Is Charlie Ergen giving up on his wireless plans? An interview with Bloomberg published this morning seemed to signal a change of stance, particularly the explicit suggestion that Ergen would be open to simply selling his spectrum to AT&T, when his (coded) message for the last year has been that AT&T would need to buy the whole of DISH (at a huge premium).
Not only has the Softbank deal enabled Sprint to escape from the box DISH had been trying to put it in, but by buying McCaw’s stake in Clearwire, and indicating to other strategic investors on the board that further equity purchases are up for negotiation, Sprint looks to have headed off Ergen’s planned deal to purchase assets from Clearwire.
Despite this, perhaps Ergen could still do a deal with other partners like Carlos Slim or DirecTV, but it is harder to see the rationale for investing in building a fifth competitive operator than it was a month ago, when Sprint (and perhaps even T-Mobile) were perceived to be in a much weaker competitive (and financial) position. In addition, we may be seeing AT&T attempt to warn off Carlos Slim from a deal with Ergen, in the form of mutterings about foreign ownership of other wireless operators.
So it looks like Ergen may now turn his attention towards a potential merger deal with DirecTV. In the short term that means putting pressure on DirecTV to come to the table, most obviously by competing with them in Brazil. Indeed, Echostar told the FCC at the end of September that it planned to move the Quetzsat satellite to the 61.5W slot, freeing up AMC-15 to be relocated to 45W in order to provide service into Brazil “as soon as January 2013“.
Wireless and broadband could also be a key part of this deal, via an alignment of both satellite TV operators with AT&T, similar to the Verizon-cable TV partnership, and that would be cemented by the sale of DISH’s spectrum to AT&T. Nevertheless, I would expect the sale to come ahead of an agreement with DirecTV, simply because being left out of a DISH-AT&T alliance would put even more pressure on DirecTV. As I noted last December, this alignment would cement AT&T and Verizon’s market leadership. However, given that the FCC backed off mandating wholesale access as a condition of approving the Verizon/SpectrumCo deal, the only plausible way to promote competition to these two players may then be to allow Sprint and T-Mobile to merge.
Of course putting a price on DISH’s spectrum is hard, but I suspect that the most appropriate comparison for the AWS-4 spectrum is the WCS band which AT&T purchased recently, because neither band is usable immediately (unlike AWS). AT&T paid NextWave more than expected for WCS, but this was offset by the fact that AT&T already owned a substantial part of the band, for which it had paid very little. I would therefore expect an AWS-4 sale to be at a fairly similar price to that AT&T paid NextWave for the usable parts of the WCS spectrum: somewhere in the range of ~$0.30-$0.40/MHzPOP (depending on how much value is attributed to the C/D blocks and AWS holdings), i.e. of order $3.75B-$5.0B plus perhaps another $1B for DISH’s 700MHz E block spectrum.
Today the House Energy & Commerce Committee has released the documents from last month’s hearing, which makes for some interesting reading.
For one, its intriguing that the FCC Chief Economist (on p33 of the PDF) highlights “just how lucrative ATC service could be”, noting that “an entrepreneur that purchased a mobile satellite property wth 34 MHz of ATC-eligible spectrum and successfully rolled out ATC service in partnership with a CMRS incumbent could obtain a rent of $11 billion, equal to the difference between the value of the spectrum ($12 billion) and the cost of developing the required satellite system ($1 billion).”
That $12B valuation of course is the same as LightSquared’s economic consultants used, back in June 2011, to estimate that the January 2011 waiver was worth $10B to LightSquared. The documents also indicate (p10 of the PDF) that the FCC Chairman was warned in September 2009 about the “equity considerations of providing a ‘windfall’ to companies who did not acquire the licenses at auction” and even at that time “changing the ATC handset requirements” was under discussion.
