As you would expect, mobile service provider executives complained at MWC25 that spectrum costs are too high and regulation too strict. The implication is that governments should release spectrum at lower costs. As the saying goes, “good luck with that.”
Spectrum licenses are a revenue source for governments who see themselves as custodians of a public resource, and are likely to be loathe to do so. Also, there are many ways for service providers to increase capacity, with other approaches. Use of smaller cells, better radios, retail pricing policies and marketing as well as other spectrum efficiency moves can increase the effective capacity of mobile networks, as important as additional spectrum allocations are.
With the caveat that the U.S. mobile service provider market typically is more robust than that of European providers, and that spectrum costs are not identical across markets, there still tend to be global benchmarks for spectrum.
And there is a difference between spectrum license cost as a matter of capital expense, and its amortization as an operating expense. Service providers capitalize spectrum license costs, meaning the actual cost burden happens as an amortization (depreciation) line item.
So it is far from clear how much cost spectrum licenses actually represent, as a percentage of operating cost, for example. For the larger European operators (Vodafone, DT, Orange, BT, for example), spectrum might seem among the bigger cost drivers, but is not among the biggest cost contributors.
And some observers might point out that the actual cost of spectrum is less the capex amount than the amortization and depreciation impact, despite the impact on cash flow.
Some industry estimates suggest that spectrum-related amortization might account for 20 percent to 40 percent of the depreciation and amortization cost for a U.S. mobile operator, for example. ,Assuming a mid-range 30 percent, that might suggest that spectrum costs (amortized) represent about five percent to 6.5 percent of service provider depreciation costs:
Verizon: ~$5.3 billion (30% of $17.6B) out of $105B opex = ~5%.
AT&T: ~$5.6 billion (30% of $18.8B) out of $97B opex = ~6%.
T-Mobile: ~$4 billion (30% of $13.2B) out of $62B opex = ~6.5%.
Compared to other cost contributors, that is relatively slight. The point is that if lower costs are important, and they are, other places seem to be more likely places to find such efficiencies. Spectrum costs are, after all, a direct reflection of operator demand.
And one value aspect is clear enough:unless an operator can acquire new spectrum to support each new mobile platform, it will eventually go out of business. So spectrum costs are an existential matter for mobile service providers, and that is unlikely to depress idemand, or willingness to pay.
