Its interesting to see that two papers have recently emerged focusing on the loss of LightSquared’s spectrum “rights”. One of these, by Lenard and White of the “Technology Policy Institute” is little more than a puff-piece for LightSquared, apparently paid for by the company (the bills submitted by LightSquared’s lawyers document their meetings with the “economist” writing a paper for them), though the authors never state that this is the case. This report was released on Tuesday (April 30) as LightSquared attempts to ramp-up its PR efforts once again in the wake of the FCC granting LightSquared the Special Temporary Authorization it has been seeking for the last two months to test LTE in the 1675-80MHz band (apparently a farewell gift that Chairman Genachowski decided to leave on his successor’s desk).
The Lenard and White paper uncritically recites LightSquared talking points, without any understanding of either MSS or GPS services, and simply recommends that the FCC do as LightSquared requests. There is no mention whatsoever of MSS providers such as Inmarsat and Iridium (implying instead that LightSquared and Globalstar are the only licensees of the L-band and Big LEO band spectrum), and bizarrely the paper asserts that for the last 25 years the MSS spectrum “was, for all practical purposes, unused”. Similarly for GPS, the paper simply asserts that the “least cost option” would have been for LightSquared to have been allowed to go ahead, leaving “individual GPS owners” to be “responsible for retrofitting or replacing their GPS devices so that they would work properly”. Of course, no mention is made of the use of GPS in aviation and the timeline and costs involved in changing out safety critical equipment.
A more serious academic paper was issued the previous week by Hazlett and Skorup of George Mason University, which makes some broadly similar points, but comes to a rather more nuanced conclusion. However, this paper also suffers from similar defects, particularly when it dismisses the cost estimates attributed to GPS disruption by asserting that “Simple Coasian analysis establishes this [$10B LightSquared license] valuation as a cap on costs to GPS users” because the “cost of any “harmful effect” is “bounded by the most efficient (least costly) mitigation technology” which would be buying LightSquared’s license. The “cost estimate” of “an astounding $245 billion” (and the similar estimates by the FAA of $60B for aeronautical users) is of course nothing of the sort, it is a loss of social welfare benefits to GPS users if GPS service was rendered unusable. Hazlett and Skorup assert later in the paper that that “easily more than $100 billion in social losses [were caused] by pre-empting the creation of new LTE band” without apparently any thought that their purported $10B cap should obviously bound these losses in exactly the same way (as an aside the $10B number was actually the windfall to LightSquared from the FCC’s waiver, not the value of the licenses themselves). If it does not bound either estimate, then of course Hazlett and Skorup are implying that the social loss of “a new LTE band” is less than the social loss that would be caused by eliminating GPS service (which hardly seems surprising given that there are many other LTE bands available), but the question of what approach minimizes the overall social losses (e.g. requiring many years of transition) is left completely unresolved.
More importantly, Hazlett and Skorup then go on to compare the situation between LightSquared and GPS (which they describe as a “tragedy of the regulatory commons”) with what they suppose was the “successful rationalization of the L-band” achieved between LightSquared and Inmarsat, in the form of the December 2007 Cooperation Agreement, which supposedly resolved the “severe in-band interference problems”. This meant that “the FCC did not test radios, seek more clarity of “harmful interference,” or determine what reliability level Inmarsat’s customers would receive due to potential “harmful interference” from LightSquared’s operations. They trusted the parties to make efficient choices with respect to these concerns.”
Of course, while that interpretation might accord with LightSquared’s portrayal of the Cooperation Agreement, it is far from reality. Up until 2007, independent of the interference disputes, LightSquared’s predecessor (MSV) and Inmarsat had been engaged in a series of disputes over 3MHz of spectrum that MSV had previously loaned to Inmarsat and which Inmarsat was keen to retain for its new I4 MSS services (note that the previous year, when Inmarsat was unable to reclaim spectrum that it had loaned to Thuraya, Inmarsat was forced to suspend certain services in the South Atlantic ocean). By signing the Cooperation Agreement, Inmarsat was granted the right to retain this “disputed” spectrum until MSV raised funding to pay Inmarsat the spectrum lease fees due under the agreement, and if MSV defaulted (which was seen as the most likely option at that time, given the enormous amount that would need to be paid to Inmarsat), Inmarsat would retain the disputed spectrum in perpetuity.
When MSV actually secured substantial further investment from Harbinger, I understand that Inmarsat came to the conclusion after internal testing and analysis, that there was no chance that the deployment of MSV’s network would actually happen, under the parameters agreed in the Cooperation Agreement, because of the interference that would be caused to GPS systems. As a result, Inmarsat decided that even if LightSquared did actually initiate payments under the Cooperation Agreement (Inmarsat was to be paid $250M to cover its costs for fitting filters to its own terminals in addition to other very substantial rebanding and lease payments), there was no point in rushing to develop and fit these filters. Indeed the process required to gain regulatory approval for new aeronautical filters was sufficiently long that approval wouldn’t even be possible until 2013 or beyond. The Cooperation Agreement was cleverly worded: Inmarsat was given absolute discretion to simply replace the affected terminals if LightSquared’s network was ever deployed, and so Inmarsat simply banked LightSquared’s payments, and incurred virtually no costs. Then in April 2012, with the deployment of a new MEXSAT system on the horizon, Inmarsat kept the Cooperation Agreement in place, because of more subtle wording which requires LightSquared to hold Inmarsat harmless from any requirement to give spectrum back to the Mexican government when MEXSAT becomes operational.
