Now that DISH’s attempts to bid for MetroPCS and do a deal with Clearwire appear to have been stymied by T-Mobile and Sprint respectively, the obvious question is what is Charlie Ergen’s Plan C? After all, last week DISH raised another $1.5B in a debt offering, “to be used for general corporate purposes, which may include spectrum-related strategic transactions”.
I’ve wondered if DISH has many options left other than to sell to AT&T, but it now appears that Ergen may have other plans, which are likely to be revealed within the next three weeks. After all, DISH asked the FCC to extend the comment period on the Sprint-Softbank deal until January 21, and DISH has 30 days from the publication of the AWS-4 Order on December 17 to decide whether to protest the proposed license modification.
While it is possible that Ergen could use the $1.5B that DISH has raised to mount a counterbid to either Sprint’s takeover of Clearwire, or T-Mobile’s takeover of MetroPCS, others think he is contemplating use of the money for a potential H-block bid, in order to persuade Sprint to enter into a more attractive hosting agreement. However, there is a far more intriguing possibility, which could explain why Sound Point started buying up more LightSquared debt at precisely the time when Clearwire decided to go with Sprint instead of DISH’s offer. That is an attempt to buy up all of LightSquared’s first lien debt, followed by a battle to oust Falcone when LightSquared current exclusivity (to propose a plan for emergence from bankruptcy) expires at the end of January.
Then DISH could propose in mid January that the AWS-4 uplink spectrum is instead converted to downlink spectrum (in line with a suggestion made by the FCC back in March), and LightSquared’s uplink spectrum would be used to provide an alternative uplink.
That would be logical, because it will be years before LightSquared is able to use its L-band downlinks, and the 1675-80MHz band is unlikely to be given away to LightSquared for nothing (as opposed to being auctioned). It would also make the full 20MHz of AWS-4 uplink spectrum usable for downlinks, and make an H-block counterbid by DISH far more plausible, because the H-block downlink (1995-2000MHz) could be combined with the AWS-4 spectrum between 2000-2020MHz, putting Sprint under further pressure. The FCC might also like to see the risk of litigation with LightSquared being taken off the table, as well as the prospect of higher bids for the H-block, even if the end result was a further delay in deployment of the AWS-4 spectrum.
Why would Ergen choose this risky path, with its inherent delays in buildout, rather than a simple sale of spectrum to AT&T at this point in time? Presumably AT&T has not yet made a knockout bid to buy the AWS-4 spectrum, while if DISH gained additional downlink spectrum adjacent to the G and H blocks, it would be far better positioned to strike a deal with Softbank to provide auxiliary downlinks for Sprint’s LTE network if AT&T doesn’t come to the table.
In addition, because Sprint is tied up with Softbank (and Clearwire) and T-Mobile with MetroPCS, there are few good partnership options available that would enable DISH to pursue a network buildout at the moment. The FCC has given Ergen 4 years for the initial AWS-4 deployment and even if that deadline is missed, the full buildout (to 70% of the population) can be undertaken in 6 years.
As a result, DISH has little to lose by taking some time to explore alternatives, seek to build up even more valuable spectrum assets, and hope that an attractive deal emerges for either a sale or network buildout in a year’s time. If DISH does go down this path, then LightSquared’s bankruptcy case, which has largely fallen off the radar screen in recent months, could be about to get very interesting.
My sources appear convinced that DISH made an informal offer to Clearwire management sometime ago, and that Sprint has been playing catch-up in its recent actions, after assuming for several years that it faced little pressure to buy Clearwire, because no-one else wanted that spectrum. It seems that Sprint reached out to Softbank in the summer, after realizing that it was facing a challenge from DISH, seeking funds to boost its position in the market (and to help acquire spectrum). At that point Sprint also moved to vigorously oppose DISH’s AWS-4 proposal, trying to delay DISH, while it sought an agreement with Softbank.
