By Peter Pitsch, Intel's Director of Global Telecommunications Policy
Brazil is in the process of holding another radio frequency auction, the so-called leftovers of the 1,800 MHz, 1,900 MHz, 2,500 MHz and 3,500 GHz bands. There was a public consultation by ANATEL (CP 20) which ended on September 2nd regarding its rules. This is a great opportunity to reflect on recent experiences in other countries.
According to the International Telecommunications Union (ITU), there are 7.085 billion mobile phone subscriptions in the world and wireless broadband has been the fastest growing communication service in the last five years, increasing by 428% since 2010[1]. These impressive statistics suggest that the value placed on spectrum for mobile broadband is high and is likely to increase. But some observers question whether these numbers actually justify allocating more spectrum to mobile broadband.
An alternative basis for assessing the likely value of additional spectrum for mobile broadband is provided by recent US auction results. Auction bidding has three important advantages: they are objective, rigorous and forward-looking. Winning bidders who propose to spend billions of dollars to buy spectrum are not making political statements. Rather, they are making a statement about the net present value of the spectrum purchased. That is, in a competitive auction, the amount of money companies are willing to bid will reflect how much they expect to be able to earn over time. In addition, rational bidders consider possible alternatives to purchasing more licensed spectrum at a particular auction, including further splitting the cells, improving existing technology, offloading Wi-Fi, or even waiting for a subsequent auction. to name a few possibilities. And because they are spending billions of dollars, bidders have a strong incentive to rigorously analyze these alternatives and market prospects.
For these reasons, recent auction results for so-called Advanced Mobile Services (AWS-3) in the US are instructive. On that occasion, 65 MHz of spectrum was auctioned. The bands were divided into six blocks. Two blocks were unpaired spectrum at 1695-1710 MHz, divided into a 5 MHz band (A1) and a 10 MHz band (B1). The remaining four blocks were of paired spectrum, where two corresponding segments of spectrum are auctioned together, uplink and downlink. Uplinks were sold in the 1755-1780 MHz range, and downlinks in the 2155-2180 MHz range. This paired spectrum was separated into three 2×5 MHz blocks (5 MHz uplink, 5 MHz downlink), called G, H, and I, plus a 2×10 MHz block (10 MHz uplink, 10 MHz downlink), called J
The US Congressional Budget Office (CBO) estimated that these bands would sell for about 20 billion dollars[2]. In the end, the tracks were purchased for a staggering $45 billion – and a study by the Phoenix Center[3] showed that they could easily have raised an additional $20 billion if the unpaired bands were sold as a pair.
The unpaired spectrum sold for considerably less than the paired spectrum. The so-called A1 went for a mere 171 million dollars, or US$ 0.11 PMP, that is, the price per MHz per person. B1 sold for $2.26 billion, or US$ 0.72 PMP. In contrast, the paired spectrum sold for significantly more. Block G sold for $7.41 billion, H for $8.45 billion, I for $8.40 billion, and J for a staggering $18.2 billion. The respective PMPs were US $ 2.40 (G), $ 2.70 (H), $ 2.69 (I) and US $ 2.91 (J).
However, the results of the so-called AWS-3 auction in the US may be even more significant than they seem at first glance. It is reasonable to assume that bidders on AWS-3 have considered Federal Communications Commission (FCC) plans to auction off 600 MHz of spectrum currently held by US broadcasters in an incentive auction in early 2016. two sides. The FCC will buy spectrum from broadcasters willing to go out of band (in the reverse auction), provided that the operators' bids (in the auction) are sufficient to pay the buy-outs and other costs. Thus, the more operators are willing to pay, the more broadcasters are likely to give up their spectrum.
While there are pros and cons on the relative merits of each of the spectrum bands up for auction – 1.7 GHz and 2.1 GHz on AWS-3 and the 600 MHz bands, both blocks are highly valued by operators and are surrogates. reasonable from the point of view of its application. Thus, rational bidders in AWS-3 have safely considered the possibility of participating in the incentivized auction to satisfy their spectrum needs rather than entering the AWS-3 itself. As the two spectrum bands can be considered functionally comparable from the operators' point of view, the proposals expressed in AWS-3 give an indication of what to expect from the proposals that will be made in the incentivized auction.
Therefore, as bids for the AWS-3 auction have exceeded expectations, it has become more likely that bids in the Incentive Auction will also exceed initial estimates – and that the Incentive Auction will indeed clear significant amounts of the 600 MHz spectrum. . However, the increase in the probability of a greater supply of spectrum in the incentivized auction should have put a strain on prices in AWS-3. Or, in other words, the AWS-3 auction prices imply high values for both spectrum auctions, even if the incentivized auction is very successful.
Thus, the results of the AWS-3 auctions suggest that additional spectrum for wireless applications in the US is in high demand, even in the presence of additional spectrum, to be auctioned in the next year. Implications for other countries may vary. For example, placing legal burdens on licenses such as network deployment requirements for spectrum access can reduce the value of winning bids. However, it seems reasonable to conclude that in developed countries in the short term and in developing countries in the future, broadband wireless spectrum will remain highly valuable.
EN 
PT 