However, far more interesting are the details of the discussions over the extension to the very short comment period on the waiver request in November 2010 (pp13-15 of the PDF). It is clear that the FCC had essentially already agreed the waiver with LightSquared and had “discussed [the timetable] previously”. The FCC was working to this “tight” timetable (later derailed by NTIA concerns about GPS interference) so that order could be issued on December 20, a date that was described as “critical”.
Of course (although it is not stated in the emails) the reason that this date was so critical was that (as I suspected at the time), the Commission intended to (and did) vote on net neutrality rules at the December 21, 2010 Open Meeting, which excluded wireless networks from net neutrality obligations. As a result, if the FCC had been able to announce the LightSquared waiver at that meeting, it would have been possible to say that at least one major 4G network provider had signed up to net neutrality principles of its own accord, effectively endorsing LightSquared as a competitor to the major cellular operators.
Despite considerable efforts by Charlie Ergen, it looks like the Softbank deal may now have enabled Sprint to escape from the box of constrained capital, limited spectrum and a second rate network that Sprint could have been confined to, if it had failed to gain access to either usable H block spectrum or Clearwire’s network on economically advantageous terms. Many thought that Sprint would move to purchase Clearwire immediately after the Softbank investment, but today sources are denying that is the intention, stating that Sprint has no intention of taking part in mergers or acquisitions until the Softbank deal is finalized in mid-2013. This timeline also implies that Sprint will not move to disrupt the T-Mobile/MetroPCS merger, which is expected to close in 2013Q2.
So the obvious question is why did Sprint need to issue a $3B convertible bond to Softbank right now? I think that can only be intended to warn off others from doing a deal with Clearwire in the interim, by offering John Stanton the carrot of improved economics and/or further investment from Sprint. Of course there are not many options for Clearwire to sell spectrum, now that T-Mobile and MetroPCS, the two operators most frequently rumored to have designs on Clearwire’s spectrum, are getting together.
As a result, I think Sprint’s actions appear to confirm that Clearwire was about to pull the trigger on a deal with Ergen, as I suggested last month, involving an asset sale and/or WiMAX customer transfer, in exchange for a combination of cash and debt. Notably, receipts from a sale of network assets (as opposed to a spectrum sale) would not have to be used to repurchase Clearwire’s first lien debt, suggesting that this could be a preferred way for Clearwire to raise funds. In addition, I’m told Ergen now holds in excess of $900M of Clearwire’s debt (not all first lien), and some of that could potentially have been traded for Clearwire spectrum.
Reports on the Sprint/Softbank deal have also suggested that both Carlos Slim and SK Telecom have considered investments in Sprint, and it is worth noting that SK Telecom invested $60M in LightSquared back in 2010, while Slim is rumored to be buying LightSquared debt. In fact I’m told that further significant purchases of LightSquared debt have taken place in recent weeks. If one or both of those two players therefore continue to maintain their interest in US telecom assets (which has obviously included MSS-ATC spectrum similar to that held by DISH), then Ergen may be the last, best potential partner available.
So has Sprint now prevented Ergen from achieving a deal with Clearwire? I’m told that (at least with the current ownership situation) Sprint would have no ability to veto such a transaction, so presumably Stanton will now be trying to extract vastly improved economics from the existing capacity agreement with Sprint in order to forego a DISH deal. What concessions will Sprint be prepared to make, and if it does give ground, where does that leave DISH? After all, it doesn’t seem that AT&T is prepared to pay Ergen’s asking price (perhaps as high as $80-$90 per share?) to purchase the whole of DISH anytime soon. Ergen must certainly be fuming at how FCC delays have prevented him from moving forward, while potential partners seem to be rapidly exiting the dance floor. At least he appears to have made a profit on his investment in Clearwire, but that may be little consolation if it now proves more difficult to find a way to monetize DISH’s other, much larger, spectrum investments.
As I pointed out last week, despite growing skepticism over “What happened to the spectrum crunch?“, the FCC Chairman has been busy making speeches about how “U.S. mobile data traffic grew almost 300% last year, and mobile traffic is projected to grow an additional 16-fold by 2016″ and asserting that
“There were many skeptics [in 2009] about whether we faced a spectrum crunch. Today virtually every expert confirms it.”