However, opinions appear to differ about whether Sprint actually wants to buy Clearwire, or is simply trying to spoil DISH’s plans. My guess is that Sprint’s preferred outcome would be for DISH to sell its AWS-4 spectrum to AT&T, allowing Sprint to pick up PCS spectrum that AT&T would have to sell, and Sprint would only later pick up some of Clearwire’s spectrum at an even lower price than is currently being offered. If DISH does achieve a deal with Clearwire then Sprint’s plans would be spoiled on two fronts: it wouldn’t be able to pick up more PCS spectrum (except the H block) in the near term, and Clearwire might not run out of money and fall into Sprint’s arms in the next few years as Sprint apparently hopes. As a result, Clearwire is now playing a central role in an intricate game of three dimensional chess between Ergen and Hesse.
Although we know what Sprint’s current offer to Clearwire consists of (namely up to $2.97 in cash for the remaining equity, assumption of Clearwire’s debt, plus a bridge loan of $800M to accelerate Clearwire’s LTE buildout), it is harder to determine what an offer from Ergen might entail. Nevertheless, considering the objectives of both DISH and Clearwire (in the absence of a compelling take-out bid for the spectrum of either company) may help to narrow down what Ergen’s alternative offer could be.
From Clearwire’s point of view, the near term objectives are to extend the cash runway, find a way to cut down on the costs of the WiMAX network (decommissioning at least half of the existing sites that will never be built out for LTE) and build out the LTE hotspot network at the lowest possible cost. The hope is that by doing all of these things, Clearwire will be able to hold onto (the vast majority of) its spectrum assets, which will become more valuable over the next 3-5 years as an international 2.6GHz band LTE ecosystem emerges.
From DISH’s point of view, the near term objectives are to deploy an AWS-4 network at minimum cost to meet the FCC’s buildout criteria, get into the wireless business sooner rather than later, come up with a residential broadband solution for its satellite TV customers, and perhaps above all persuade AT&T that DISH is serious about moving forward with a buildout (so AT&T will have to purchase DISH or at least pay up for the AWS-4 and 700MHz E block spectrum).
Clearwire and DISH therefore have a clear alignment of interests with regard to Clearwire’s existing WiMAX network: a sale of these assets to DISH would reduce Clearwire’s expenses, raise a significant amount of cash (that wouldn’t have to be used to pay down the first lien debt), provide a network sharing agreement for the LTE buildout, get DISH into the wireless business more quickly and at lower cost, and provide a fixed broadband solution for satellite TV customers (via Clearwire’s original fixed wireless service, perhaps integrated with a satellite TV antenna to extend the range compared to an indoor modem).
My guess is that DISH would likely pay around $1.5B for Clearwire’s WiMAX network (which cost roughly $4B to build), which might therefore fall below the 20% of asset value cutoff point at which a sale would require approval by Clearwire’s board. DISH would likely also acquire Clearwire’s retail WiMAX customer base and presumably also provide service to the wholesale WiMAX customer base (adding another point of leverage over Sprint) – perhaps DISH would pay up to an additional $500M for these customers.
What is more uncertain is what would happen about the spectrum required for DISH to run the network. Clearwire has a strong interest in establishing a high valuation benchmark for its spectrum, likely $0.30 per MHzPOP or above, and DISH would also want to ensure that perceived spectrum prices are high, if it still hopes for a knockout bid from AT&T. DISH likely needs 20-30MHz of spectrum, covering perhaps 200M people, implying that DISH might have to spend at least $1.8B to $2.7B if it was to buy this amount of spectrum from Clearwire. On the other hand, DISH might opt for primarily leased spectrum, reducing the price somewhat, or simply agree say a 5 year lease with Clearwire, perhaps with a fairly significant prepayment.
Overall, I could envisage DISH paying Clearwire anything from $2.5B in the near term (based on a spectrum lease), up to perhaps $4B+ (assuming a reasonably significant spectrum purchase). Part of this payment would presumably be made by contributing the substantial amount of Clearwire debt already owned by DISH (which cost $750M), and going forward Clearwire would then presumably have a network sharing deal with DISH, so that Clearwire could rollout its LTE hotspots in urban areas and DISH could roll out wide area coverage in the AWS-4 band.