All of that hardly seems to be a “successful rationalization of the L-band” as Hazlett and Skorup suggest. It sounds to me more like a sophisticated player (Inmarsat) taking advantage of another operator that is desperate to have some agreement to show to a gullible hedge fund manager, so that he would invest more money to keep them afloat.
I guess at least from that point of view it was successful for MSV’s then-management, because Falcone proceeded to commit a further $600M in the following seven months to complete and launch the MSV satellites, presumably taking advice from his due diligence advisers (who I’m told specialized in analyzing “fast food restaurants”). My guess is that by the time the GPS problems became apparent to Harbinger, all their money had been spent, so all that was left was to persuade other hedge funds to put their money in alongside Harbinger’s and try to keep up the pretense that all was well. After all, LightSquared’s CEO repeatedly told investors that GPS was not a problem, so that more money could be raised, when it was clear to me and others that was simply not the case.
The real question is when LightSquared came up with the end game of litigation or a spectrum swap to escape from this self-inflicted debacle. Was Falcone so delusional that he thought it wouldn’t matter if there was interference to the GPS system? Or was he expecting compensation from the government all along, once Harbinger’s money had all been spent in 2009?
If you’ve been wondering what I’ve been up to for the last couple of weeks (like Phil), I’m pleased to announce that I’ve been hard at work on a new report, released today, on DISH’s bid for Sprint, which discusses DISH’s plans, alternatives and the most likely outcomes from the battle with SoftBank. You can also read some of my thoughts in a GigaOm article (though its not very accurately titled – that wasn’t my choice) or listen to my interview on CNBC last week.
If you are interested in getting a copy of the report, and discussing my conclusions, then an order form and report summary can be found here or please email me for more details. If you’d like to meet in person, then do get in touch: I’ll be in New York next week from May 7-9, on a panel session discussing “Wifi – Friend or Foe of the Mobile Network?” at the Jefferies TMT conference (May 8th) and then at CTIA in Las Vegas from May 21-22, speaking at a Tower and Small Cell Summit panel on “Spectrum Acquisitions and Their Impact on Cell Towers” (May 21st).
UPDATE: A few early comments on the report:
“It was excellent; a really deep analysis. I cannot fault your conclusions and I actually learnt a lot from reading it.” Wireless company CEO
“Read your report this morning. It’s very good…I think your ideas are dead on” National tech reporter
Its interesting to note that Inmarsat has been competing much more aggressively against key competitors in the last few months. First, I’m told that Inmarsat offered a bounty to Telemar to capture Anglo Eastern, a key Iridium Open Port customer with 350 ships, from Globe Wireless, in the fourth quarter of 2012.
Now (perhaps showing a little pique at losing the recent tender for the AT&T Genus replacement contract) Inmarsat is going after Globalstar, with new North American ISatPhone Pro regional voice plans which will start on May 1, and match Globalstar’s recently announced Orbit and Galaxy plans (though without Globalstar’s “double time minutes” promotional offer). Inmarsat is once again offering a huge bounty to service providers for these new signups, equivalent to multiple months of service revenue.
All of these developments suggest that Inmarsat is determined to seek topline growth in its L-band business and is no longer reluctant (as in the past) to explicitly target its competitors with selective pricing, even though this runs counter to Inmarsat’s recent tendency to increase list prices. Of course, it is less clear whether the new deals will be profitable for Inmarsat, given the incentives needed to achieve these sales.
But with Inmarsat’s investors focused intently on whether the wholesale L-band Inmarsat Global business has returned to growth, and apparently willing to overlook the recent significant contraction in margins within Inmarsat’s Solutions business unit (blamed on a transfer of margin from retail to wholesale operations), that might not matter for now. However, if Inmarsat wants to make more acquisitions (and it is hard to see in the long term who else might end up operating LightSquared’s satellites), then regulators might wonder whether industry consolidation could give Inmarsat even more market power.
The FCC’s proposed FY2014 budget, released on Wednesday, now appears to have sounded the death knell for LightSquared’s September 2012 spectrum swap proposal, that it should be granted access to the 1675-80MHz band, in exchange for giving up “rights” to deploy a terrestrial network in the 1545-55MHz downlink part of the L-band. Instead, the budget proposes that NOAA’s radiosondes should be relocated (presumably to the 400-406MHz band which is also used by weather balloons) and that the 1675-80MHz should then be repurposed using “either auction or fee authority” in 2017 to raise $230M.
