Spectrum has traditionally been very highly priced as a precious non-renewable resource. An experiment by the FCC to incentivise existing spectrum holders to sell their assets for cellular use has shown that operators are shifting their capital spending priorities elsewhere. With the auction imminently about to close, we report on the auction process and speculate on the longer term implications.
What is an Incentivised Spectrum Auction?
Generally speaking, when spectrum has been auctioned or allocated to mobile operators in the past, it is either previously unused or been freed up by regulator mandate. Analog TV was switched off and replaced by much more efficient Digital TV, while government, military and other users have been required to free up some of their allocations. Some of the proceeds from spectrum auctions have been used to compensate or address technical problems for those affected at the discretion of the regulator.
In a radical new approach, the US FCC created a scheme whereby broadcasters who currently operate in spectrum suitable for cellular could offer to sell some of it at a price. Market forces should mean that it is allocated for the most valuable purpose.
So a back-to-back auction process was created whereby broadcasters could place some of their spectrum assets up for auction and mobile operators could bid for it. There are two separate but interdependent auctions – a reverse auction determining the sale price which broadcasts would accept and a forward auction determining the price cellular operators would pay. A “repacking” process arbitrates between the two, reorganising and reassigning channels to the remaining broadcast stations in order to clear contiguous blocks of spectrum for cellular use. The process was organised into four separate phases or stages.
Quantity and Timescale
A total of 126 MHz of spectrum in the 600MHz band was offered up by broadcasters, who initially priced this at $86 billion. This was repackaged into 80MHz of "clean" spectrum after allowing for guard bands and other adjustments. A limit of 40MHz was placed on those operators who already have a lot of spectrum assets; effectively this means Verizon and AT&T are likely to get at most 20MHz each.
Congress authorised the plan back in 2012 and the auction formally began in March 2016. There are four stages or phases of bidding, starting in May 2016 and completing in January 2017. The latest package is 118MHz repackaged into 70MHz of clean spectrum.
The fourth and final stage of bidding indicates a final price of around $20 Billion, substantially less than the $45 Billion paid for AWS-3 spectrum in 2015 or the $86 Billion initially requested by broadcasters.
The appetite for cellular spectrum is diminishing
Lower frequency (sub 1GHz) spectrum can be very useful to increase the capacity of longer range macrocell towers serving large areas. The benefit is to avoid the need for additional towers and instead just pay for the spectrum and additional equipment onsite. These lower frequencies also penetrate inbuilding to a better extent, also preserving the value of macrocells, but they can’t always be re-used as closely together as with higher frequencies thus limiting total system throughput.
There is also a benefit for dominant players to hold as much of this spectrum as they can, to avoid new entrants (such as the Cable TV companies) rolling out their own networks. For this reason, limits were set on how much the incumbent operators could acquire.
In areas of higher traffic capacity, simply expanding the spectrum at existing sites isn’t enough. Sector splitting and other techniques help, but there is a point beyond which service quality deteriorates and system throughput cannot be increased so easily.
Instead, capital investment programs look to re-use existing spectrum by in-filling with small cells. Savings from not buying additional spectrum will offset the high capacity gains. Even a relatively small number of additional urban small cells, say two or three per macrocell site, can double total capacity.
Growing options for in-building spectrum
There are already many buildings constructed with very high RF isolation from the outside which offer a relatively easy method of re-using existing spectrum. These are often the locations with high traffic/high quality demand and which are more difficult to serve efficiently from outdoor sites.
There is no immediate need to acquire additional spectrum for this purpose and there is already a pathway to boost speeds and capacity further using LAA, which can add extra channels on unlicensed 5GHz bands on-demand.
Furthermore, the promise of CBRS 3.5GHz spectrum, essentially free or at minimal cost, and MulteFire (operating completely in free unlicensed spectrum) open up business models that don’t have to factor in the huge upfront cost of spectrum licences.
The economics of the cellular industry are changing. These auction results demonstrate there is a price ceiling for even the most precious spectrum and that alternative solutions are gaining mindshare. The next visible turning point will be when we see a shift of capital investment programs over the next 3-5 years.