How much did 3 pay for EE's spectrum?

Plenty has been written about Everything Everywhere selling 30MHz of 1,800Mhz spectrum to the UK's smallest operator 3 (including this, which is really good). Much of the aforementioned text is a bit sensationalist or misinformed, so I thought I'd summarise what's gone on and also attempt to answer the one thing that hasn't been covered: the price.

In summary:

• Everything Everywhere (the result of the merger between Orange and T-Mobile in the UK, for those of you who haven't been keeping up :) ) were forced by the EU to sell a proportion of their collected spectrum in order that the market remained "fair"
• The spectrum is at 1,800MHz ("1800") and was originally used for carrying 2G mobile signals
• Although that frequency can be used for LTE, it isn't often. The more common bands are 800 and 2600, although others - including 2100 and 1800 - are sometimes used
• In any case, the spectrum won't be fully released to 3 until September 2015, long after the long awaited "digital dividend" auctions have come and gone
• Everything Everywhere have also obtained permission from Ofcom to deploy an LTE network on some of their remaining 1800 spectrum
• LTE is roughly 4 times faster than the fastest 3G broadband and therefore theoretically enables operators using it to outcompete those who don't have it
• Although EE's competitors are furious about this, in reality its effect on the market should be somewhat limited by device availability. 1800 is not supported by the only LTE device sold in the UK - the iPad HD - and it may not be supported by the next iPhone, the international varient of which is more likely to have 700, 800 & 2600 compatibility
• That said, there are some LTE handsets - most notably the Samsung Galaxy S2 - that do support this type of LTE and there are also modem ("dongle") products available that would bring high speed wireless data to users acquiring one
• Coverage could be quite reasonable to start with, since Huawei have been upgrading Everything Everywhere's 2G basestations to give them HSUPA (3G) and in future, LTE capability
• In my mind there is a radio planning question that remains to be played out, in that the most valuable spectrum for LTE is at 800. This is because lower frequencies give better coverage due to their lower propensity to be attenuated by objects in the environment. 800 is worth about 4 times 1800 on a per-MHz basis
• Anyone planning an LTE network would therefore be better off planning for 800, which won't necessarily create the best 1800 network
• Ultimately, the effect on the UK market of both the EE LTE launch and 3 acquiring extra spectrum will be very small. In reality, getting 1800 outside of an auction enables 3 to obtain a lower price than they may have got in a competitive process for 2600 spectrum. They will still need 800 to be genuinely competitive

All this brings me to the sticky issue of price. Based on my international benchmark for spectrum price, I estimate that a "fair" price for 30MHz of 1800 would be £200Mn - £250Mn.

But I think 3 might have paid less than this. Firstly, the sale was effectively uncompetitive - no other operator could buy the spectrum because Vodafone and O2 are already too close to the limits on the amount they can own. Secondly, 2600, which is the more common high-frequency LTE frequency is worth about half as much as 1800 in an auction. As mentioned about, 1800 is also not a common LTE frequency. Thirdly, I understand that 3 are very happy with the price they paid - sounds like it was low.

So I'm going to revise my estimate down to £175Mn - £200Mn. Cheap, but not significant. I still expect 3 to be consolidated by another player soon after the auctions complete.