The 700 MHz auction is over at last. In the next 6 weeks, 8 Canadian carriers will be scraping together five and a quarter billion dollars to pay for their spectrum. Some were quick to compare this to the $4.25 B raised in the 2008 AWS auction, without realizing that the license being issued this time around are 20 years, not 10 years.
The final bids varied widely between carriers and comparative analysis is challenging without considering which blocks were acquired, whether the spectrum was adjacent to a block that can be aggregated to provide better performance, the demographics of the geography and the compatibility with existing networks for handsets and partnerships.
Not all spectrum is alike.
For example, SaskTel issued a press release expressing its concerns with the outcome. SaskTel paid just $7.5M for the C1 block in its home province, covering about 10.4 million megahertz pops (population times the bandwidth). But SaskTel said “device options are not available for the C1 band”, so its “rural customers will not see a benefit for some time”. As it has consistently claimed for a couple years (since the auction framework was announced), SaskTel said “This incredibly complex auction format is clearly biased against regional carriers, in addition to having a number of other systemic flaws.” It called for the Federal Government to continue to focus efforts on ensuring no spectrum remains unused in rural Canada.
Rogers acquired dual blocks in most of the most populous regions of the country, gaining the A and B blocks everywhere but Manitoba and Saskatchewan, Northern Ontario, Northern Quebec and the Northern Territories. As Minister James Moore said in the press briefing, Rogers paid the most ($3.3B), but they got the highest quality spectrum that was being offered. With aggregation possible across the two blocks, the purchase should allow Rogers to get higher performance from its network, and many of its existing customers already have compatible handsets.
I won’t begin to speculate on Videotron’s spectrum strategy. I will just observe that in 2008, the company spent $555M to acquire AWS spectrum, mainly in Quebec, and on a 10 year license. Today, the company spent 60% less, just $233M for a 20 year license on Quebec, Southern and Eastern Ontario, Alberta and British Columbia. It picked up spectrum in the most populous and most valuable markets in the country, with spectrum that has far better performance characteristics than its current inventory. Videotron spent less for its 20 year 700 MHz spectrum than Mobilicity spent for less coverage with its 10 year AWS purchase ($243M). Allocated on a population basis, Videotron’s “out-of-territory” 700 MHz spectrum cost about $150M of the $233M that they spent.
Does this set an upper bound for the value of Mobilicity’s spectrum?
how do the $/mhz/pop prices compare with US 700? AWS?
an interesting exercise would be to revisit the Telus(?) paper that explored the overpayment in the AWS auction as a result of the set-aside and compare with the outcome of this auction.
This certainly is a much less transparent process. The SMRA format was open and anyone could watch the round by round results to understand what was going on. Not only do you practically need an economics PHD to understand the combinatorial clock format, but it seems to be a complete black box – at least to the public.