This news seems to confirm what happened at the February 20 meeting between the FCC and LightSquared’s lawyers, is that the FCC told Latham & Watkins that they were going to propose reallocation of the 1675-80MHz spectrum for commercial use, but would deny the requested spectrum “swap”, so LightSquared would have to bid in an auction just like anyone else. Although the FCC has not yet issued a ruling (perhaps to avoid tarnishing Chairman Genachowski’s recent speeches about his legacy), release of an order may now be quite close, given the public disclosure of the FCC’s intentions in this week’s budget document.
In a statement to Communications Daily (who first reported this issue), LightSquared continued its usual practice of pretending that all is well, asserting that “What happened today is certainly consistent with our filings” because there would be costs associated with any “sharing agreement”. However, even if LightSquared did pay $230M for the spectrum (and I find it hard to conceive that the FCC would simply award the spectrum to LightSquared if other operators want to bid on it at auction), the 2017 allocation date is obviously far too late for LightSquared to proceed with its fantasy of an LTE network buildout.
UPDATE 3 (4/13): LightSquared’s ex parte for an April 3 meeting with Commissioner Pai indicates a different (and more telling) story, going back to LightSquared’s generic appeals from 12 months ago about the “compelling need to find spectrum-based, technical, legal, or other solutions to the current issues” and abandoning the 1675-80MHz proposed “swap”. Notably, this ex parte was not even filed in IB Docket 12-340, relating to LightSquared’s September 2012 request to modify its ATC Authorization, which would have been required if the 1675-80MHz “swap” was part of this discusssion. Instead LightSquared is now apparently going back to its demand for “regulatory approvals” to move forward with its original L-band plan and threatening legal action if that is not forthcoming.
Given the Mexican coordination issues, the implausibility of LightSquared gaining any more spectrum for free, and the potential costs for continuing LightSquared’s Cooperation Agreement with Inmarsat after March 2014, its fascinating to see how much excitement has built in the distressed debt community in the wake of the WSJ’s report last week that Sound Point had been buying up LightSquared’s debt and preferred stock. That is despite the fact that several large holders in the Ad Hoc Secured Group (not just Fortress) have now exited their positions.
I’m told that a target valuation of 130 cents on the dollar for LightSquared’s LP Term Loan debt is now being floated by some people, based on an assumption that Ergen would convert the secured debt to equity. Even if one believes that Ergen will get involved (as opposed to using LightSquared as a bluff to secure a deal with T-Mobile, MetroPCS, CLWR or Sprint), then that seems excessive, given that buying out the TL debt at par plus accrued interest would only give a valuation of 115 cents on the dollar at present, and any bankruptcy auction will presumably look a lot like TerreStar, where DISH bid just enough to buy out the secured debt (including Echostar’s holdings) at par, just as DISH would presumably buy out Sound Point and other holders for cash at par in a LightSquared auction.
After all, DISH has $10B in cash, and potentially not much to spend it on right now. Any DISH deal with T-Mobile would most likely have come before the recent price adjustment to TMO’s MetroPCS bid (not least because it would have made it more difficult for Paulson and PSAM to hold out for alternative suitors), and a counterbid for MetroPCS is harder now that MetroPCS’s stock price has risen. Similarly, a potential deal with Clearwire seems to be slipping away, as it takes more money from Sprint, despite the efforts of Crest to block Sprint’s bid.
What would be the form of a DISH deal with TMO? The four possibilities are:
a) a joint venture (like Clearwire) which wholesales capacity to DISH, TMO and potentially other players
b) a hosting agreement (like Sprint and LightSquared) with TMO taking payment in the form of capacity
c) an investment by DISH in T-Mobile (like Softbank and Sprint) through the contribution of cash and spectrum in exchange for a majority equity stake
d) a full merger of DISH and T-Mobile.
Of these, I think the first can be ruled out, for the same reasons that Clearwire has had so many problems: TMO would take most of the capacity, so would have too much influence, and it would have an incentive to only use the JV capacity after exhausting its own resources. The second can also potentially be ruled out, simply because the deal would have to be more advantageous to DISH than the LightSquared-Sprint agreement: TMO would cover most of the costs and take payment in capacity, not cash, and then DISH wouldn’t need $10B in cash. The fourth option is probably also less appealing (as DISH has hinted previously), because DT might insist that Ergen gives up the dual class share structure which gives him tight control over DISH, and the synergies between DISH’s satellite TV business and T-Mobile’s wireless business are far less compelling than a merger between DISH and AT&T (with its fixed line and U-verse business) would have been. If DISH stays as a separate company then a merger with DirecTV also remains a possibility, perhaps with DISH’s T-Mobile stake being spun off to Ergen if DirecTV buys out DISH.
