The FCC’s 2.5 GHz auction (FCC Auction 108) ended Monday, after 73 rounds, reaching net proceeds of only $427.8 million (M). The FCC found winning bidders for 7,872 of the 8,017 licenses offered. The FCC holds the remaining 145 licenses. Proceeds were much less than anticipated before the auction. Pre-auction estimates had run as high as $3B, or in the range of $0.10 to $0.20 per MHz-POP. In reality, the end result was just $427.8M in aggregate proceeds, and less than two cents per MHz-POP on average.
“After some extended bidding in Guam today, Auction 108 finally came to an end,” wrote Sasha Javid, BitPath chief operating officer. “While the end of this auction should not be a surprise for those following activity on Friday, it certainly ended faster than I expected just a week ago.”
With no assignment phase, Javid predicted the FCC will issue a closing public notice in about a week, with details on where each bidder won licenses. T-Mobile was expected to be the dominant bidder as it fills in gaps in the 2.5 GHz coverage it’s using to offer 5G. AT&T, Verizon and Dish Network qualified to bid but weren’t expected to acquire many licenses (see Craig Moffett’s comments below).
New Street analysts significantly downgraded projections for the auction as it unfolded, from $3.4 billion, to less than $452 million in its latest projection. New Street’s Phillip Burnett told investors Sunday Guam Telephone Authority was likely the company making a push for the license there. The authority owns citizens broadband radio service and high-band licenses “but lacks a powerful mid-band license” since “no C-Band or 3.45GHz licenses were offered for Guam,” he said. “We still assume T-Mobile won essentially all the licenses,” Burnett said in a Monday note. The auction translated to just 2 cents/MHz POP, 8 cents excluding the areas where T-Mobile is already operating, he said: “This will make it the cheapest of the 5G upper mid-band auctions at the FCC to date, both in terms of unit and aggregate prices. However, given how odd these licenses were, we wouldn’t expect to see the auction used as a marker for mid-band values going forward.”
Craig Moffett of MoffettNathanson wrote:
While we won’t know for sure who “won” the licenses in question for another week or so, it is universally assumed that T-Mobile was far and away the auction’s principal buyer. They are the only U.S. company that uses 2.5 GHz spectrum (2.5 GHz is the backbone spectrum band of their 5G network), and the licenses at auction were best seen as the “holes in the Swiss cheese” of T-Mobile’s otherwise national 2.5 GHz footprint. There was a great deal of spectrum here for sale, but it wasn’t geographically contiguous, and thus it would be difficult for anyone other than T-Mobile to use it. Nor should one expect spectrum speculators to have played a large role; after all, if there is but one true exit – i.e., to sell to T-Mobile – then bidding more than T-Mobile was willing to pay would seem an ill-advised strategy. Usually, we refrain from using the term “winner” when discussing auction results.
Winning, after all, depends on price paid. In this case, however, there can be little argument that T-Mobile is the auction’s big winner (assuming, again, that it was indeed T-Mobile that bought almost everything here). They will have significantly expanded their already-large spectrum advantage versus Verizon and AT&T and they will have done so at a much lower price than had been expected. Remember that not only does T-Mobile enjoy a spectrum quantity advantage versus Verizon and AT&T– they already had more mid-band spectrum than either VZ or T, and now they will have significantly augmented their already prodigious holdings – but they also have a spectrum quality advantage, inasmuch as 2.5 GHz spectrum propagates better than the 3.7 GHz C-Band spectrum used for 5G by Verizon and AT&T, and therefore promises better coverage and fewer dead spots with less required capital spending for density/coverage.
T-Mobile just a few weeks ago invested about $3.5B in low frequency spectrum, allocating about the same amount of capital many had expected them to spend on Auction 108. But their low (600 MHz) frequency spectrum purchase – done at what we assume is $2.53 per MHz-POP in a two-part acquisition from private equity owners – is for spectrum they were already leasing, so it represents a direct substitution of capital for opex without changing the amount of spectrum employed in their network. Margins should be higher, as what was previously leased is now owned. And, happily, they got the deal done just before the Inflation Reduction Act eliminated the cash tax shield from spectrum purchases as it relates to the 15% minimum corporate tax rate on future spectrum purchases.
If there is one additional takeaway here, it is the reminder that spectrum is NOT a commodity, where prices inherently reflect some immutable “intrinsic value.” Instead, they are highly volatile, reflecting much more the dynamics of supply and demand for each individual spectrum band at any given moment, factoring in not just how much different carriers might want the spectrum, but also what their balance sheets will bear.
Our long-term tracking of spectrum transactions, sorted into low-band, mid-band, upper mid-band, and millimeter wave cohorts, now updated to include both Auction 108 and T-Mobile’s private market transactions for 600 MHz spectrum, tells the tale: