The market landscape and auction structure make the Canadian 3500 MHz midband spectrum auction an interesting one to watch. The auction will kick off on June 15 to release an average of 111 MHz of which around 50 MHz will be set aside for regional wireless operators. This is relatively small in terms of what 5G requires to deliver the promised Gbps throughput. Supply constraints could suggest high prices. But the background is complex and the outcome will depend on the potential demand.
Supply Throttle
The 3500 MHz midband spectrum auction is a first instalment in a two-phased release of midband spectrum. A chunk of 330 MHz in the C-band (3650 – 3980 MHz) is scheduled for 1Q2023 [see here]. However, as 80 MHz of the C-band spectrum will be assigned on shared basis, it leaves 250 MHz for contention among bidders.
A few countries released midband spectrum in instalments: the UK, Hong Kong and a few countries in Eastern Europe are examples. In these auctions, the number of bidders was limited to primarily the incumbent operators. As such, most of these auctions did not yield much of a premium over the reserve price, except in the UK. There, the first instalment of 150 MHz fetched 81.5% higher price than the second instalment of 120 MHz. In effect, the operators have discovered their financial boundary and play book for 5G, and acted accordingly.
High Pricing Floor
The Canadian 3500 MHz midband spectrum auction defines different reserve prices depending on population in Tier 4 service areas. Bidding starts at C23.2 cents/MHz-PoP in the three metro areas of Toronto, Montreal and Vancouver. This is the highest starting price of any spectrum auction I have ever seen. In fact, the average final price for 25 recent midband auctions is C$0.2391/MHz-PoP and the median is C$0.1086/MHz-PoP. [The FCC C-band auction skews these results: see here.]
Perhaps the intent of such price is to limit the number of rounds leading to a quick outcome. However, in gaming the auction process, such a price may lead to excess bidding, especially when the price increment is also high. We’ll have to wait and see how much of an impact the initial price has on the final outcome.
The Game for Incumbents
Bidders entering the 3500 MHz midband spectrum auction will have to carefully consider how to play their hand. For the three incumbent operators, the 50 MHz set aside spectrum will limit their allocation to an average 90 MHz in the top 3 metro areas. With both Bell and Rogers each in possession of [almost] nationwide 30 MHz, including the top 3 metro areas, the stakes are high to develop a dominant position over Telus which has important holdings of its own, despite being more specific in select markets.
The dynamics among the incumbent operators will be interesting to watch for several reasons. First, Bell and Telus have a RAN sharing agreement to complement their strengths on geographic basis. Second, Rogers is in process of acquiring Shaw, whose Freedombile business unit is the 4th largest operator with 2 million subscribers. Third, Bell withdrew from the last major auction (600 MHz) - I would think they plan very differently this time.
The Game for Regionals
This auction may be a defining one for regional operators. Spectrum assets - and especially midband spectrum - are more important than ever to compete effectively in the era of mobile broadband. As the average spectrum holdings of each of the regional operators is 50 MHz, winning the entire set-aside allocation in their respective regions will double their current holdings. However, this would leave each regional operator with about half of the holdings of the national incumbents! This translates into half the peak speed that an incumbent vendor could provide. With the licensing process for the 3800 MHz auction is still to be determined, the regional players have to carefully consider their long-term strategy on how to scale. The absence of Shaw/Freedom Mobile makes it easier than ever for the regionals to acquire spectrum in this auction.
The Urban-Rural Divide
It is a given that the metro areas will be highly contested among the major bidders. What is less clear, is how much emphasis will be placed on the rural and remote markets. The Tier 4 service area licenses for this auction are relatively small – there are 172 such areas across Canada. The intent was to encourage small service providers to participate in the auction process and to eventually serve these rural and remote areas. But will that happen?
Consider that the top 3 metro areas account for 40% of the total population and 50% of the MHz-PoP. The top 10 markets account for 61% of the population and 68% of the MHz-PoP. The major bidders will allocate their firepower to the large markets, and view areas with small population as non-strategic. However, the small markets will account for relatively little cost, hence, the large bidders could sweep these markets with comparatively little cost. How will the bidding for rural and remote areas unfold, and will small players find a niche?
Concluding Thoughts
These are some of the dynamics at play in this auction. But there are many more to consider in the finer details of this auction. For instance, the Rogers-Shaw deal leads to many wrinkles. There is also the question of how the bidders will approach the allocation stage in light of the upcoming 3800 MHz auction. How ever this auction ends, it will set an interesting background for the upcoming C-band auction in early 2023!