Annual spectrum fees and price metrics

Plum produced a study on annual fees for mobile spectrum. The study considered the different types of annual fee, evaluated the metrics used for reporting and comparing fees and compared the level of spectrum fees across countries. This analysis was then used to make recommendations on best practice for setting spectrum fees.

Regulators have different rationales for charging annual fees. In general, annual fees are motivated by a desire to recover administrative costs and to promote spectrum efficiency (particularly in Europe). In some cases they are used to generate a return to the state. Other rationales have also been advanced, for example to address asymmetric profit shocks arising from liberalisation in the UK. Annual fees have also been suggested in the context of licensed shared access.

Annual spectrum fees are growing in prominence, particularly as existing mobile spectrum licences in many countries near their expiry. Other potential triggers for the introduction of annual fees or fee increases include: new spectrum allocations where competition is expected to be limited e.g. Malta, or pricing associated with a tranche of spectrum for a new entrant where an auction is not proposed.

Spectrum fees can be classified according to the way they are administered – auctions and non-auctions. Auctions have the primary aim of achieving an efficient initial allocation of spectrum and where bidders determine the price (though administratively determined reserve prices may in some instances play a role in determining bids, or a decision not to bid). Auctions can in turn be divided into those involving purely upfront payment and those with an upfront payment (determined by the market) and a recurring annual payment (determined by regulators). Non-auctions include recurring annual fees and lump sums.

Benchmarking is a widely used approach in spectrum valuation alongside business and cost modelling. Spectrum prices are generally reported in terms of price/MHz/pop, which may then be adjusted based on exchange rates and inflation to provide a common basis for benchmarking. An alternative metric to price/MHz/pop is price/MHz/GDP. Since GDP reflects population and income per capita, this metric allows for both factors compared to price/MHz/pop, which only factors in population.

There is no clear trend in the level of fees across time, region or frequency bands. Annual fees do not always vary across frequency bands within a country.  There is also large variation in the size of the annual fees compared to the size of the lump sum fees. In Mexico, the annual fees account for 79% of the cost of PCS 1900 MHz frequencies, while in Taiwan (which has similar annual fees in terms of price/MHz/pop) the annual fees only account for 12% of the cost of the 1800 MHz band.

Best practice for managing annual spectrum fees hinges on three key considerations:

  • The balance between annual and upfront fees. Annual fees are not necessarily superior to upfront fees, So long as there is equivalence in present value terms.
  • The comparison of fees. There is no single correct approach to comparing and benchmarking fees. A variety of methods should be applied, judgement exercised and potentially the relationship between fees and possible explanatory variables explored using statistical (econometric analysis).
  • The discount rate. The use of an appropriate discount rate is necessary for any benchmarking exercise for adjusting fees to a common duration to enable comparison, or if annual fees are converted to a net present value (or vice versa).

If fees are fixed the economic cost of setting fees too high versus too low should be considered. Given uncertainty it is appropriate to err on the side of caution, particularly given that benchmark values may include not just opportunity cost but also the strategic value of spectrum to operators.