The twisted saga of US telco M&A over the past few years has centered on one issue above all, the hunt for mobile broadband spectrum. AT&T’s aborted bid for T-Mobile USA and the latter’s subsequent merger with MetroPCS; a host of acquisitions of small cellcos or individual licences by the big two; Verizon’s deal with the group of cablecos; and now the Sprint/Softbank/Dish/Clearwire quadrille ‘ all are primarily motivated by the need to gain capacity, competitive advantage and economies of scale via superior spectrum holdings.
There is a sense of panic about spectrum in the US, echoed by the FCC itself. Under outgoing chief Julius Genachowski the regulator set itself the target of opening up 500MHz of additional bandwidth for mobile broadband, and has cast far and wide for sources, including broadcast and licence-exempt bands, TV white spaces, and those which could be shared with federal agencies. His successor designate, Tom Wheeler, is likely to intensify those efforts, and indeed, acting chairwoman Mignon Clyburn (standing in until Wheeler’s appointment is confirmed by the Senate) assured the CTIA Wireless conference this week that the FCC was not slowing down its quest for spectrum during the interregnum. ‘The FCC wireless agenda remains focused,’ she said. ‘The national broadband plan put the issue of spectrum on the map. And the agency is doing a lot to make more spectrum available’, including, of course, the upcoming incentive auction of broadcast frequencies.
One of the bands which Genachowski initially seemed to regard as low hanging fruit was the underused mobile satellite (MSS) spectrum in the L-band (1.5GHz) and the renamed AWS-4 (formerly S-band) area of the 2GHz spectrum. Successive generations of providers had failed to deliver a convincing business case for these services outside niches like emergency response and ultra-rural broadband, and even where those offerings were running, there was plenty of bandwidth going to waste. However, the MSS idea proved to be poison fruit in the case of LightSquared, the venture to build a wholesale LTE network in the L-band, supporting both satellite and terrestrial services. The FCC rushed to grant a waiver, removing the usual requirement that all services should include a satellite component, the limitation on previous initiatives to introduce cellular networks to these frequencies (previous regulatory changes, and technologies like ATC, had opened the way to hybrid satellite/terrestrial options, but this was the first time terrestrial-only was permitted in these airwaves).
The failure of LightSquared, because of concerns about interference with nearby GPS services, highlighted the dangers of acting too hastily in the race for new spectrum, and put a damper on the excitement about MSS bands providing a ripe and relatively uncontroversial source of prime frequencies. However, it did not turn operators or the regulator off the potential of satellite spectrum for long, as seen in the waiver granted to satellite-TV giant Dish to run LTE-only services in its own MSS frequencies (obtained from two bankrupt MSS carriers, and a safe distance from GPS).
Moreover, Genachowski recently said he still believes the L-band will be usable in future, with technology changes to protect GPS. No friend to satellite, his comments seemed to downplay the complexity of those changes, highlighting the retiring chairman’s firm belief that all satellite (and broadcast) spectrum is underused and should be reallocated for, or shared with, terrestrial mobile broadband services. For instance, the Satellite Industry Association responded angrily last month to his proposals that the 3.5GHz C-band should be allocated to small cells (though these would have to coexist with existing users). The SIA said it had concerns over ‘whether introduction of small cells in the 3.5 GHz band would be compatible with current and future satellite operations’ and added it had ‘significant disagreement on virtually every aspect of the regulatory framework’ proposed by the FCC, and expressed ‘doubts about whether the proposals for shared access to the band are feasible, particularly in the near term’
But the rising interest in harnessing these frequencies is growing and is not confined to the US. The European Union has initiatives focused on broadband in the L-band, and on using it for supplemental downlink, to add download capacity to conventional LTE networks (this is being trialled by Orange in Toulouse, France). And when Samsung recently announced an experimental breakthrough in ‘5G’ technology, it achieved 1Gbps transmission speeds using a more far-out satellite option, the 28GHz Ka band, mainly used for fixed satellite uplink. While most LTE-over-satellite projects so far have focused on frequencies close to the cellular ones, between 1GHz and 2.5GHz, this demonstration highlighted the potential to use higher ‘millimeter wave’ bands, already in the spotlight for backhaul, for small cell access.
All this activity, and the enthusiasm of regulators to colonize satellite spectrum for 4G, tends to imply that the mobile satellite communications business itself is a dying one. Eventually, it seems, its airwaves will be largely conquered by radio, as cellular technologies get better at expanding into the rural and ad hoc applications where satellite broadband has thrived. Earlier this month saw a Qualcomm proposal to the FCC, to use the Ku band spectrum for a radio-based air-to-ground system which could compete in another key preserve of satellite, in-flight services.
