CBRS Auction 105: MNO Results and The Changing Landscape

Spectrum decisions can have a profound impact on coverage, capacity, user experience, and ultimately network cost; all of which are determining factors of market share. With the recent T-Mobile + Sprint merger and whirlwind of mmWave, CBRS, C-Band, and L-Band spectrum opportunities; we are witnessing the largest re-shuffling of market power in recent telecom history.

Prior to the T-Mobile + Sprint merger both Verizon and AT&T had led T-Mobile with 115MHz and 150MHz of nationwide spectrum. However, heading into the CBRS auction T- Mobile surpassed AT&T and Verizon to become the new spectrum king. With the closing of the merger, T-Mobile had amassed ~300MHz of spectrum thanks in large part to Sprint’s increasingly valuable 2.5GHz spectrum. The CBRS auction offered the first of several opportunities for Verizon and AT&T to partially close the spectrum gap. However, surprisingly AT&T was absent from the auction likely leaving their dry powder for a more aggressive position in the upcoming C-Band auction.

CBRS (Citizens Broadband Radio Service) opens up 150 MHz of spectrum in the 3.5 GHz band for commercial use. CBRS employs the use of a Spectrum Access System or SAS to dynamically share and priority usage across three tiers of licenses.

• Tier 1 is reserved for incumbent users including federal government, military, or fixed satellite. Incumbent users maintain the highest protection and always receive priority over Tier 2 and Tier 3 users.

• Tier 2 or Priority Access Licenses (PAL) were awarded during the auction to the highest bidders. Each Priority Access License covers 10 MHz of spectrum in a county with a maximum of 4 licenses (40MHz) held by any given licensee. PAL users receive priority and interference protection from GAA users.

• The GAA Tier offers offer at least 80 MHz of shared spectrum and as much as 150 MHz of spectrum for GAA when not in use by a PAL licensee. GAA users receive spectrum allocations from the SAS, but do not receive guaranteed protection from other GAA users.

Spectrum valuations are often expressed as a $/MHz/Pop with the intrinsic value reflective of the propagation characteristics of the auctioned band. Lower frequencies command a higher price due to a lower cost of deployment with wider area coverage. As networks continue to evolve to provide higher quality coverage, increased capacity, and new peak speeds they require both an increasing amount of spectrum and higher cell density. As cell density increases the value of low-band spectrum is diminished making mid-band increasingly popular. Compared to prior auctions two additional factors contributed to a lower value for CBRS. First, the non-exclusive nature of a PAL license and second the power restrictions.

At the close of auction 105 the FCC had raised ~$4.6B at an average price of $.22/MHz/Pop. The smaller license area offering lower entry prices, the availability of dynamic spectrum sharing, and the allure of new or improved business cases with private LTE and fixed wireless attracted 257 bidders with 228 licensees walking away with licenses.

As expected, there was evidence of a general relationship between the number of individual bidders and the price paid per MHz per Pop. However, what may be surprising are the markets that attracted the largest number of bidders. Its no surprise that markets like Los Angeles, CA were popular and garnered the attention of the largest number of bidders, however, Mills, IA tied Los Angeles for the highest number of bidders. Beyond the usual MNO and cable MVNO suspects in large markets like Los Angeles, Orange County, and San Diego there was the presence of large utility companies including Southern California Edison and San Diego Gas & Electric as well as smaller players like Ten-Four and Bee Wireless. In Iowa we saw the presence of several regional or fixed wireless providers including Nextlink, Western Iowa Networks, and Michigan Wireless, Atlantic Tele-Network, and RSA 1.

With the shift in market dominance and the mid-band spectrum advantage of T-Mobile there is an increased pressure on competitors to quickly close the spectrum gap. CBRS is the first spectrum auction to occur post merger and as result we wanted to map out impact of spectrum strategy for nationwide MNOs.

Dish

Heading into the auction Dish held ~100MHz of spectrum nationwide and a MVNO subscriber base of approximately 8M from their acquisition of Boost Mobile. During the auction, Dish acquired 5,492 licenses across 3,128 counties at a price of $913M. Their CBRS spectrum licenses covered 309M people or 98% of population with an average depth of 1.8 licenses per county basis at a price of $.15/MHz/Pop.

Dish ensured that it got at least 1 carrier in every county. The only counties where Dish lost in bidding were in the Island of Puerto Rico.

Dish has no existing wireless or wireline infrastructure though they have subscribers and spectrum in place. With the low power requirements for CBRS and its existing spectrum holdings, along with the need to build a greenfield network, it will be interesting to see how Dish’s network deployment strategy evolves.

Verizon

Verizon entered the auction with 115MHz of nationwide spectrum spending $1.9B to acquire 557 licenses across 157 counties covering 143M people or 46% of the US population with an average of 3.5 licenses per county. Verizon paid an average $.39/MHz/Pop.

Verizon’s approach to spectrum acquisition was very focused, bidding on a limited number of counties containing major urban areas. Verizon’s focused approach ensured they they received at least 3 carriers in every county where they bid with the exception of Arlington, Virginia where they were limited to 2 carriers.

Heading into the auction Verizon had been active in deploying CBRS and stands to rapidly benefit from the combined benefit of GAA and PAL licenses.

T-Mobile

T-Mobile paid only $5.6M for 8 licenses in 6 counties with 1 license across 4 counties in Essex, Hudson, Mercer New Jersey and Bronx NY and 2 additional licenses in Osceola, Florida and Charleston, SC. T-Mobile paid an, average $of .14/Mhz/Pop covering 3.8M people or 1% of population. Their most expensive license was in the Bronx at price of $3.7M or $.27/MHz/Pop.

Additionally, T-Mobile is in the process of trying to acquire Shenandoah Telecommunication’s (Shentel) wireless assets. During the CBRS auction Shentel was an active bidder walking away with licenses across 74 counties covering a total of 3.6M people at a price of $16M and $.13/MHz/Pop. Shentel operates a wireless network in rural Virginia, West Virginia, Kentucky, Ohio and Pennsylvania serving 1.1M subscribers.

During the CBRS auction T-Mobile bid opportunistically, initially bidding on up to 53 markets, but ultimately walking away with a limited number of licenses in only 6 counties. T-Mobile did not appear to be a serious bidder during these auctions which was likely driven by their significant mid-band holdings.

All 4 competitors took very unique approaches to CBRS. Dish purchased a limited the number of licenses, but went wide in terms of coverage and population density resulting in a lower cost per MHz per Pop. Verizon purchased a large number of licenses in select markets ultimately ensuring the maximum amount of spectrum where it mattered the most. T-Mobile already well situated with mid-band picked up licenses opportunistically, while AT&T was not an active participant in the auction.

Spectrum decisions can have a profound impact on coverage, capacity, user experience and a strong correlation with market share. At the end of the CBRS auction T-Mobile remains in pole position with 300MHz of spectrum while Verizon partially closed the gap with an average of 3.5 carriers pushing them to 150MHz in key markets. While Verizon continues to lag in valuable mid-band spectrum, CBRS provides quick relief and significant capacity and peak speed enhancements to their existing LTE network.

The next big opportunity for buying mid-band spectrum licenses is the C-band auction, which starts in December and offers up to 280MHz of valuable spectrum.

The opinions expressed in this case study are based on publicly available information and do not claim to represent any companies views or strategy.