surprising that both Sprint and Clearwire’s stock prices have fallen sharply, and expectations of a Clearwire bankruptcy have heightened significantly. However, though Sprint’s CEO pointedly remarked that “no bankruptcy case involving a wireless company has resulted in a disruption of service”, that has not been true for (fixed) wireless broadband companies such as Teligent and Winstar.
This situation also brings to mind the @Home bankruptcy in the fall of 2001 (my first introduction to US bankruptcy cases), where @Home threatened to turn off its cable modem network, thereby forcing its cable partners to pay a significant premium for “transition services” while they built out their own networks (indeed AT&T’s customers were actually disconnected because no agreement was reached). In addition, AT&T was subsequently sued for its involvement in driving @Home out of business, and settled that suit for $340M in 2005.
Despite the ever louder assertions from LightSquared’s CEO that he is “absolutely confident” that the company “can raise the capital that’s needed” and that the FCC “will approve” its buildout later this year, the end game for LightSquared has now become pretty clear. The company is down to its last “several hundred million dollars” and “isn’t actively raising funds”, because as LightSquared admitted at the Goldman Sachs conference last week, they need an FCC approval before any more money can be raised (and even then would need to raise a staggering $3.5B within the next 2 years!).
With payments to Inmarsat totaling $155M due between now and February and $9M of lease payments each quarter for its other spectrum, not to mention interest on its $1.6B of first lien debt and other costs of operating the company, that money will be gone within the next six months, and given the timeline set out by the NTIA, its all but inconceivable that the FCC will be able to give the approval by the end of the year, as LightSquared asserts (and would require in order to have any chance of raising more money before they have to file for bankruptcy).
As I’ve expected for several months, it appears that LightSquared are therefore planning to sue the FCC for what the company will presumably assert is the unconstitutional “taking” of its property rights, as granted in previous FCC rulings dating back to 2005. However, as one regulatory lawyer put it to me, suing the FCC is not like shooting yourself in the foot, its more like shooting yourself in the head. This is especially true in a situation where there are all sorts of references in the various FCC rulings to LightSquared’s obligations to avoid interference, and the problems could easily have been uncovered if LightSquared had decided to test their system back in 2005, rather than waiting until 2011. Indeed the FCC now seems very likely to simply wait this one out, and point to the need for further tests to justify holding off on any decision, which would potentially deprive LightSquared of any cause of action.
Today we’ve also seen further Congressional pressure building on LightSquared, which will be amplified by the GPS industry’s decision to focus attention on LightSquared’s $10B spectrum windfall that I noted two weeks ago. This letter from Trimble even points out that President Obama’s jobs bill “would require the FCC to recover a significant portion of the value of new terrestrial broadband deployment rights in certain spectrum frequencies that were originally set aside for satellite services either through competitive bidding procedures or spectrum fee authority”:
SECTION 274. REQUIREMENTS WHEN REPURPOSING CERTAIN MOBILE SATELLITE SERVICES SPECTRUM FOR TERRESTRIAL BROADBAND USE.
To the extent that the Commission makes available terrestrial broadband rights on spectrum primarily licensed for mobile satellite services, the Commission shall recover a significant portion of the value of such right either through the [auction] authority provided in section 309(j) of the Communications Act of 1934 (47 U.S.C. 309(j)) or by section 278 of this subtitle.
SECTION 278. AUTHORITY TO ESTABLISH SPECTRUM LICENSE USER FEES.
(B) In addition, the Commission shall, by regulation, establish a methodology for assessing annual user fees and a schedule for collection of such fees on entities holding Ancillary Terrestrial Component authority in conjunction with Mobile Satellite Service spectrum licenses, where the Ancillary Terrestrial Component authority was not assigned through use of competitive bidding. The Commission shall not collect less from the holders of such authority than a reasonable estimate of the value of such authority over its term, regardless of whether terrestrial services is actually provided during this term…
As a result, its now going to be much harder for the White House to defend the FCC’s waiver, which provides even more of a reason for the FCC to simply defer any decision authorizing LightSquared’s network and giving rise to such a windfall. Thus I suspect a lot more of LightSquared’s future will be dictated by what happens at 1 Bowling Green, rather than in any DC Circuit Court litigation against the FCC. More importantly, I strongly doubt that buyers will be lining up to take over LightSquared’s spectrum rights and keep paying Inmarsat’s spectrum lease contract, while the GPS interference situation remains unresolved.
Yesterday, Globalstar announced that it had “placed an order with Thales Alenia Space for an additional six satellites…in addition to the 25 satellites already ordered under the 2006 second-generation constellation contract”. The press release states that:
According to the terms of the contract these satellites will cost approximately EUR 55 million and should be delivered by mid-2013. The fast track delivery, agreed upon when the contract was signed in 2006, is due to the fact that Globalstar has previously purchased EUR 12 million in long lead items that facilitate the prompt manufacture of these six spacecraft. Additionally, the incremental purchase amount is particularly price competitive due to the fact that Globalstar has already prepaid over EUR 53 million as part of its 2009 COFACE financing. Thus, combined with the EUR 12 million long lead items, Thales has already received approximately EUR 65 million for these satellites. Further payments will not be made until Thales Alenia Space initiates the manufacturing process.
This wording is somewhat strange, because Globalstar is indicating that the low cost and fast track delivery are “according to the terms of the contract”, but there is no indication that TAS has agreed to this price and timescale (or even a quote from TAS welcoming this new order). Indeed the order was placed “nowithstanding the previously announced arbitration proceedings” which according to Space News concern “pricing terms for a second batch of 24 second-generation Globalstar satellites that Cannes, France-based Thales Alenia Space had agreed to build under a contract that has since been amended”. This arbitration panel will convene in January 2012 and Globalstar “wants the arbitration panel to oblige Thales Alenia Space to pursue work on the 24-satellite batch under [the] terms of the amended contract”.
Presumably, in the absence of a negotiated resolution of the dispute, the actual manufacturing cost and timeline for these satellites will therefore remain subject to the outcome of the arbitration proceeding. It is possible that because the order is for six more satellites to be “fast tracked”, some people might now speculate that the problem with the momentum wheels on the first 6 second generation satellites, launched last October, is more serious than previously disclosed. However, I suspect that the order timing (at the end of September) is more likely attributable to an existing contract deadline, together with Globalstar’s long planned requirement to supplement their constellation once the eight first generation spares, launched in 2007, reach the end of their life, and in that case Globalstar’s need for additional satellites would not be as urgent.
UPDATE (10/5): An 8-K filing by Globalstar confirms that TAS has rejected the order “because Thales believes that Globalstar has no right to place the order”. The 8-K also describes various amendments to the terms of the credit facility and seems to indicate (by its reference to final in-orbit acceptance of 18 satellites) that (at least for the time being) Globalstar does not intend to accept the first 6 second generation satellites, presumably because the momentum wheel problem is one of the issues under dispute that will be addressed in the arbitration.