I assume the remaining Clearwire debt would be refinanced, allowing the equity holders to pursue their bet that Clearwire’s spectrum will increase in value. Of course that may not be an expectation that Intel, Comcast and Bright House necessarily share, so Sprint obviously hopes it can persuade these strategic investors to block a deal with DISH. If not, then the question remains, will the DISH deal go through, or will it be derailed by a knockout bid from another party: either Sprint paying the $5+ that shareholders are demanding for Clearwire or AT&T buying DISH for $80 per share?
[UPDATE 12/17] So Sprint has convinced the Clearwire board to accept its offer of $2.97, and it looks like Hesse may have come out on top in this round of the chess game. The determining factor appears to have been the lack of confidence from Clearwire’s management and board that there would ever be a second major wholesale customer for its TD-LTE rollout, as Verizon, AT&T and T-Mobile weren’t interested in buying capacity from Clearwire. In the presentation discussing the deal, Clearwire confirmed that there had been another “credible, but preliminary, proposal” in the “past several weeks” presumably from DISH, but all potential options for spectrum sales had “values well below those recently speculated”. Clearwire also noted that the “existing governance arrangements” left the company “unable to secure new partnerships”. Of course, the deal locks up Clearwire pretty well, because the $80M per month of financing that is being advanced by Sprint is convertible to equity at only $1.50 per share, and Sprint is not obligated to go through with the purchase of Intel, Comcast and Bright House’s stakes if the acquisition is rejected in the independent shareholder vote (but might then come back with a lower bid).
So now Ergen appears to have struck out twice in his attempts to enter the wireless market, being rejected by both MetroPCS and Clearwire. Will he follow the FCC’s signal and sell his spectrum to AT&T? If so, the price may not be as attractive as many hope: if there are few other options then my earlier estimate of $0.30 to $0.40 per MHzPOP sounds closer to the mark than the inflated $1 per MHzPOP speculation we saw last week. Those expecting a higher bid for Clearwire were banking on a spectrum crisis forcing other operators to make use of Clearwire’s capacity in the next few years, which Clearwire management clearly didn’t believe would happen. Now the question appears to be whether Ergen still believe his spectrum will become more valuable over time, or if instead he will just take the money as Clearwire did.
Today Sprint made an SEC filing stating that it had offered to buy the remaining shares in Clearwire for $2.90 per share, far below the $5+ that some think Sprint will have to pay to buy out the hedge funds that have invested in Clearwire with an expectation that the company’s spectrum is hugely valuable. Of course, as many have noted, Clearwire’s spectrum is seen as very desirable by Softbank, and could enable Sprint to rollout a much higher capacity urban LTE network, given the huge amounts of spectrum that Clearwire holds.
However, the underlying story here is far more complicated: why has Sprint made such a low offer, and why now? After all, even if the remaining strategic investors (Comcast, Intel and Bright House) are willing to take the $2.90 and run, they wouldn’t be able to command the majority of non-Sprint-affiliated holders of Clearwire’s equity needed to approve the deal. It appears that one (possibly the key) motivation is that if the strategic investors do accept Sprint’s offer then this would potentially block any alternative deals by Clearwire for several months (until the Softbank deal closes), as according to Sprint’s filing:
“Under the terms of the Sprint Proposal, each of Comcast, the BHN Entities and the Intel Entities would enter into a voting and support agreement with Clearwire with respect to the Proposed Transaction, which will provide that such Equityholders will vote (i) all voting shares of Clearwire owned by such Equityholders in favor of adopting the definitive merger agreement and approving the transactions contemplated thereby and (ii) against other acquisition proposals. These voting obligations would apply whether or not the Proposed Transaction is recommended by the Board of Directors of Clearwire. The voting and support agreement would also provide that such Equityholders would provide any waivers and consents needed to effectuate the Proposed Transaction.”