One would assume that the FCC Chairman regards the CTIA as experts, because after all they have been amongst the foremost cheerleaders for the spectrum crunch over the last three years. Indeed, upon releasing their latest wireless industry statistics, CTIA’s press release on Thursday proclaimed that “Wireless Network Data Traffic Increased 111 percent and Highlights Industry Need for More Spectrum”.
However, the press release buried the far more significant message in the CTIA’s actual data (which counts wireless network traffic from 97% of all US wireless subscribers – it is unclear if any WiFi traffic is included, but it appears not), namely that growth in data traffic per device came almost to a full stop in the first half of 2012 (compared to the previous six months), presumably due to a combination of:
1) changes in user behavior (offloading to WiFi),
2) dilution from less active users buying smartphones, and
3) data caps beginning to impact some high end users.
According to the CTIA survey, total wireless data traffic in the US in the first six months of 2012 was 635B Mbytes, up only 21% on the 526B Mbytes recorded in the previous six month period. This compares to the 54% growth seen in wireless data traffic between H1 and H2 of 2011, and the 51% growth in traffic between 2010H2 and 2011H1.
Even more remarkably, with the number of smartphones and wireless connected tablets growing by 16% between 2011H2 and 2012H1, data traffic per device was only up about 3% (based on the average number of devices in each 6 month period), compared to 29% between H1 and H2 of 2011 and 23% between 2010H2 and 2011H1.
As shown in the chart below, if we extrapolate this same growth in traffic per device to the second half of 2012, then total data traffic during the period will be about 750B Mbytes, and total traffic for 2012 will have increased only 60% compared to 2011. That is only half of Cisco’s Feb 2012 estimate of 118% growth in North American mobile data traffic between 2011 and 2012, and would clearly force a complete re-evaluation of future traffic projections. However, just from the data in the first half of 2012, we can already see that traffic growth is clearly not “exponential” and has now passed its peak. More importantly, if growth in traffic per device has been brought under control, then a spectrum crunch is no longer possible in the foreseeable future, because smartphone penetration is already approaching saturation.
As others have pointed out, “there is no more scarcity of wireless spectrum than there is a shortage of, say, the color purple”. Its therefore perhaps ironic to note that back when the last Emperor Julius (Caesar) was in charge, “virtually every expert confirmed” that the color purple was expensive, difficult to find and limited by regulatory fiat to a tiny elite group.
The history of the MSS industry, like most other parts of the telecom and technology sectors, has revolved around a theme of “faster, better, cheaper” as technological advances have dramatically improved satellite throughput and enabled significant reductions in the price per minute and per bit of voice and data communications. However, it seems that at least in L-band, we may have reached the limit of economically justifiable technological advances, and the future for the traditional MSS market could now be one of slower development cycles, simpler satellite systems and more expensive services.
Some of those technological dead ends are obvious, like the huge satellite antennas built for LightSquared and TerreStar, which did nothing whatsoever to make their satellite services more viable. However, more importantly, Inmarsat sets the tone for the entire MSS sector and could now make it much clearer that we have reached the end of the historical development pattern for this industry.
We’ve already seen Inmarsat pushing up prices this year, and with its new focus on the Ka-band Global Xpress system, Inmarsat has also indicated that it expects to delay capital expenditures on a next generation L-band (I6) satellite system and that the system itself will be cheaper and simpler than the I4 satellites. After feeling a backlash from distributors and customers, Inmarsat has been at pains to suggest that this year’s price rises represented a one-off adjustment, rather than the start of regular yearly price rises. Nevertheless, Inmarsat has seen little if any impact in terms of maritime customer losses because Inmarsat holds such a dominant position in the market and Inmarsat’s 2012 wholesale revenues are likely to be boosted by at least 2%-3% as a result of the price rises.
