What Do Current 5G Spectrum Auctions Tell Us About its Business Case?

On April 5, 2018, United Kingdom’s telco regulatory authority Ofcom completed auction of 5G spectrum in the 3.4 GHz band. The four winners included Vodafone, O2, Three UK, and EE Limited, and the auction raised US$1.64 billion in revenue. The results were the following for a total of 150 MHz:

• O2 won 40 MHz, paying about US$447 million (US$11.18/Hz).
• Vodafone won 50 MHz, for which it paid US$532 million (US$10.65/Hz).
• EE paid US$427 million for 40 MHz, and (US$10.68/Hz).
• Three UK won 20 MHz, paying US$231 million (US$11.57/Hz).

The auction experienced a delayed start primarily due to regulations on the amount of spectrum an operator could bid for in the auction and the amount it could hold in total after the bidding. While legal appeals by Three UK and O2 to reduce the cap were rejected, Three UK had a year earlier completed the acquisition of U.K. Broadband and its spectrum in 3.4 GHz to 3.8 GHz, 3.9 GHz, 28 GHz, and 40 GHz. Following U.K.’s spectrum auction, South Korea as well as the United States announced 5G spectrum auction details on the spectrum band, bandwidth, and minimum opening prices. South Korea is expected to hold its 5G auctions in June 2018 and the United States in November 2018.

Licensed spectrum is the most important mobile service provider asset, and it has in some cases affected mergers and acquisitions. So even prior to spectrum auctions, Mergers and Acquisition (M&A) activities in 2017 and even earlier (when the spectrum holding played a key part) will allow some operators to be early movers in 5G deployment. As a most recent example, this is also clear in the latest T-Mobile/Sprint merger announcement in which Sprint’s spectrum was a key component for the decision to merge.T-Mobile has stated its plans to use its 600 MHz spectrum holdings to deploy 5G. And now with more than 160 MHz of Sprint’s nationwide 2.5 GHz spectrum, which was included by the 3rd Generation Partnership Project (3GPP) in its 5G New Radio (NR) specifications in December 2017, T-Mobile will be able to address both coverage and capacity in the key U.S. markets for its 5G deployment.

While acquisitions have played an important role in bolstering an operator’s spectrum holdings, 5G spectrum auctions will ultimately make relevant spectrum bands available to more operators. Currently, about 40 countries globally are in different stages of planning for 5G spectrum auction, but only a handful of countries have announced plans to hold 5G spectrum auctions in 2018 and in 2019. Most 5G auctions are expected to take place after World Radio Conference (WRC) 2019, to be held between October and November 2019, when 5G bands will to be finalized. The earliest mobile 5G services are expected in the second half of 2019 after 5G handsets become available.

The M&A activities provide some indication of the value of 5G spectrum; however, the recently completed auction is the clearest indication available yet. Currently 5G auctions are less expensive than 4G auctions, and this trend has not changed with the United Kingdom’s spectrum auction. While this may differ across countries as individual regulators may have different objectives in setting minimum prices and auction rules in the context of 4G competitive landscape, it does provide a directional guidance for the sub-6 GHz band.

Table 1 provides an overview of the sub-6 GHz 5G-auctions completed so far and the upcoming 5G auctions in 2018. It does not include 5G auctions in the similar bands completed by Slovakia as that was a regional auction with national-level licenses already in place. It also does not include the 600 MHz auction in the United States even though T-Mobile has stated that it plans to use that band for 5G mobile services. To compare with 4G spectrum auctions, the most relevant auction was researched in terms of timing as that would be key in determining the value of the spectrum. As evident from the data, on a US$/MHz-Pop basis, currently, 5G spectrum auction is less expensive than 4G. South Korea’s higher minimum opening price in the sub-6 GHz band does not necessarily imply significantly higher final auction revenue as has been the case with some of its auctions in the past. As recent as 2016, the final auction price increased for only one out of the four spectrum bands being auctioned for US$260 million. The lower opening prices on the Millimeter Wave (mmWave) band both in the United States and South Korea possibly indicate that these are expected to be very competitive auctions due to the wide bandwidths available at those frequencies and that the regulators are more confident of exceeding a certain target amount in those bands. It is possible that mmWave band spectrums will earn higher revenues for the regulators than 4G auctions.

The current lower US$/MHz-Pop in the sub-6 GHz range could also, however, indicate the lack of a strong business case for 5G services in these bands. Ofcom released only 150 MHz for auction, which left each operator with less than 100 MHz. Even in the case of South Korea’s auction in the 3.5 GHz band, where each operator is expected to acquire at least 100 MHz, the same US$/MHz-Pop values between the United Kingdom and South Korea for different spectrum bandwidth indicate that currently the sub-6 GHz bands are not seen as offering a strong business case to operators. One of the key distinguishing features of 5G is the use of wider spectrum bandwidths both in mmWave and sub-6 GHz bands. It is possible that Ofcom will hold another auction in the sub-6 GHz bands after 5G services have been launched.

While 5G mmWave bands hold wide swathes of contiguous bandwidths, the sub-6 GHz 5G bands, especially the C-band (3.7 GHz to 4.2 GHz) is crucial for coverage for 5G enhanced broadband services. It does provide a vast improvement in capacity over LTE bands, but the substantial increase is expected to come from mmWave bands. A wider coverage can also be realized with sub-1 GHz bands. Furthermore, technologies such as uplink and downlink decoupling can even help operators leverage existing LTE bands. Still for many operators, the C-band may the best choice for deploying 5G networks for ease of deployment and cost-effectiveness as they may be able to use existing LTE macrocell sites, and a lower bidding price may further provide momentum to this trend.

As is the case with South Korea, it is likely that many countries will auction sub-6 GHz (including sub-1 GHz) and mmWave bands simultaneously or, like the United States, multiple mmWave bands in the same auction. What that auction scenario and the current data indicate is that operators will need a clearer five- to eight-year 5G deployment plan utilizing different spectrum bands on which a sounder spectrum strategy can be built. These factors ultimately impact the proliferation of 5G networks and the kind of business cases that can be built using different spectrum bands.