The Finnish authorities (Ministry of Transport and Communications and Traficom, the Finnish Transport and Communications Agency) have drafted plans to auction a key 5G capacity band, 26 GHz, in June 2020. In comparison to any current mobile frequency allocations the new band is exceptionally wide; it spans from 24.25 to 27.5 GHz and thus includes a total of 3.25 GHz of frequencies. The lower part, ie. 24.25 – 25.1 GHz is to be reserved for local, unlicenced uses and is left out from the auction. This leaves a total of 2.4 GHz to be awarded in the auction.
In general, the auction plan follows a similar three-way split as the earlier mobile frequency auctions in Finland. The band is divided into three blocks of 800 MHz as follows:
25.1 – 25.9 GHz (block A);
25.9 – 26.7 GHz (block B); and
26.7 – 27.5 GHz (block C)
One might conclude that the implicit objective is to support the current market setting with three major mobile network operators, each possessing equal frequency assets. In a sense, this objective is understandable since the Finnish mobile market is highly developed, availability of mobile broadband is good and prices are acceptable. However, the the plan attracted some criticism during the consultation round.
One source of criticism was the intended balance between national & licenced and local & unlicenced spectrum. 5G is believed to enhance possibilities for new kind of businesses and local networks that may be run at least partly independent of the established mobile operators. To facilitate this kind of development some respondents, for example Digita, the major broadcast network operator, would like to have a bigger share of the band allocated for unlicenced uses. Similarly, Finnet, an association of the local Finnish operators, would like to advance spectrum sharing. These views are supported by the fact that due to its poor propagation characteristics the 26 GHz band is not suitable for building national coverage in countries like Finland. The established operators see the case differently and would like the whole of 3.25 GHz to be allocated for licenced uses.
Another source of criticism was the planned auction design and poor reasoning in the draft award documentation. First, the band is divided into three big blocks; ie. the authorities wish to fix the allocation in advance. A more market-oriented approach would be to have small slices and let the markets decide on the allocation in the auction. The documentation provides about no reasoning on why the 3-way split would be an optimal choice. This criticism was presented especially by Aalto University. In addition, the auction is planned to follow the same principles as the previous auctions despite their poor track record. With three blocks and three bidders in each of the previous auctions (800 MHz, 700 MHz and 3.5 GHz bands) there was no auction-like pricing mechanism and the frequencies were sold at only slightly above their reserve prices. With the planned approach, this is the most likely outcome in this case as well. In consequence, it seems rather unnecessary to run an auction; a beauty contest might be a simpler and, surprisingly, even more transparent way to reach the most likely outcome. Another option would be to change the auction design.
The licence period is 1 July 2020 – 31 December 2033. The reserve price has been set at EUR 7 million per 800 MHz and the operation must be started within two years but there are no specific roll-out obligations. The auction is planned to start on 8 June, 2020.