So the third option sounds like the most plausible, and the question is whether Ergen can purchase a majority stake in a merged T-Mobile/MetroPCS in exchange for $10B in cash and potentially $8B to $10B in contributed spectrum value. This becomes a harder task after the MetroPCS merger goes through and the merged company has a larger enterprise value (estimated at $28B-$30B by BTIG). However, depending on the remaining debt load after DISH’s investment, it still seems like a reasonable objective. Having a public market valuation for T-Mobile also makes DT’s decision about what share of the company DISH would receive a lot more transparent.
Of course, this new development also makes it rather more reasonable to regard a potential DISH bid for LightSquared as another bluff by Ergen, or at the very least a far less attractive fallback option (and if you think a deal might still be done with LightSquared as well as T-Mobile, note that DT is currently embroiled in its own LightSquared-like ATC scandal in India, with some of the original SkyTerra investors). It also confirms that DISH has moved on from its interest in Clearwire, and will most likely just look to cash in its sizeable profits (of up to $300M) on Clearwire’s debt when the Sprint merger eventually goes through.
UPDATE 2 (4/12): Clearwire has filed a proxy statement this evening, noting that it was approached by Party J, a strategic (i.e. non-financial) buyer, earlier this week with an offer to buy 5B MHzPOPs “in large markets” for $1.0B-$1.5B minus the NPV of the associated lease payments. Given the NPV of Clearwire’s lease payments is $1.8B of which I’d guess at least two-thirds is in large markets, and 5B MHz POPs would be around half of the leased spectrum in these markets (assuming this is the top 100M POPs in the US), that would mean a deduction of $600M or more, leaving a net price of ~$400M-$900M.
Walt Piecyk at BTIG thinks that Party J might be Ergen once again and I tend to agree (perhaps this approach is through Echostar?). It would fit with the move on LightSquared last week as another effort to establish a potential backup to T-Mobile, by persuading Clearwire stakeholders to hold out for more money and reject the Sprint bid. Clearwire seems to be hinting that it will file for bankruptcy on June 1 if the Sprint bid is rejected and that would also provide Ergen with far more options in the event that he can’t pull off a deal with T-Mobile, because of his substantial holdings of Clearwire debt. However, given the difficult relationship between Sprint and DISH and the numerous challenges associated with LightSquared, both look far less attractive options than a deal with DT for DISH to invest in T-Mobile.
For those of you not following my Twitter feed (@TMFAssociates), last week there were a couple of interesting developments related to Globalstar, which is currently negotiating with its noteholders under a forbearance agreement (lasting until April 15), after holders of $70.6M of the 5.75% notes exercised their rights to require repurchase of the notes, and Globalstar did not pay the $2M of interest due on April 1.
Firstly, Globalstar has been granted received approval from the FCC for the experimental authority it was seeking for tests of its proposed S-band TLPS service, on March 25 for testing in Cambridge, MA and on April 1 (no joke) for testing in Cupertino and Sunnyvale, CA. Globalstar has also told the FCC that it intends to submit further experimental applications “in the near future”. However, the authorizations are for testing only and are “subject to prior coordination with the Society of Broadcast Engineers”, because both locations are within BAS Channel A10 pickup areas. This coordination has apparently not yet taken place, and because the BAS community is rumored to be very unhappy with the situation, it could take some time to reach an agreement. The timeline for the FCC to issue an NPRM setting out the proposed rule changes to permit commercial use of TLPS also remains unclear, but it seems to be taking longer than originally hoped.
A second recent development is that last week AT&T notified the (less than 1000) subscribers to the TerreStar Satellite Augmented Mobility (SAM) service, that DISH has decided to shut down the service effective May 1, via the letter below, and has “made an arrangement with Globalstar” to offer a discounted replacement rate plan and Globalstar satellite phone.
Of course, those with long memories will recall that TerreStar Networks filed for bankruptcy in October 2010, less than a month after AT&T started selling the Genus phone. Given Globalstar’s current financial challenges, let’s hope that AT&T’s decision to start selling Globalstar service now is not a bad omen for the company.
No I’m not talking about Intelsat’s current IPO! Although some might raise questions about the company’s debt levels, its also pretty clear that Intelsat’s mobility strategy and its high throughput Epic satellites are the envy of its FSS competitors. What I’m actually referring to is Row44, which went public at the end of January through a merger with Global Eagle, which had raised $190M in a blank check IPO. The end of year results from Row44 and Global Eagle have gone completed unreported, but were filed with the SEC last month, and highlighted just what a disastrous business Row44 really is.
In particular, as installations of the Southwest fleet neared completion, Row44′s revenues in Q4 dropped by 27% compared to Q3 (from $20.1M to $14.8M). Moreover, Row44 only generated total Internet connectivity revenues of $11.4M in 2012 while spending $19.6M on buying bandwidth from Hughes, and produced no revenues from its much ballyhooed content and portal services (despite spending $1.9M on video licensing fees).