It’s surely not a coincidence that this offer came on the same day that Charlie Ergen received approval from the FCC for terrestrial use of his AWS-4 spectrum, with DISH stating that it “will consider its strategic options and the optimal approach to put this spectrum to use for the benefit of consumers”. After all DISH appears to have had a potential deal with Clearwire on the table for several months, held up only by delays in the FCC’s approval (which were largely caused by Sprint’s intervention via efforts to gain access to the H-block), and has bought $750M of Clearwire’s debt. The existence of other offers also seems to be confirmed, at least obliquely, by Clearwire’s indication that the Special Committee of the Clearwire Board of Directors, was “previously formed to review potential indications or proposals, including from Sprint”. However, DISH’s offer to buy some of Clearwire’s network and spectrum assets may not provide much if any near term value for the equity investors, perhaps making the strategic investors think that Sprint’s “bird in the hand” will be a better bet.
So what is Ergen up to? Rumors have persisted that Carlos Slim is allied with Ergen and that he has also been looking closely at Clearwire in recent weeks, while I’m told that today Sound Point has suddenly started moving aggressively to buy up more of LightSquared’s debt. Could Ergen and Slim be trying to acquire spectrum across multiple bands, by purchasing assets from Clearwire and positioning themselves to use some of LightSquared’s spectrum after a resolution (in the distant future) off GPS interference issues?
I’ve wondered recently whether Ergen has little choice other than to sell DISH’s spectrum to AT&T because there are few alternative partners available. However, it seems that Ergen may not yet have given up on a deal with Clearwire, making Sprint ever more desperate to block it. If Sprint is now able to tie up Clearwire for several months, then that would probably delay any deal beyond Ergen’s window for making a decision about how to proceed. Sprint has also been spreading rumors of its own potential deal with Ergen as a way to pressure Clearwire’s strategic investors to accept the offer from Sprint: the story was reported by Bloomberg last Friday, but first emerged somewhat earlier, just when Sprint delayed its proxy filing so as to negotiate with Clearwire.
UPDATE (12/14): Reuters is now reporting that Clearwire “is also in talks about other strategic alternatives besides the Sprint offer”, apparently confirming that an alternative deal is on the table. Its hard to see what that could be other than an offer from DISH. Reuters also notes that Softbank is capping Sprint’s offer at $2.97 per share, the same as paid for Eagle River’s stake, which indicates that Softbank is certainly not willing to pay whatever it takes to buy Clearwire at this point in time.
Some have asked me what is Sprint and Softbank’s alternative to buying Clearwire. My view is that Sprint will save its money for buying PCS spectrum, where its need is far more urgent. In particular, Sprint is going to have to pay $1B+ to buy the 10MHz H block in the auction next year, and if DISH is left with no alternative other than to sell out to AT&T, Sprint would expect to pick up another 10-20MHz of PCS spectrum that AT&T would need to sell in order to get a DISH deal approved by the FCC. What is key to understanding this strategy is that Sprint wins simply by blocking DISH, regardless of what happens to Clearwire, because if DISH has no alternative other than to sell the AWS-4 block to AT&T, Sprint will be able to buy PCS spectrum from AT&T.
If Clearwire’s strategic investors don’t take Sprint’s bait then would DISH be able to pull off a deal with Clearwire? Even then would Ergen follow through and build a wireless network? My guess is that the answer very much depends on how much AT&T is prepared to pay for the AWS-4 spectrum, but its certainly not yet time to take it for granted that Sprint will end up acquiring Clearwire in the next few months.
On November 13, Globalstar submitted a Petition for Rulemaking to the FCC seeking permission to use its spectrum for terrestrial services, without any of the restrictions imposed under the current ATC “gating requirements”, and shift its authorization to Part 27, as used for other standard terrestrial mobile services.
The petition envisages two parallel rulemakings, the first to consider designation of Globalstar’s downlink spectrum (Globalstar refers to this as the “upper Big LEO band”) as an AWS-5 band, permitting flexibility for any wireless service to be offered, such as TD-LTE similar to Clearwire’s planned deployment in the adjacent BRS/EBS band. Globalstar also suggested that it be permitted to offer a Terrestrial Low-Power Service (TLPS), which would effectively be a separate channel for licensed WiFi service, using both Globalstar’s upper band spectrum and the adjacent unlicensed spectrum.
