Your browser is not supported. Please update it.

1 February 2019

The USA’s much-vaunted mmWave auction is attracting lukewarm interest

The quantity of spectrum available in the millimeter wave (mmWave) bands will, of course, mean these never generate the same bidding wars, and resulting high prices, that are sometimes seen in lower bands. However, after the FCC placed 28 GHz at the heart of its Spectrum Frontiers initiative – anointing itself as the world leader in opening new bands for 5G – the results of the still-ongoing (and aptly named) Auction 101 are a damp squib.

The FCC is offering two licences, L1 and L2, each for 425 MHz, in each county-sized licence location. The same operator is allowed to buy both, which would give a total of 850 MHz per market – sufficient to support some interesting 5G applications.

As of Thursday, the auction had generated just $700m in bids – which compares to $44bn raised in the AWS-3 sale in 2015, and $20bn in 2017’s 600 MHz incentive auction. But 28 GHz had three factors working against high prices. In the areas where is is available, there is plenty to go around, so no need for escalating rounds of counter-bidding.

Second, the two big telcos already have 28 GHz and 39 GHz spectrum, which they have acquired by buying companies that owned licences in these bands dating back to the broadband wireless bubble around 2000 – and there are other licences out there, which could probably be acquired relatively cheaply from their current owners. So AT&T and Verizon have limited requirement to bolster their arsenals – Verizon has said it has plenty of spectrum for its 5G needs for the foreseeable future. Indeed, Verizon is estimated to own about 75% of the USA’s total 28 GHz spectrum thanks to its acquisition of XO Communications, so Auction 101 has been dubbed the “crumbs from Verizon’s table”. Meanwhile Sprint has been somewhat cool on mmWave because it has so much of its 2.5 GHz spectrum still unused and available for 5G. It is not bidding, partly because it would not want to bid against its potential new parent, T-Mobile.

Third, the 28 GHz auction only made spectrum available in certain markets, and many major metro areas, such as New York City, Los Angeles, Chicago and Philadelphia – were missing, though dense urban networks are the main application most operators see for high band 5G (once they look beyond the initial roll-outs, like Verizon’s, for fixed wireless access). The top markets, in terms of bid value, in Auction 101 have been Dane, Wisconsin ($12.5m); Honolulu, Hawaii ($10.2m); and Linn County, Iowa ($10m) – unlikely to be the lead markets in any operator’s 5G business plan.

Analysts expect the auction to have been dominated by Verizon filling gaps in its footprint, and AT&T and T-Mobile possibly bidding mainly to try to prevent that, and so weaken their rival’s enviable mmWave position. However, the bidders’ target markets and bid amounts are not revealed until the final results are clear, though the FCC does release the list of qualified participants – AT&T, Verizon, T-Mobile, US Cellular, Frontier, Windstream and a long list of small or rural operators and other players.

T-Mobile also has some 28 GHz holdings, acquired when it bought regional operator MetroPCS in 2013. That company had 28 GHz licences covering a total of 100m POPs, in a rather fragmented pattern, but including some of the nation’s top markets. So, according to estimates by AllNet Insights & Analytics, TMO owns 20% of the 28 GHz spectrum in New York City, while Verizon owns the rest.

Verizon’s initial 5G Home fixed 5G service uses 400 MHz, but the carrier plans to double this to 800 MHz before year end – hence why it will be looking to bolster its holdings, however substantial these already are, in the auction, to maximize the speeds it can support, and prevent TMO, in particular, from achieving a competitive position in the 28 GHz band.

The next auction, Auction 102, will start as soon as this one finishes, which could be as soon as this week. It will focus on the 24 GHz band, which is expected to generate greater interest because there is spectrum available across virtually the whole country -AllNet predicts takings anywhere between $2.4bn and $5.6bn, assuming prices of around $0.01 to $0.02 per MHz/POP.

Licences will be available in most major cities, with a few exceptions such as Las Vegas, Phoenix and Albuquerque (where a company called Skyriver owns the spectrum – and must be calculating which MNO might want to pay handsomely to buy its assets).  Auction 102 will offer seven licences of 100 MHz each, in each market, and one operator would be allowed to acquire all 700 MHz.

The licence sizes will be larger than 28 GHz’s counties – they will be available in Partial Economic Areas (PEAs), which are bigger and therefore more attractive to large players. Small and rural telcos like small areas, which they can align better to their areas of operation and which are cheaper. But national MNOs prefer the scale and simplicity of large areas. For instance, the New York City metro equates to just one PEA, but 28 counties. Among the qualified bidders in 24 GHz are AT&T, Verizon, T-Mobile and Windstream, cableco Cox and fixed wireless start-up Starry, among others. TMO is expected to be a big bidder, to shore up its mmWave reserves, especially as a fallback option if it fails to acquire Sprint. TMO’s initial 5G plans center on a coverage-oriented roll-out in its 600 MHz spectrum, but if it does not get its hands on Sprint’s plentiful 2.5 GHz reserves, it will need to get high capacity spectrum elsewhere.

The next mmWave auction in the USA will be in 39 GHz. This is the band in which AT&T is making its first 5G deployments, courtesy of its acquisition of FiberTower last year. The operator will have the same choice as Verizon in 28 GHz – sit the auction out because it already dominates this band, or move aggressively to ensure it controls the lion’s share of this spectrum, and to raise barriers to rivals.