In 2013, Row44′s satellite connectivity commitment to Hughes will go up to a minimum of $28M (and more likely well over $30M, because of additional commitments made in early 2013 to add coverage in Russia, and additional trans-Pacific coverage planned later this year). Row44 has also changed its deal with Southwest, so it now only receives fees for passengers using connectivity (between $5 and $6 per user, while Southwest charges $8 for the service), rather than Southwest paying for every boarded passenger under the original agreement, and as part of the new contract, Row44 is adding more capacity, apparently to counter passenger complaints about “slow or restricted services”.
At an investor conference on March 13, Global Eagle’s CFO suggested that paid take rates for other providers “range from high single digits to high teens” and that inferences could be drawn from that for Row44′s take rates on Southwest. Of course that is completely untrue: take rates for Gogo are around 5%, and Global Eagle itself said back in November that the targeted take rate on Southwest was 6.5% in 2014, so its hardly likely that the take rate is higher than that today. I therefore have to wonder if Global Eagle’s CFO actually understands this business at all.
Indeed, the company seems to have a hard time predicting the outlook for its business even in the very short term: this March 13 presentation indicated that the 2012 adjusted EBITDA from Row44 and AIA respectively was -$26M and +$18M respectively, whereas on November 27, 2012, with just one month left in the year, Global Eagle forecast that the full year adjusted EBITDA results for Row44 and AIA would be -$25M and +$22.3M respectively.
If we look forward to 2013, then it seems certain that Row44′s revenues will decline sharply, because of the slowdown in equipment installations (from $61M of equipment revenues in 2012 to ~$35M in 2013, and potentially to only $20M in 2014, based on Row44′s current business plan). More importantly, the company is set to make another significant loss on provision of connectivity services: I estimate $20M-$25M in revenues, depending on the take rates that Southwest achieves, compared to connectivity costs of well over $30M and more likely close to $40M (just in payments to Hughes, ignoring Row44′s own operational costs). In addition, content revenue is still minimal (Global Eagle described the product as “brand new” in mid March), with a very limited selection available, and by my estimate content and portal revenues will be no more than ~$5M this year.
In the longer term, Row44′s new deal with Southwest has locked the airline into a deeply unfavorable competitive position vis-a-vis JetBlue, who intend to offer Internet access for free. Southwest can’t possibly do the same, if they have to pay Row44 at least $5 for every passenger who uses the service, despite the two airlines competing very actively on other passenger amenities (like free checked bags). Moreover, if Row44 is ever to break even, it may have to cut back on the amount of capacity it buys from Hughes (rather than boosting capacity as Southwest currently plans), making Southwest’s Internet service even less attractive compared to JetBlue, which uses much cheaper Ka-band capacity.
In recent days, we’ve seen yet another report predicting rapid growth for in-flight connectivity deployment and revenues. However, none of these reports have got to grips with the issue of what services can be profitable for providers but still affordable for the airlines. Of course, if there’s one lesson to be learned from the Connexion-by-Boeing debacle, its that you can’t just assume companies will be able to continue providing service at an enormous loss indefinitely. After scraping through 2012, raising money at effective interest rates of up to 421%, Row44 has escaped for the time being by capturing Global Eagle’s cash balance. Perhaps unsurprisingly, given its management’s apparently tenuous connection to reality, Global Eagle now intends to “use its balance sheet” to pursue consolidation of the in-flight connectivity business, which might provide some relief from Row44′s current unsustainable business model, but on the other hand may simply drag down other companies in the sector.
We’ll be publishing our own detailed report on the in-flight connectivity business next month, which takes a very different approach to the analyses we’ve seen so far: instead of simply projecting how many planes might be fitted out with each technology, we’ll look at both the revenue and cost base of the various IFC providers, and discuss which of them may be able to find a sustainable long term business model.
That’s an interesting question after the WSJ reported yesterday that Sound Point had now acquired over $600M of LightSquared’s $1.7B LP secured debt. Speculation has been rampant once again that Ergen is behind this move, which involved buying all of Fortress’s $161M of LP secured debt (as the WSJ noted – indeed if Sound Point now owns over $600M of secured debt, then it must have bought around $100M of secured debt elsewhere in recent weeks, as Providence only owned $7.4M of this debt), but I was told also Fortress’s $87M and Providence’s $38M of LP Preferred Units. The secured debt was apparently bought at par, and the preferred units at 95 cents on the dollar, and assuming the Preferred units turn out to be the fulcrum security, this $125M will give the owner over 80% of the $164.6M that is outstanding in the LP Preferred class, and potential control over LightSquared’s bankruptcy reorganization plan.
Most people appear to believe that Ergen is behind Sound Point’s investment, despite the doubts that were expressed by people close to the situation last May. The fact that the deal was apparently struck immediately after the end of Q1 (and therefore any public company, such as DISH, would not have to reveal a substantial stake until the second quarter 10-Q in August) might support that possibility: although DISH could not hold the LightSquared LP debt itself due to limitations in the debt agreements (and therefore Sound Point would be the vehicle for this investment), Ergen might well decide to acquire the preferred stock through DISH, to give him a direct voice in proposing a bankruptcy reorganization plan after July 15.
However, it remains an intriguing question as to whether Carlos Slim has any involvement, given the challenging Mexican coordination issues facing LightSquared, which the company has reluctantly acknowledged by hiring a Mexican law firm this week to “ provide legal services with respect to LightSquared’s activities and/or negotiations in Mexico and with Mexican authorities”. After all, it hardly seems likely that Ergen would want to take on the major challenge of making LightSquared’s spectrum usable, without some sort of arrangement in Mexico, if a large part of the spectrum could soon be lost to the MEXSAT system.
Incidentally, further confirmation of LightSquared’s difficult February 20 meeting with the FCC has also emerged in the $271K invoice and February billing records submitted by Latham & Watkins a couple of weeks ago: there was no preparation for this meeting until the day of the meeting itself (implying that the lawyers were summoned by the FCC, rather than arranging the meeting on their own initiative), and the meeting was billed as related to ATC licensing, not to 1675-80MHz sharing issues (which was how LightSquared’s ex parte filing characterized the discussion). More importantly an associate was then immediately tasked to “research and analyze FCC opinions and orders related to the dismissal and withdrawal of applications at the request of applicants”, implying that LightSquared was so concerned about the possibility of an imminent rejection by the FCC that it contemplated pulling its September 2012 application. The lawyers then switched immediately to focus on “draft[ing] legislation” and “talking points regarding license mod advocacy” for LightSquared’s government relations team to push at “[Capitol] Hill meetings”. They also appeared to drop the analysis they had been working on with an “economist”, in favor of exploring the hiring of a new “consultant”, presumably to reflect their new emphasis on trying to go around the FCC.
Given all these issues, another possibility is that a further investment in LightSquared is a relatively cheap signal to T-Mobile that Ergen still has alternative ways forward to make his spectrum assets more valuable, even if the possibility of a deal with Clearwire seems to be fading. After all, as I wrote a couple of weeks ago, T-Mobile could secure a deal with DISH as a way to make it clear to MetroPCS shareholders that staying independent is not a good alternative to the T-Mobile merger. It certainly seemed that a DISH-TMO deal of some sort could be close at that point, but T-Mobile’s event the following week focused instead on its new no-contract pricing plans and initial LTE network launch.
However, after the recommendations by two proxy advisory firms that MetroPCS shareholders should reject the merger, T-Mobile is now considering whether to change the terms of the deal, and one way to do that might be to bring DISH into the picture. Certainly T-Mobile’s partnership offer to DISH ought now to be more attractive than it was a couple of weeks ago, but if it is still unacceptable to Ergen, I could envisage DISH making a bid for MetroPCS before the April 12 merger vote, using the $2.3B it raised this week. In that context, any takeover of LightSquared certainly ought to be seen as a fallback option for DISH, which need not even be considered until the situation with T-Mobile and MetroPCS has been resolved one way or another.
For the last week, since DirecTV’s bid for GVT fell through, rumors have been circulating about a potential merger between DirecTV and DISH. However, to me that seems rather implausible in the near term: while a merger might well be feasible from a regulatory perspective, there would almost certainly be months of jockeying for position in Brazil and elsewhere, before it became clear which of the two companies would be in the stronger position to dictate the terms of a merger.
Instead, if DISH isn’t to be left holding its spectrum assets for many years (with no potential buyer on the horizon, after AT&T and Verizon both declared they would not be buying any more spectrum outside of FCC auctions in the next few years), DISH now has to focus on trying to bring its spectrum plans to fruition in the immediate future. This means seeking either to block the Sprint-Clearwire deal (to gain leverage for a network sharing agreement with Sprint) or to do the same in the T-Mobile-MetroPCS deal, which is now set to close in the next few weeks, unless MetroPCS shareholders vote it down.
I suggested last month that if Clearwire drew down the funding line from Sprint at the end of February, then DISH might “potentially move on MetroPCS immediately thereafter”. While no public offer has emerged, in recent days DISH has sounded noticeably less enthusiastic about its prospects of blocking the Sprint-Clearwire merger (and note that DISH is likely to make a $250M-$300M profit on its investment in Clearwire simply by allowing Sprint’s deal to go through). It therefore seems to me that if DISH does emerge with a deal in the near future (which perhaps might have prompted the 5% rise in DISH’s stock price on Wednesday), then the only plausible possibility is that T-Mobile got worried that its MetroPCS merger would be voted down, and negotiated a network sharing deal behind the scenes with Charlie Ergen.
That would be a great move for T-Mobile, at one stroke removing both potential alternative merger partners for MetroPCS (by taking DISH off the table and ensuring that Sprint is tied up with its Clearwire merger) and also cementing its position as the wireless operator best able to offer large, low cost data packages for the foreseeable future. Of course it would also be disastrous for those hedge funds betting that they can force T-Mobile and Sprint to increase the financial consideration being offered for MetroPCS and Clearwire respectively.