This would only be one possible option under an AWS-5 designation, but what is pretty smart about the low power TLPS service (similar to WiFi use, which already overlaps with these BAS channels) is that Globalstar may not have to relocate legacy BAS users who currently operate in the 2450-2500MHz band and could be impacted by a new wide area high power network deployment (as was seen with some of Open Range’s towers, due to lack of coordination on Open Range’s part). However, it is possible that even if the FCC permitted the TLPS service to begin immediately, it might require further actions to be taken (or impose other coordination requirements) before full flexibility was granted in the 2483.5-2495MHz band.
UPDATE (12/11): I’m told that legacy BAS users do experience interference from existing WiFi channels above 2450MHz, but that to date the FCC has not taken action to address concerns about interference from unlicensed spectrum users. Whether this will have implications for quick authorization of TLPS is unclear.
The second rulemaking that Globalstar envisages would then extend the AWS-5 designation to include Globalstar’s uplink spectrum (which it refers to as the “lower Big LEO band”), thereby enabling FD-LTE across the whole of the Big LEO band, which Globalstar considers to be the “highest and best terrestrial use” of this spectrum. The reason for separating the two requests is that the lower band (uplink) spectrum is close to the GPS band, and so the FCC is likely to be cautious about permitting terrestrial services in this band after the LightSquared debacle. As a result any authorization of high power LTE usage in the lower band spectrum (even though it would be for uplink only) would require considerable testing and therefore it would take some time before any approval could be granted.
If granted permission to provide TLPS by the FCC, it appears that Globalstar would look to monetize the TLPS offering (prior to gaining authorization for a standard LTE service) by providing spectrum for a small cell buildout, most likely by a major wireless carrier, but possibly by a tower company or technology player instead. Though such a buildout could be undertaken with existing licensed spectrum (e.g. Clearwire’s 2.5GHz band), or new unlicensed or shared spectrum (such as TV white spaces or the 3550-3650MHz band that the FCC intends to auction for shared usage), Globalstar’s advantage is that WiFi capability is built into the vast majority of smartphones, and Globalstar estimates that its licensed channel would be expected to offer around three times the range and speed of similar access points in the existing uncontrolled WiFi spectrum. However, Globalstar would need to move quickly to take advantage of the installed base of WiFi-capable devices, before capabilities to use longer range unlicensed spectrum (White Spaces) or other licensed small cell bands become widely available in smartphones.
As a result, Globalstar would need both quick action from the FCC and to strike a partnership (or long term spectrum lease) in 2013 or early 2014 enabling rapid deployment of a small cell network. In that regard, the fact that the FCC has acted much more quickly to put Globalstar’s proposal on public notice (2 weeks) than the recent LightSquared petition (which took 6 weeks) suggests that the FCC may well consider Globalstar’s proposal with rather more urgency. This certainly marks a significant turnaround in Globalstar’s relationship with the FCC, which was rather difficult (to say the least) back in September 2010, when the Commission suspended Globalstar’s ATC authority.
Globalstar believes that because of the availability of an existing WiFi device ecosystem, its spectrum should be more highly valued than alternative small cell spectrum, such as that owned by Clearwire. Indeed, Globalstar apparently considers that the future possibility of using LTE within its spectrum band could make this spectrum worth even more than the $0.20 to $0.30 per MHzPOP valuation seen in recent transactions such as NextWave’s WCS spectrum.
However, that is very dependent on a major cellular operator deciding to choose Globalstar as the solution for a small cell rollout (as well as future LTE licensing), and it remains uncertain what will happen to the value of “small cell” spectrum in the next year or two, as more spectrum is brought to market (a rulemaking on the 3550-3650MHz band will be considered at the FCC Open Meeting later this week), especially if data traffic on existing LTE networks grows more slowly than expected. Some think that the value of the 3550-3650MHz band will be very low (perhaps $0.01/MHzPOP or less), as has been seen internationally, which could lead operators to decide that putting a multi-billion dollar valuation on Globalstar’s spectrum for use in a TLPS service would be totally ludicrous.
