There was little surprise that the three national telcos were also the three big winners in the USA’s recent auction of C-band spectrum in 3.7-3.98 GHz. The results of Auction 107 showed that, for all the talk of diversification of the mobile business in 5G, this is not yet affecting the outcomes of major auctions, unless a regulator intervenes to ensure this.
What was a surprise was the vast sum Verizon paid, investing over $45.5bn in gross bids in its effort to make up for the lack of midband spectrum that has severely constrained its 5G efforts in the first phase. The need to acquire midband frequencies, which support high 5G capacity without the engineering and propagation challenges of millimeter wave bands, has become even more urgent since T-Mobile USA acquired Sprint, and so gained its rich dowry of 2.5 GHz 5G-suitable spectrum.
Verizon spent at least $10bn more than analysts had expected on the eve of the auction, and more than Morgan Stanley Research, last year, predicted as the total takings of the whole auction. The company forecast a final tally of $26bn, with $35.2bn being the top estimate, before revising their numbers as the level of interest in the C-band became clear, and the US saw the onset of the latest outbreak of that well-known mobile industry disease, auction fever.
The FCC completed the second, assignment phase of the auction a week ago, matching the winning bidders up with specific frequencies. On Wednesday, it announced the winning bidders. A total of 21 organizations paid $81.2bn between them to acquire all 5,684 licences that were on the table.
Verizon’s total was $45,454,843,197. AT&T, which has the same challenges in the midband as its rival (but more low band coverage spectrum in 600 MHz) spent $23.4bn. There was a significant gap between these two high sums and the third largest spender, TMO, which paid $9.3bn for its allocations.
After that, bids from smaller MNOs or new entrants were considerably smaller. Only US Cellular, which pledged over $1.28bn, and NewLevel II, a private equity firm specializing in broadband, on $1.27bn, topped the $1bn mark.
The auction broke US records even without significant competition from the large cablecos to drive up the prices. Comcast and Charter, which cooperate on their mobile network services and build-out, were initially expected to be heavy spenders, but in the end, decided to sit it out, while Dish Network was also a surprisingly inactive player, spending less than $3m under the bidding entity name of Little Bear Wireless.
In terms of the numbers of licences acquired, Verizon gained 3,511 and AT&T 1,621. US Cellular gained 254, well ahead of TMO’s 142 even though it spent less, reflecting the different locations that each operator prioritized (US Cellular tends to focus more on rural and suburban areas where bidding was likely to have been less intense).
In fifth place in licence numbers was Canopy Spectrum, another private equity spectrum investor, which gained 84. The FCC has introduced rules to prevent private equity buyers from selling on their spectrum immediately for a quick profit, but they can still get an attractive interim return via leasing.
Meredith Attwell Baker, CEO of the US operator trade body, CTIA, said in a statement: “These record-breaking results highlight the demand and critical need for more licensed midband spectrum and demonstrate the importance of developing a robust spectrum auction pipeline.”
The FCC has also released its proposed rules for another midband auction, in the 3.45-3.55 GHz band (currently occupied, though underused, by naval radar). Michael Calabrese, director of the Wireless Future Project at New America’s Open Technology Institute, called for an overall spectrum aggregation limit to apply to prevent the big telcos becoming unassailable in their positions. He said: “Unfortunately, the auction’s $15bn reserve price, most of that needed to pay the cost of clearing Navy radar off the band, has led to an auction format that like C-band is designed to exclude all but the very largest mobile carriers. Since we also learned today that Verizon and AT&T together have acquired over 80% of the C-band’s 280 MHz, it will be crucial for competition that the FCC adopt a 3 GHz spectrum aggregation limit as part of the 3.45 GHz auction rules.”
The FCC is proposing to structure the 3.45-3.55 GHz auction in a similar way to the C-band process in terms of partial economic areas, band sizes and technical rules. Some analysts speculate that Dish spent so little on the C-band (just one licence in Cheyenne, Wyoming, where it has a test site) because it is saving itself for Auction 108, which should be more affordable now that the big two have gained such significant midband holdings.
Rick Engelman, an engineering consultant with telecom law firm Wiley Rein, told Light Reading that A Block licences across the biggest cities gained the highest bids. The 100 MHz in the A Block will be available for commercial 5G operations by the end of 2021, while the B and C Blocks will not be available until 2023 as they have to be vacated by satellite incumbents like SES and Intelsat.
Verizon will have to spend an additional $53.4bn to pay those satellite players to clear out their operations from the C-band. The telco has announced a $25bn loan from JPMorgan Chase Bank to help with the costs.
One forecaster who was proved right was T-Mobile USA’s CEO Mike Sievert, who told an investor event last year: “I think AT&T and Verizon will absolutely kill each other over C-band. I think they’re going to spend tens of billions of dollars they don’t have to stress out their balance sheets and put at risk their share buyback and dividend plans in order to not be left out of the party on 5G. Because they’re stuck. They’re way behind [in spectrum ownership] and they can’t stand it. So they’re going to have to do things that are probably going to be uneconomical.”
Despite the big sums spent by the two largest telcos, TMO still has the largest overall spectrum holdings below 6 GHz, largely thanks to Sprint’s 130 MHz or so of 2.5 GHz frequencies per market.