What would such a deal look like? I would guess very much like DISH’s desired deal with Sprint, namely that T-Mobile hosts the DISH spectrum on its LTE network, while taking most if not all of the financial consideration in the form of capacity, providing DISH with a roaming deal plus up to half of the AWS-4 capacity to launch its own wireless services. A deal like that, with no spectrum actually changing hands, would provide no basis for further FCC review of the T-Mobile/MetroPCS transaction, so the perfect time to announce it would be in the next couple of days (now all approvals have been received), giving MetroPCS investors time to contemplate their fate ahead of the April 12 vote.
According to Wikipedia, “In financial circles, the Mexican standoff is typically used to connote a situation where one side wants something, a concession of some sort, and is offering nothing of value”. That’s presumably how the FCC feels about LightSquared’s request for a so-called spectrum “swap”, as I noted last month when the FCC’s General Counsel summoned LightSquared’s lawyers to tell them that there was no way LightSquared would get access to the 1675-80MHz band for free. LightSquared’s lawyers were forced to promise that the company would at least pay for relocation of NOAA’s radiosondes “consistent with the Commission’s emerging technology” doctrine.
The FCC must therefore have been even more annoyed when LightSquared followed up this meeting with a request for an STA to conduct four months of tests to examine radiosonde relocation, and essentially stated that NOAA should pay for the relocation itself. In particular LightSquared asserted that “continued use of this [1675-80MHz] band by radiosondes may not be compatible with the expected operation of the new GOES-R satellite system that is expected to be deployed in the next few years” and so “the modification of current radiosonde operations in and around the 1675-1680 MHz band…is required by the “downshift” of a portion of the GOES-R satellite downlink channels”. Given that NOAA has spent five years studying this “downshift” to free up the 1695-1710MHz band for auction (in that case to raise money for the government) and has never indicated that relocation of radiosondes out of the 1675-80MHz band is required by this downshift, that is quite a remarkable assertion for LightSquared to make, and is all but guaranteed to alienate NOAA as well.
As a result, a cynic might conclude that LightSquared’s attempt to obtain a four month testing window is intended to delay any negative ruling from the FCC until after LightSquared tries to convince the bankruptcy court that they are well positioned to secure the spectrum they need, and persuade the judge to allow them to cram down the LightSquared LP debtholders, on the grounds that Harbinger’s equity is supposedly still “in the money”. Indeed, the FCC may well be tempted to consent to LightSquared’s request, in order to push out any litigation beyond the end of Chairman Genachowski’s term in office.
However, there is yet another, potentially even more serious, Mexican issue looming on the horizon (which will quite possibly emerge before mid July), namely the MEXSAT-1 satellite that Boeing is constructing, which is scheduled for launch at the end of 2013 or early in 2014. As I noted in December 2010, when the MEXSAT order was placed, after the MEXSAT launch, LightSquared will no longer have “any authority to share spectrum with [Mexico's planned next generation] system in the absence of coordination”.
I’m informed that as part of this new coordination agreement, the Mexican government will now demand the right to use one third of all L-band spectrum in North America for the MEXSAT-1 and 2 satellites, based on Mexico’s huge $1B investment in this system. This amount (~22MHz in total) is of course far more than Mexico’s current allocation for its older satellites. However, LightSquared’s Cooperation Agreement with Inmarsat guarantees that Inmarsat will be held harmless from any future reallocation of spectrum to Mexico, and thus all of the additional spectrum demanded by the Mexicans would have to come out of LightSquared’s allocation. That quite possibly explains why Inmarsat wanted to keep the Cooperation Agreement in place last April, while suspending further payments by LightSquared for two years, because now Inmarsat won’t have to give up any spectrum at all to Mexico. It might also indicate why Carlos Slim was rumored to be interested in LightSquared, although given recent developments, Slim might now have bigger issues to worry about in Mexico.
UPDATE (3/21): Here is a description of the MEXSAT system, confirming that it will need at least 15.4MHz of L-band spectrum (2.2MHz per beam, 7 color re-use) to operate as designed.
LightSquared could potentially reject the Mexican request, which would result in a return to the status quo under the original Mexico City L-band coordination agreement from the late 1990s. However, then the L-band would be broken up into small chunks, with the Mexican spectrum (which is largely in the lower L-band where LightSquared wants to operate terrestrially) preventing LightSquared from gaining any contiguous 5MHz blocks of spectrum that could be used for an LTE network deployment. In adddition, LightSquared would have no right to generate any interference whatsoever into the Mexican system from its terrestrial operations.
In other words LightSquared’s L-band MSS spectrum assets now run the risk of either becoming either completely worthless or at the very least being significantly reduced, perhaps to as little as 10-15MHz in total (excluding the spectrum leased from Inmarsat). This of course makes it far less plausible that DISH would be willing to countenance using LightSquared’s spectrum (for uplinks) in conjunction with its own AWS-4 spectrum (for downlinks), as LightSquared’s debtholders hope. After all, harking back to another long ago post of mine, it seems DISH has a pretty good understanding about “which of DBSD/TerreStar, Clearwire and LightSquared” turned out to be “the good, the bad and the ugly”.