Clearly the potential value of Globalstar’s spectrum is a critical component in securing investors for the new financing that Globalstar is trying to complete in the early part of 2013, because even though duplex revenues are slowly starting to recover, growth in the SPOT business, which has carried the company for the last few years, has recently fallen short of (at least my) expectations. Globalstar needs to raise money to pay for its EUR150M contract with Thales Alenia Space (TAS) to build a further six satellites, plus the launch and insurance for these satellites. At some point Globalstar will presumably need to complete its second generation ground segment upgrade contracts with Hughes and Ericsson, and Globalstar also has to address the April 1, 2013 deadline when $71.8M of Globalstar’s 5.75% convertible notes may be redeemed for cash at the option of the holders (Globalstar stated in its 2011 10-K that it assumed these notes “will be refinanced in 2013 by issuing additional debt”).
Much of the upcoming capex program would presumably be funded by an increase in Globalstar’s COFACE loan facility (which typically would cover 85% of the costs, although it is unclear if this would relate only to work carried out by French companies). It is worth noting that the FCC licenses do not form part of the security package for the COFACE loan, which could make Globalstar’s fundraising easier, if additional funding (from non-COFACE sources) was secured against the FCC license subsidiary. However, if Globalstar does not succeed in raising the required funds, this also poses the question of whether (in the event of a bankruptcy) the existing convertible note holders could sell the spectrum licenses and receive a recovery, even if the COFACE loan is not paid off in full by a sale of the satellite assets. Such a possibility may complicate negotiations over the upcoming refinancing of the 5.75% convertible notes.
More importantly, it will be very important to see whether the FCC follows the DISH model, and simply grants Globalstar a separate terrestrial license (alongside its satellite license) which could be monetized at a later date (regardless of the ultimate fate of Globalstar’s satellite system), or if it follows the LightSquared path, where (because LightSquared is operating under an ATC waiver) the L-band MSS spectrum cannot easily be disentangled from the continued provision of satellite services. After all, past bankruptcies in the MSS sector have shown how hard it is for creditors to achieve a large recovery from billions of dollars invested in MSS satellite hardware, and at least in the case of DBSD and TerreStar (albeit in a situation with no meaningful existing satellite business, unlike Globalstar), the in-orbit satellites were seen as a potential cost associated with obtaining the spectrum licenses, rather than valuable assets in their own right.
No not his assertions that “I still believe that we had a great, and we still do have a great vision” and that “a bankruptcy would not necessarily wipe out the equity holders of LightSquared because the spectrum it owns retains value”, which look like they are moving even further away from being realized as time goes on.
LightSquared’s latest plan is to relinquish terrestrial rights in part of its L-band spectrum in exchange for the right to use additional government spectrum used by NOAA. However, I’m told that although NOAA may be prepared to allow terrestrial use of the 1675-80MHz spectrum band (subject to payment of relocation costs to move their radiosondes to the 401-406MHz band), LightSquared’s plan may now be derailed by the apparent intention of Congress to mandate an auction of “at least 15MHz” of spectrum in the 1675-1710MHz band, thereby requiring LightSquared to buy this additional spectrum at auction rather than being given (what they are portraying as) a “swap”.
After all, most Democrats appear to have concluded that “What happened to LightSquared is disappointing…But unfortunately that ship has sailed”. NOAA would presumably also like to maximize the auction proceeds to fund the relocation of its other systems in the 1695-1710MHz band which will also be auctioned as (lower value) uplink spectrum. As a first step, NOAA’s position may become clearer next week once the initial comments are filed in response to LightSquared’s petitions for rulemaking.
What I’m actually referring to in the title of this post is Phil’s comment on my blog post back in August 2011 that “Everyone knows Ergen is not going to build out a network. No one trusts him, including the FCC. They are not going to put their eggs in that basket because they know he will make them look foolish.”