But is this ‘radio beats satellite’ scenario really the only one in broadband access? For instance, set the initiatives above against that is the imminent launch of ViaSat-2, which promises a massive step forward in capacity and speed, and whose backers insist will more than justify its cost in terms of new communications services.
Undoubtedly, spectrum sharing techniques are improving all the time ‘ the TV white spaces projects have been an important testing ground ‘ and mobile satellite bands have been underused. However, there are fears that some regulators are belittling the technical and political issues that are involved in the near term, and are disregarding the commercial potential that still lies in satellite broadband.
This is best demonstrated by ViaSat, which answered critics who claim there is no profit model in satellite broadband recently, when it returned to the black and reported a 28% year-on-year increase in first quarter revenues, crediting a rise in broadband subscribers in the residential, broadband and airline sectors. It sees sufficient growth potential to have launched a mighty satellite, ViaSat-1, recently, and to plan a second, upgraded bird, ViaSat-2, in 2016. The former has total capacity of 140Gbps, which supports the firm’s Exede broadband service for remote and rural areas of north America as well as in-flight connectivity. Via-Sat 2 will enable twice the capacity and a sevenfold increase in coverage. Exede offers services of 12Mbps, comparable to Verizon’s recently announced fixed LTE offering, HomeFusion, which promises rates between 5Mbps and 12Mbps. Verizon and, in particular, AT&T have indicated they will use LTE to phase out DSL, in areas where they do not have fiber coverage, but AT&T still relies on satellite in ultra-rural areas, and there are no indications that will change.
Across the Atlantic, the European Union calculates that almost 10m households have no access to broadband ‘ over 95% of European homes are connected, but the figure falls to 78.4% in rural regions. Despite putting LTE and unlicensed wireless systems at the heart of universal coverage projects, the EU still sees a significant role for satellite, as witnessed by its SABER (satellite broadband for European regions) program. This aims to set out a framework for authorities to use satellite to achieve 100% broadband coverage objectives by 2020, and has received €510,000 in EU funding.
In other words, while satellite spectrum will certainly be a valuable source of additional capacity for mobile networks in future, regulators should not be too hasty the wireless industry’s voice is loud and frequently rational, but agencies like the FCC must recognize the continuing role for satellite and, while working on important breakthroughs in spectrum sharing, must also guard against disasters like the LightSquared episode, which only serve to put valuable frequencies into limbo.
And even those holders of less controversial MSS spectrum cannot yet make an LTE business case in that alone. Dish has been open that, if it cannot gain additional frequencies, and/or a strong cellco partner, it will not have enough capacity to achieve its national LTE-Advanced network plan profitably, and would likely sell on its AWS-4 licences. However, its first choice is to have its own LTE network, to add value to its services and challenge the emerging quad plays of the telcos and cablecos.
Hence Dish’s bold bids first for Clearwire ‘ spectrum-rich in the 2.5GHz TDD band, but in need of financing for its own migration to LTE-A ‘ and then for Sprint, whose Network Vision multi-technology infrastructure could save Dish the costs of rolling out its own network from scratch while also bringing PCS, 800MHz and Clearwire spectrum to the party.
And so the US’s spectrum wars go on, as Dish awaits the outcome of its two related bids. It is unlikely to win Clearwire ‘ not only was it always going to be hard to get round Sprint’s 50.1% stake, but the cellco, which had previously made an offer for full control of its 4G venture, has now topped its rival’s bid. Clearwire was a key attraction for Japan’s Softbank when it made its own bid for 70% of Sprint, which was then countered by Dish’s offer for 100% ownership.
This week saw Sprint raising its price for Clearwire as though to speed the closing of the deal, and ensure its own success and the approval of Softbank. It has increased its offer for the remaining shares from $2.97 per share to $3.40, topping Dish’s unsolicited $3.30 per share offer and responding to various large minority shareholders, which had argued that the $2.97 price undervalued Clearwire’s spectrum assets. A meeting that was scheduled to take place today for shareholders to vote on Sprint’s original proposal has been pushed back to May 30.
