The 4G spectrum auction results were announced back in February. Five telcos won spectrum: Telefonica O2, Vodafone, EE, 3 and BT. MLL Telecom and HKT (UK) Company lost out. Before looking at the merits of each deal it is worth understanding the pros and cons of each spectrum band.

Two bands were up for grabs: 800MHz and 2,600MHz or 2.6GHz. 60MHz was available in the 800MHz band and 185MHz in the 2.6GHz band. There was therefore more capacity available and in bigger blocks at the higher frequency than in the lower. The data throughput that can be achieved in a mobile network is proportional to the amount of spectrum you can throw at it.

The higher frequencies are potentially more valuable from a network capacity perspective than the lower. Where there is a 35MHz block available you can also decide whether to use the whole block to offer a faster service to fewer people or to divide it into smaller packages and serve more subscribers with lower speeds. 45Mbps versus 15Mbs say (my guess).

The downside for the higher speed spectrum is that it has poorer in building penetration and a lower  reach and is therefore not as useful for providing a fixed line broadband replacement service as the lower 800MHz band.

One lot in the 800MHz band was designated by Ofcom as being saddled with a coverage obligation with a requirement to reach 98% of the population with a 2Mbps service by 2017.

Before rural dwellers get excited it is worth noting that the coverage obligation states that “a minimum download speed of 2Mbps should be available with 90% confidence in 98% of houses (residential properties) covered by the mobile broadband service when the network is lightly loaded. Lightly loaded is defined by Ofcom as a “single user demanding service within the serving cell, and the surrounding cells of the network are loaded to a light level (by which we mean the common channels only are transmitting at 22% of the maximum cell power)”.

In my mind that means that 2Mbps is the absolute maximum anyone will get under the coverage obligation. If this was introduced to support the government’s 2Mbps for all pledge then look out for weasel words galore when that number is not achieved by “the end of this parliament”.

Next let’s look at who bought what.

Winning bidder	Spectrum won	Base price
Everything Everywhere Ltd	2 x 5 MHz of 800 MHz and 2 x 35 MHz of 2.6 GHz	£588,876,000
Hutchison 3G UK Ltd	2 x 5 MHz of 800 MHz	£225,000,000
Niche Spectrum Ventures Ltd (a subsidiary of BT Group plc)	2 x 15 MHz of 2.6 GHz and 1 x 20 MHz of 2.6 GHz (unpaired)	£186,476,000
Telefónica UK Ltd	2 x 10 MHz of 800 MHz (coverage obligation lot)	£550,000,000
Vodafone Ltd	2 x 10 MHz of 800 MHz, 2 x 20 MHz of 2.6 GHz and 1 x 25 MHz of 2.6 GHz (unpaired)	£790,761,000
Total		£2,341,113,000

 

Vodafone

At £790 million Vodafone spent the most dosh in the auction and came away with the best spread of spectrum with 20MHz of the 800MHZ and 65MHz of the 2.6GHz spectra. Basically roughly a third of what was available. This should give them the most optimal flexibility to provide a mix of in town and rural services. Vodafone, which is sharing infrastructure with O2 under a venture known as Cornerstone Telecommunications Infrastructure Ltd has like O2 stated a goal of hitting the 98% population coverage before 2016.

EE

The next biggest spender at £589 million was EE bringing it 70MHz of the 2.6GHz spectrum and the “minimum buy” of 10MHz from the 800MHz band. EE is already offering 4G services in the 1.800MHz band so whilst its 800MHz holding may be a little light the company still has a good spread of spectrum.

EEs head start in this game also adds an additional competitive dimension to the whole business.

O2

The O2 spectrum allocation is an interesting one. Paying £550 million for the 20MHz coverage obligation lot in the 800MHz spectrum it has paid more for this band than either EE or 3 if we use the reserve price of £225 million paid by 3 as a benchmark. It isn’t as simple as that but without spending days analysing the finer points of the auction it serves a purpose as a rough guide.

O2 is obliged to cover 98% of subscribers by 2017. In my mind this is something they will have been wanting to achieve in any case so the “obligation” is unlikely to be a particular burden. The relatively high cost of the spectrum combined with what I imagine to be a higher cost of serving rural districts might raise an eyebrow but I am not party to O2’s infrastructure cost model and plan.

What is potentially more likely to be an issue for O2 is the absence of a holding in the 2.6GHz band. O2 may be pinning its hopes on being given future permission to use its 2,100 MHz spectrum holding or there may be a clue in the recent announcement that BT will be working with O2 to provide the backbone network for O2s 4G services. Having the high capacity backbone is fine but not particularly necessary if you don’t have the spectrum capacity to drive traffic.

Hutchinson 3G UK Ltd

There isn’t that much to say about 3 really. The company was guaranteed some spectrum in the Ofcom process. It paid the reserve price of £225 million for the smallest allocation of all the networks. 3 is currently by far the smallest mobile operator in the UK ignoring the fact that BT seems likely to re-enter the market and one wonders what the long term plans are likely to be. Will they change their name to 4? 🙂

BT

BT’s acquisition of 2.6GHz spectrum at £186 million represented by far the lowest cost per meg and so on that basis the incumbent fixed line operator seems to have got the best value out of the auction.

However the obvious spectrum for BT to have gone for would have been in the 800MHz range, assuming its intention is to use 4G to improve its broadband penetration to rural areas. This would be consistent with BT’s increasingly monopolistic position in the “final third” of the country as seems to be suggested in the awarding of Government BDUK contracts.

I don’t have any insight into their plans but when people look back with the benefit of hindsight at some event or other they often say “of course it was obvious”.

If we look at the BT and O2 positions in respect of spectrum it would seem obvious that both parties should share their respective spectra. I wouldn’t rule out some kind of re-convergence of the twain/merger bearing in mind that BT sold off O2 at some stage in the medium term past.

O2 has just offloaded its fixed broadband business to Sky. BT is getting back into mobile. O2 owner Telefonica is a Spanish company and therefore likely to be suffering from the woes of the Spanish economy. BT could buy O2! Pure speculation but makes sense to me.

That’s it as far as an analysis of who bought what in the 4G spectrum auction. A bit overdue but sometimes these things benefit from digesting the information for a little while before rushing to gain first mover/publisher advantage in the SEO stakes. In the meantime developments have allowed me to add an additional dimension to the analysis.

I doubt we will have to wait much long to find out where this whole space is going. You can read up about my thoughts on time to market here.

28/9/13 update – comparison of O2, EE and Vodafone 4G networks in London