That does appear to be a pretty good summary of the underlying rationale behind the FCC’s draft AWS-4 Order, circulated just before Thanksgiving, because the Order does not impose any anti-windfall conditions that would stop DISH from selling the spectrum to another operator, but instead attempts to mitigate this windfall (a concern the FCC is acutely sensitive to) by ensuring that the proceeds of an H-block auction can be maximized. If sold to existing operators, then the 50MHz of spectrum in AWS-4 and H-block combined would help to preserve the current four player wireless market in the US and would also mean that (as the FCC Chairman’s former advisor put it this week) “we’re not in the rush we originally thought” to address the purported spectrum crunch.
Despite this encouragement to cash in, DISH objected to the draft order, with Ergen appealing in person to each of the five Commissioners to make changes. However, the FCC fired back at DISH and has now scheduled the order for a public vote at the meeting on December 12, sending a signal of unanimity to DISH, and perhaps also discouraging any Commissioners who might be wavering from taking a position that would increase DISH’s potential windfall at the expense of H-block auction proceeds for the Treasury. Sprint has maintained its opposition to the latest changes proposed by DISH, and appears to have been successful in fending off any modification to the original draft (if Sprint was losing ground then I would have expected to see more vocal objections, as opposed to the current strategy of providing behind the scenes briefings and a rather non-specific ex parte filing).
Thus I’m left wondering whether DISH will now be persuaded that its best interests lie in selling out (especially given the diminishing number of available partners), or if instead DISH will continue to fight with the FCC (and challenge the legality of the Order as outlined in last week’s ex parte filing), potentially holding up the deployment of both AWS-4 and the H-block?
Could Sprint have designs on the AWS-4 spectrum as well (after all, Ergen is known for fighting vigorously right up to the time he does a deal with an opponent), or will it simply want to take advantage of disposals that AT&T would be forced to make after buying DISH’s spectrum? We should know more about Ergen’s intentions very soon, and Sprint’s proxy filing, now delayed until December 21, should make even more interesting reading than that of MetroPCS, because it will have to reveal if discussions have been going on with DISH behind the scenes over any potential deals.
The latest Global Eagle investor presentation, which was filed with the SEC on November 27, makes for some pretty entertaining reading, much like the script for one of those Hollywood disaster movies that Row 44′s new owners may be fond of. Of course it requires some pretty convoluted logic to argue that Row 44, which is selling equipment at cost and losing money on every megabyte of data carried across its network, is worth $250M.
That’s especially true when the company fully intends to continue to sell equipment at cost and intends connectivity to be a “lower margin” service. In addition, despite a new contract with Southwest, which is expected to lead to a near 50% increase in connectivity revenue per passenger by 2014, the connectivity service is projected to produce less than $5M of gross margin (not EBITDA) that year. Instead, Row 44 and Global Eagle “believe the next frontier for growth will be providing quality entertainment, vast entertainment, first through the airlines into a projected multi-billion-dollar marketplace in the air” and so are projecting that Row 44 will make $49M of gross margin in 2014 from “TV/IPTV/VOD” and portal services.
What’s wrong with this picture? At its simplest, Row 44 is projecting that the take rate for connectivity on Southwest will grow to 6.5% (not outrageous, but perhaps ambitious given recent trends for Gogo and the predominantly leisure orientation of Southwest’s customer base) and the take rate for the “TV/IPTV/VOD” service will be 5.75% in 2014. Recall that passengers are paying $5 for internet connectivity on Southwest, which is the same as Row44′s assumed $5 per passenger fee for TV/IPTV/VOD. That’s all well and good if this was seatback IFE, available (and very visible) to all passengers with attractive early window content (very few people pay for live TV onboard planes in the US today), but remember that the only way to get access to the “TV/IPTV/VOD” service is through an Internet-enabled portable device.
Why on earth would almost as many passengers decide to pay (the same amount) for access to just the walled garden “TV/IPTV/VOD” service as pay for access to the rest of the Internet (where they might expect to have access to any content they choose – unless Row 44 decides to cripple the Internet service, which of course will lead to its own problems)? Will Row 44 have some incredibly compelling early-window content? Well studio executives apparently “Lol’d” when asked about whether they would allow early window content to be streamed over wireless IFE networks. And remember that while you might happily access email and social media on your iPhone, its far more difficult to watch long form video on a phone for an hour or two, so in reality even less devices may be suitable for watching the content services than for accessing the Internet.
