The proxy statement filed by Clearwire on Friday morning made for interesting reading, not least in guessing the identity of some of the companies that Clearwire has talked to over the last couple of years. Some are obvious (A=T-Mobile, B=AT&T, C=MetroPCS, D=China Mobile, F=LightSquared, H=Verizon) and some are more speculative (E=Google? G=one of the hedge funds invested in Clearwire, I=Samsung or Qualcomm?), but what stands out is the lack of bids for Clearwire’s spectrum at an attractive price.
In particular, Clearwire didn’t find the bid from T-Mobile (in fall 2010) to be “compelling” and was unable to reach agreement with MetroPCS (in fall 2011) on the spectrum to be included in any deal (i.e. owned vs leased) or the price. Clearwire notes that even DISH’s bid at a price of $0.19/MHzPOP “related to the acquisition of higher quality spectrum assets of Clearwire and would leave Clearwire with less valuable spectrum assets”, implying that it is mostly for owned and/or contiguous spectrum (as I suspected), and implying that the price for the spectrum that Clearwire would be left with could be rather lower.
However, the fundamental reason for Clearwire to acquiesce to the Sprint bid appears to be that Clearwire has been unable to find a second wholesale customer for its network. That target customer appears to have been AT&T, given the financials for the MCC (Multi-Customer Case) set out in the proxy, which assumed that the second customer would generate approximately 1.5 times the revenue produced by Sprint, or in other words would have had around 90 million customers who could use the Clearwire network.
AT&T certainly took quite a lot of interest in Clearwire, conducting “extensive due diligence” in fall 2010 (though ultimately declining to submit an offer, presumably opting instead to pursue the T-Mobile merger) and then resuming discussions (at AT&T’s initiative) in February 2012:
In February 2012, Party B approached the Company about restarting discussions about a possible spectrum sale and commercial agreement. The conversations between the parties focused on technical issues and the spectrum that the Company could potentially make available for sale. Party B made it clear during the discussions that a transaction with the Company was one of the several options it was pursuing in order to satisfy its spectrum needs. Party B terminated further discussions with the Company in May 2012 after it had determined to pursue one of the other options to satisfy its spectrum needs.
AT&T’s deep dive into “several options” in the first half of 2012 (after rejection of the T-Mobile bid) appears to have been very extensive, and focused largely on options to meet longer term spectrum needs (3+ years out). Of course we know that AT&T ultimately decided to buy NextWave, after approaching NextWave initially at the beginning of April. However, I’m told that AT&T also approached LightSquared in early 2012 with a view to buying the 1670-75MHz spectrum block (when LightSquared indicated it was considering a sale of these spectrum leases), and its inconceivable that AT&T didn’t also discuss with DISH the possibility of buying the 2GHz MSS spectrum (when it actively tried to interfere in DISH’s FCC proceeding).
Was AT&T scared off by the FCC’s denial of DISH’s ATC waiver request in early March, or simply by Charlie Ergen’s high asking price. More than likely it was the latter, given AT&T’s initial low ball $350M offer for NextWave’s spectrum in April 2012 (after the waiver denial, but presumably when discussions with DISH were still ongoing), followed by a decision to offer rather more in early June, after AT&T had reached a decision on its preferred spectrum option. Around the same time DISH also began exploring options with Clearwire and MetroPCS, leading up to its parallel bids to take over MetroPCS and buy spectrum from Clearwire in August 2012.
I’ve often thought that AT&T’s interest in WCS could be independent of a possible purchase of DISH’s spectrum, with WCS providing a “high band” option for dense urban networks (as a direct alternative to Clearwire’s BRS/EBS spectrum), while DISH’s spectrum provides a “mid-band” alternative to PCS or AWS. Indeed, after AT&T gave up a large slice of its AWS spectrum to T-Mobile as part of the break fee, it was plausible to think AT&T would have a potential shortfall in its mid-band spectrum assets, which would make DISH’s spectrum particularly attractive. However, it appears that AT&T may instead have looked at DISH’s spectrum more as “what do we need for 3+ years out”, considering it alongside WCS and BRS/EBS, which would almost certainly lead to a mismatch of valuation expectations with Charlie Ergen.
Instead AT&T now appears to be focused on a combination of its 700MHz LTE network (bulking up with the B block purchase from Verizon) plus smaller amounts of cellular, PCS and AWS spectrum (including the acquisition from ATNI/Alltel and the AWS spectrum included in the NextWave purchase) to meet its near term needs. The 12MHz of Qualcomm 700MHz D/E block spectrum would then be used for supplementary downlink to relatively narrow 5x5MHz cellular, AWS or PCS LTE deployments in urban areas from 2014 (though DISH’s 700MHz E block holdings could spoil any prospects of deployment outside the 5 major metropolitan areas where AT&T owns the entire band). This would explain why AT&T has retained 10MHz of AWS spectrum in Los Angeles, splitting the AWS A block spectrum with Verizon as part of the 700MHz B block deal.
Although its not been widely recognized, AT&T has already started to deploy LTE within its AWS spectrum in a few markets, and is now emphasizing the capacity enhancements available from small cell technology. Indeed, given the backlash against just the two modestly sized spectrum purchases from Verizon and ATNI, its hardly conceivable that AT&T could be planning to buy DISH’s spectrum in the near term as well. However, if AT&T was going to do another spectrum deal in the near term (which may now be unlikely), I’d bet that using Leap or US Cellular’s spectrum (perhaps even selling their customers to Sprint or T-Mobile?) would be more in line with AT&T’s current spectrum strategy than a deal with DISH.
So where does that leave Charlie Ergen? Perhaps he really does need to secure a deal with Clearwire not just to make it impossible for Sprint to get control of Clearwire’s spectrum, but also so DISH has a way forward to a near term deployment? Alternatively, moving towards a deal with LightSquared and a reorganization of the AWS-4 band to create additional downlink spectrum (as I suggested in December) could continue to create problems for Sprint (given its desire to purchase the H block) without committing DISH to a near term buildout. Either way, it seems that in the near term, Ergen might be more likely to be a buyer than a seller of spectrum.