Looking forward, the outlook in most parts of the MSS market is fairly depressed, with M2M services providing the main source of growth, and overall wholesale L-band revenues are only likely to grow by perhaps 4% p.a. in the next few years. Although we’ll find our more specifics on Tuesday about Inmarsat’s expectations for Global Xpress, it also remains hard to see how GX will meet the original target of $500M in incremental wholesale revenues within 5 years, suggesting that Inmarsat’s move to Ka-band will not move the needle on overall MSS market growth very far. Reasons for this include the apparent lack of military Ka-band frequencies (which were supposed to be secured by Boeing and would be needed to allow GX to operate as a seamless supplement to WGS), the fact that XpressLink customers are being given the option but not the obligation to upgrade to GX when it is launched and, most importantly, the threat posed by Intelsat’s new Epic satellites, which have secured some key anchor customers (Panasonic, MTN and Harris Caprock) and caused many distributors and end customers to reconsider the “inevitability” of a move to Ka-band.
As a result, it seems there is now a fairly clear case for Inmarsat (as the price leader in the MSS sector) to push through regular annual price rises on the 50% or so of its wholesale L-band revenue base that is least likely to move to alternative solutions and has the lowest price elasticity. This would mean price rises for low and mid range maritime customers plus many land customers, while leaving aeronautical and high end maritime customers (who are more at risk from VSAT competition) largely untouched. Today there is an essentially flat outlook for L-band revenues, as modest growth in M2M and handheld is being offset by the migration of high end maritime and aeronautical customers to XpressLink (and other VSAT solutions), together with reductions in defense spending/event revenues. However, a price rise of 5%-10% p.a. would potentially allow Inmarsat to grow its L-band business at 2%-4% p.a. for the next several years.
Other MSS operators would also benefit through (modest) gains in market share and more pricing freedom, and the overall MSS sector could perhaps then return to something closer to the 7% p.a. revenue growth rate seen before the downturn of the last couple of years. Even distributors, who have seen their margins pressured over the last decade, would appreciate some ability to increase their overall revenues (given that demand elasticity is quite low), if margins can be sustained or even increased.
However, in order to execute such a change in strategy, Inmarsat would also need to repair its relationship with independent distributors, which has taken a significant knock from Inmarsat’s acquisitions of Segovia and Ship Equip and its decision not to pass on some wholesale price increases through its own direct distribution channels. There have also been several instances where Inmarsat has made its own bids for key contracts which massively undercut the bids from the incumbent independent distributors providing the same services. Indeed it often appears that Inmarsat has the explicit intention of driving its leading independent distributor, Vizada (now owned by Astrium Services) into the arms of other satellite operators such as Intelsat and Iridium. It seems to me that only with the cooperation of independent distributors can Inmarsat present a united front to end customers, and explain that price rises are necessary for a healthy market, as opposed to having their assertions undermined not only by (the expected) sniping from competitors, but by distributor dissatisfaction (and potential defections due to increased margin pressures) as well.
Is Inmarsat willing to change this dynamic and restrain its direct distribution arm, or will it resort to a “bunker mentality” and either miss the chance to boost revenues or push further price rises onto its independent distributors, amidst a rising tide of opposition? Time will tell, but its perhaps worth remembering that the full title of the book was “The End of History and the Last Man”, the last man being someone who “is tired of life, takes no risks, and seeks only comfort and security“.
As I noted in an earlier post, it was quite surprising to read LightSquared’s assertions in court last Monday October 1 that:
“We’ve made substantial progress on our regulatory issues…Short of a few sections of the country, dead zones we will attempt to resolve through other means, this would give us 4G LTE coverage throughout the country. It’s not the proverbial home run everyone said we’d hold out for, but it is a significant terrestrial network.”