Row 44 also intends to generate $19M ($0.15 per passenger carried) from its portal business by 2014, even though it has “only nominal revenue from the portal business today”. As another point of comparison, Row44 expects to make more than half of its passenger revenue from the video and portal services by 2014, when today Gogo only generates 2% of its Commercial Aviation revenue from “Gogo Vision, Gogo Signature Services and other service revenue” (i.e. video and portal services combined, plus other services such as VOIP for flight crews).
Fundamentally, I simply can’t understand why on earth Global Eagle think that this business is a good investment (though lack of understanding might be one reason). The history of inflight connectivity is littered with failures, and even Gogo, the market leader, is facing challenges in getting a return on its investment, let alone completing a successful IPO. In every respect, Row 44 is a worse business than Gogo: it has substantial ongoing bandwidth costs, far more expensive equipment, and (in Southwest) probably the least attractive airline in the US from the point of view of demand (few business travelers, short flights and no power outlets). I can only conclude that, as in the picture above, Global Eagle is suffering from the Icarus Syndrome, and flying too close to the sun for its own good.
Apologies for the lack of posts over the last couple of weeks – I’ve been buried in writing my latest MSS industry report, which is bigger and better than ever, and includes not only all the latest MSS industry developments such as an analysis of Inmarsat’s investor day, but 30 pages on everything you want to know about the current spectrum issues involving DISH, LightSquared, etc. I’ll be writing blog posts about that plus some of the latest inflight connectivity developments over the next few days, but I’ll start with a little noticed fact that emerged while I was analyzing MSS subscriber growth: surprisingly enough, the various MSS operators use very different definitions for what they count as a subscriber.
Now you might think that a subscriber is simply someone who is paying the operator for service (perhaps indirectly via a distributor) and if the customer is paying for x terminals, then the MSS operator will report that they have x subscribers. That is basically what Iridium do, now that there is a charge each month even for suspended terminals. However, until recently Inmarsat didn’t have a monthly access charge for most terminals, and only got paid for airtime. As a result, Inmarsat has always defined its subscriber count as terminals that have accessed the network in the last 12 months. Now that Inmarsat is charging monthly fees for most services, this leads to anomalies such as in its 2012Q2 results, where Inmarsat noted that:
“At the time of our consolidated financial results for the three months ended 31 March 2012, we announced having reached over 55,000 IsatPhone Pro subscribers. However, in our reported active terminals for land mobile, we included a lower number of approximately 49,800 terminals, the difference being the elimination of subscribers who had not used their IsatPhone Pro terminal in the preceding twelve months…”
Even more significantly the number of Satellite Low Date Rate (M2M) terminals reported by Inmarsat has declined quite noticeably over the last year, but as far as Inmarsat’s distributors like SkyWave are concerned, the number of subscribers is actually going up. However, once you realize that a key application for ISatM2M is stolen vehicle recovery, its pretty obvious that only a small proportion of terminals (i.e. those cars that are actually stolen) will need to access the Inmarsat network each year.
That’s a positive for Inmarsat, because their market share in the SLDR/M2M sector is actually quite a bit higher than many assume. However, Globalstar’s counting methodology goes the other way: SPOT customers are included in the published subscriber count even if their terminal is “suspended” for non-payment, because those terminals still have access to the network and Globalstar is attempting to collect payment for the service (although of course no revenue is actually being recognized for those subscribers unless and until collection occurs). The number of suspended SPOT subscribers has increased consistently since this statistic was first reported in early 2010, and by 2012Q3 amounted to 29% of SPOT subscribers. I’ve generally been pretty optimistic about the long term potential of the personal tracking market, but worringly, in the third quarter of this year the number of paying (i.e. non-suspended) SPOT subscribers actually fell from the previous quarter for the first time ever.