ADDENDUM (3/23): Apart from its spectrum, LightSquared’s other assets mainly consist of its two new satellites, SkyTerra-1 (in orbit) and SkyTerra-2 (still on the ground). Many assumed that the SkyTerra-2 could be sold to Boeing, for use in the MEXSAT program (since MEXSAT-1 and 2 are very similar to SkyTerra-1 and 2). However, I’m told that it is now far too late for that, as the MEXSAT-2 build is already well underway, and so it is likely to be even more difficult to raise money from the sale of LightSquared’s satellites.
Another option for LightSquared has always been its supposed litigation claims against the FCC. I understand that LightSquared believes it has some communications in its posession which would make this case a “slam dunk”, and so it expects to succeed in extracting concessions from the FCC, including the purported spectrum “swap”. However, note that the Mexican coordination is formally a government to government agreement (with the FCC at least in theory responsible for negotiating on behalf of LightSquared). Thus, I wonder how much leverage LightSquared currently has with the FCC, when a negative outcome (or simply no progress) from the coordination negotiations would likely ruin LightSquared’s prospects of deploying a terrestrial network in the L-band. In particular, LightSquared’s ATC licenses from the FCC could then potentially become useless, even without further action on the GPS-related issues, based on LightSquared’s acceptance of the FCC’s (separate) March 2010 Mexico re-use approvals. This might even provide a convenient way for the FCC to park the GPS issues, while completely avoiding any liability to LightSquared.
Yesterday, two interesting pieces of news emerged relating to LightSquared’s proposal that it should be granted “shared” access to the 1675-80MHz band in exchange for giving up “rights” to operate a terrestrial network in the 1545-55MHz spectrum block, which is closest to GPS and caused the biggest concerns during the 2011 testing.
The first of these was the GAO report on options to improve receiver performance mandated in the JOBS bill last February in response to the LightSquared debacle. Both this report and an associated paper by the FCC’s Technical Advisory Committee (TAC) published earlier this month suggest that the FCC should do more to explore setting “harm claim” thresholds at which receiver manufacturers could assert that a new service was causing interference. Spectrum bands in which these thresholds could be trialled are identified, but unfortunately for LightSquared, no mention is made of the L-band/GPS band boundary (and the rationale given for selecting a trial band appears to suggest that the LightSquareed spectrum would not be a good candidate for initial experimentation).
Instead, one of the leading candidates is Globalstar’s proposed S-band TLPS service, which perhaps explains Globalstar’s confidence that the FCC will soon move forward with an NPRM, despite the opposition Globalstar faced from the WiFi community, Clearwire and others in response to the TLPS proposal. On the other hand, the fact that such a proceeding would be an experiment to try and determine an appropriate interference threshold may well mean it would still be very difficult for Globalstar to undertake any large scale deployment in the near term or receive approval by the end of 2013 as Globalstar hoped.
The second, and more significant, development was an ex partefiling by LightSquared which documented a meeting on Wed Feb 20 between LightSquared’s regulatory lawyers and the FCC’s General Counsel and the Acting Chief of the FCC’s Office of Strategic Planning. Curiously, this meeting didn’t involve any technical personnel, despite the fact that the “much of the discussion” focused on “the proposed shared use of 1675-1680 MHz spectrum”. That seems to imply that the FCC was focused on the political problem associated with any perceived spectrum “giveaway” and appears to be confirmed by LightSquared’s offer that it would undertake “relocation of NOAA’s radiosondes…in a manner consistent with the Commission’s emerging technology and other applicable precedents”. In other words LightSquared offered to pay for this relocation.
Even that may prove insufficient, given that LightSquared’s proposed “sharing” with NOAA (based on exclusion zones around certain satellite receiving stations, after relocation of the radiosondes) actually involves even less sharing than will be needed in the 1695-1710MHz band, which has now been confirmed by NTIA as being suitable for auction. In the 1695-1710MHz band there will be protection zones around 18 satellite downlink sites, significantly more than the four locations (Fairbanks, AK, Wallops Island, VA, Suitland, MD, and Greenbelt, MD) that LightSquared would have to protect in the 1675-80MHz band.
As a result, it seems that the development of rules for interference limits at the L-band/GPS boundary is unlikely to be a high priority in the immediate future, and there are still major roadblocks to any spectrum “swap” involving the 1675-80MHz band. We will apparently soon see how much LightSquared is offering to pay for relocation of NOAA’s radiosondes, but if this spectrum is deemed just as suitable for auction as the 1695-1710MHz band, then that may be far from sufficient.
LightSquared now has a deadline of May 31 to come up with a plan to emerge from bankruptcy and after July 15 it will lose its exclusivity to propose a plan. However, if the FCC continues along the lines indicated yesterday, the main asset of the estate may end up being its legal claims, whether against Harbinger or the FCC.