The new offer is up about 14.5% on the previous one, and is close to Clearwire’s closing price on Monday of $3.26 per share. Clearwire’s board has urged shareholders to approve the Sprint offer as its last, best chance to survive financially, but several minority shareholders have insisted there are better choices for the company and their weight was clearly heavy enough to persuade Sprint to increase its bid. The move may not please Softbank, which sees Clearwire as a key attraction in its own proposal to control Sprint. It needs to sign off on any increased Sprint price, and its CEO Masayoshi Son said recently that he did not see a need for Sprint to sweeten its offer. He argued that, since significant investors Comcast, Intel and Bright House Networks have agreed to sell their collective 13% voting stake in Clearwire to Sprint, it would give Sprint the ability to block other parties.
Further muddying the waters is an unsolicited offer from Verizon Wireless to buy Clearwire’s spectrum leases in major markets for up to $1.5bn.
Sprint also agreed to open up confidential information to Dish. The cellco had been resisting Dish’s requests to see its detailed books, as part of its bid for Sprint, partly because of commitments to Softbank, whose 70% offer has been approved by the board. Sprint has also questioned some of Dish’s claims about the cost savings which would result from its own proposal, and its ability to finance its 100% offer.
However, Softbank ‘ which has vigorously defended its own bid, but refused to raise it ‘ has now granted the waiver which enables Sprint to provide confidential information to Dish, whose chairman Charlie Ergen said in a statement: ‘We look forward to engaging in full due diligence and continued discussions with Sprint. We remain confident that this process will confirm the superiority of our proposal, the reasoning behind our synergy projections and our vision for a competitively superior Dish-Sprint.’
While Sprint has set up a special committee of directors to examine Dish’s proposals, it said it had ‘not changed its recommendation with respect or, and continues to support, the company’s pending transaction with Softbank’. It said it retains the right to cancel the Softbank deal if it receives a superior offer, but insisted it ‘has not determined that the Dish proposal in fact constitutes a superior offer under the existing merger agreement’. Softbank’s offer is scheduled to close in about six weeks’ time.
Meanwhile, Dish is keeping its options open and has been casting far and wide for new spectrum, it seems. According to Bloomberg, he has made a $2bn bid for the L-band spectrum owned by LightSquared. If Dish could secure the spectrum cheaply, out of backruptcy court, it might be able to add the frequencies at a safe distance from GPS to its existing mobile satellite licences, while leaving those close to GPS untouched. However, Dish’s bid is understood to be a ‘stalking horse’ offer, designed to set a minimum price in a bankruptcy sale, and could be topped by other candidates in future.
The LightSquared spectrum would hardly be a simple way to boost Dish’s position. It is ‘certainly not a near term alternative to Clearwire, given that it will take years for use to be approved, and an L-band ecosystem won’t develop until well after that for Dish’s AWS-4 spectrum,’ wrote Tim Farrar of satellite analyst firm TMF Associates in a blog post. ‘On the other hand … it could be a long term back-up option, if Dish plans to sit on its AWS-4 holdings for the next several years and wants to rebind it to downlink spectrum, so it’s worth forcing Harbinger’s hand and ensuring that LightSquared’s assets are ultimately auctioned off.’ Buying Sprint, LightSquared and Clearwire would give Dish a spectrum position to rival that of the big two, though a fiendishly complicated one in terms of multiband support, device ecosystem and carrier aggregation, potentially including five different areas of the spectrum.
So, despite technical challenges, expect the cellular industry’s battle for the satellite spectrum to continue, in the US and elsewhere ‘ but we should watch the progress of ViaSat closely, and not write off the satellite broadband business just yet.
LightSquared’s debts mount:
LightSquared posted a $48.1m loss in April, bringing the company’s total losses since its May 2012 bankruptcy filing to $660.8m. It attributed much of the deficit to interest payments on its debt, which totalled $31m in the month, bringing total payments during the bankruptcy period so far to $363.2m.
It also said it spent $1.1m for its spectrum reuse fee in April, bringing that total to $10.4m during its bankruptcy, plus restructuring charges of $1.2m, bringing that total to $27.5m since May 2012.
The amount it spends on its network could rise in the next three months, the period the FCC recently said it would allow LightSquared to share a band of spectrum with the National Oceanic and Atmospheric Administration (which uses the frequencies for its meteorological work) to test its technology.
The FCC said LightSquared would bear the costs of the test and would have to discontinue use of the spectrum if it interferes with NOAA work but the decision shows the agency’s eagerness to find alternative bands in which the LightSquared wholesale 4G plan ‘ seen as a potentially valuable boost to competition as it could support small and disruptive service providers ‘ could be revived.
LightSquared has until May 31 to file a plan of reorganization, which could include spectrum sale, and has sole control over its reorganization plan until July 15.