4G Auctions: Making the Spectrum Pay

POSTED BY Nitin Bhas
Just under 13 years ago, the mobile industry went briefly but entertainingly bananas. At the announcement that spectrum would be auctioned off to the highest bidder in several countries, it persuaded itself that 3G would be the Holy Grail, the Philosopher’s Stone, the Golden Goose and Angelina Jolie all rolled into one: a wondrous, beautiful panacea for all ills that would set their world and bank accounts to rights for years to come, no matter how much it cost to begin with. They weren’t entirely sure what they were going to do with it when they got, but hey, they were sure that something would work out. Several months later, and five UK 3G licences had raised £22.5 billion (then around $34.4 billion); in Germany, they went for the small matter of €50.1 billion (then $46.6 billion); in Italy, for   €12.2 billion (then $11.3 billion). And, indeed something did work out for the mobile industry: thanks in large part to overspend on 3G licences (and that other business with dark fibre lease commitments), it nosedived sharply for a few years. The 3G licences cost the MNOs, collectively, $196 per head of the Italian population, $567 per German and $586 per UK resident. It then took the respective licensees in Germany and the UK the best part of a decade after launch to recoup the licensing costs – and that was without factoring in infrastructure and running costs. One reason here was that, disappointingly for the MNOs, video calling had failed to be a killer app, the networks weren’t up to delivering streamed video and thus the licensees were rather strapped for ideas on how to monetise said networks. Then along came the OTT players and WiFi and the rest is history. And so to 4G. The environment  - both technologically and economically - is rather different now than it was in 2000: since late-2008, Western economies have been fluctuating in and just out recession, which is hardly conducive to inducing operators to spend big on spectrum; secondly, the operators, have paid way beyond their means in 2000, are typically more cautious in their approach; thirdly, while the OTT players (net spend on licensing – nil) would welcome more bandwidth with open arms, the MNOs (net spend on licensing – a fair whack) aren’t entirely sure how to extract value from the networks once they’ve rolled them out. Hence, in Germany, when the 4G spectrum was auctioned off in 2010, the net spend on licences was a rather more modest $56 a head. Now, had the UK MNOs paid at that level , then they would have raised $3.53 billion for the Treasury, or £2.31 billion – as it was, they actually paid marginally more, at just under $57 per head. Which was some way short of the UK Treasury estimate, which required spend of around $85 per head of population to realise the £3.5 billion ($5.3 billion) it was hoping for. Given the prevailing economic conditions, given that the MNOs had had their fingers burnt 13 years ago, given that LTE premiums from existing services elsewhere are moderate rather than spectacular, and given that there are mobile broadband alternatives just about everywhere in urban areas (hello, WiFi hotspots) such spend levels were highly unlikely. As my colleague Nitin Bhas has observed in his recent report on 4G LTE, even at these significantly reduced levels of investments, the operators will take several years to recoup their expenditure. Furthermore, the transition to an all-IP environment – where the OTT providers will gratefully piggyback to deliver not just content, but voice and messaging services in direct competition with the core MNO offerings – poses a huge challenge to the established MNO business models: hence the fact that the old “voice ARPU” KPI is becoming less and less important as MNOs adjust their offerings accordingly. But at least we should be thankful that the UK auctions are finally done and dusted, and that – following EE’s launch last year – the UK market will at last see some competition. Whether the operators can successfully develop the business models to successfully monetise the spectrum is another matter entirely.