As far as I know, no-one has pointed out that these “few sections of the country” where LightSquared would have problems using the 1675-80MHz band it has asked to share, actually include 69 fixed radiosonde launch stations in the lower 48 states (plus additional mobile launch sites when severe weather conditions such as hurricanes are anticipated), from which the weather balloons can drift 180 miles or more, depending on which way the wind is blowing. A quick look at a map of these locations shows a near uniform distribution of sites around the country and that a 180 mile exclusion zone around each station would make it impossible to provide service almost everywhere in the US.
Of course in the past LightSquared demonstrated its skill at predicting and even controlling which way the political winds were blowing. So maybe they now intend to see if that capability extends to real life weather as well?
FCC Chairman Genachowski took a trip to Wharton today, to tell a bunch of students about “the incredible world of mobile communications”. However, he also gave away an enormous amount of information about the FCC’s spectrum agenda, which so far has gone almost completely unreported. As first sight one might be distracted by such nonsense as “U.S. mobile data traffic grew almost 300% last year, and mobile traffic is projected to grow an additional 16-fold by 2016″ and the boast that he alone knew “that something was up”, “did the math” (wrongly) and “sounded the alarms…about the looming spectrum crunch”. Incredibly Chairman Genachowski even makes the ludicrous claim that:
“There were many skeptics [in 2009] about whether we faced a spectrum crunch. Today virtually every expert confirms it.”
Of course this comes just at the time when journalists are starting to ask “What happened to the spectrum crunch?”
Once you’ve stopped laughing at all of this, the meat of the speech is in fact very useful, as the Chairman indicates just how he hopes the “audacious” target in the National Broadband Plan of freeing up 300MHz of spectrum by 2015 will be “exceeded” by a combination of auctions, removal of regulatory barriers, clearing the TV bands and spectrum sharing. First of all, 75MHz of AWS spectrum will be auctioned, including, in 2013, the 10MHz of H-block spectrum desired by Sprint. This confirms that DISH has lost the battle to avoid a 5MHz shift in its uplinks, but in compensation DISH will at least be authorized to use the full 40MHz of spectrum (2005-2025MHz up and 2180-2200MHz down) for a terrestrial network “later this year”.
Secondly, an additional 50MHz of AWS-3 spectrum (desired by T-Mobile) will be auctioned, based on spectrum sharing with the DoD in the 1755-1780MHz uplink band. Finally, AT&T will get its rebanding of the WCS spectrum approved. The Chairman even indicates that the FCC is “working with stakeholders to enable use of the portions of the mobile satellite spectrum in the L- and Big LEO bands [i.e. LightSquared and Globalstar] for terrestrial service” although notably this spectrum is not included in the 2015 total, indicating that these efforts may not be concluded quickly.
The most obscure reference is in the unstated 15MHz balance of AWS spectrum planned for auction before 2015. Given the short timeframe, this can only be the 1695-1710MHz spectrum being reclaimed from NOAA. Presumably this block will be made available as uplink spectrum (because it is adjacent to AWS-1 uplinks at 1710-1755MHz) and as such it will be attractive for AT&T to pair with the WCS spectrum (which will probably all be converted to downlinks). However, this leaves LightSquared in a bind over the spectrum “swap” it proposed last Friday, because LightSquared does not want more uplink spectrum (let alone having to buy it in an auction), and after giving up the 1695-1710MHz block, NOAA will need to use the 1675-80MHz band even more intensively for weather balloons.
Looking at the bigger picture, the situation may be made more difficult not just for LightSquared, but for DISH and Clearwire as well, because the FCC’s actions appear designed to give all of the major wireless operators the spectrum they are hoping for in the near term. Specifically, the FCC intends to free up the H block for Sprint, the AWS-3 block for T-Mobile and 1695-1710+WCS for AT&T, while Verizon has already had its SpectrumCo purchase approved. Especially in the wake of yesterday’s T-Mobile/MetroPCS merger, this makes me wonder just how many attractive alternatives Charlie Ergen still has to a deal with Clearwire for buildout of his 2GHz spectrum?