Why mobile data should be more like mustard

Industry analysts and commentators continue to marvel at the “growth in demand” for mobile data and indeed some mobile networks are groaning under the strain. It appears however that the meaning of the word demand has been forgotten. Demand, in an economic sense, relates to the willingness to pay. One could say that there is a “high demand” for Ferraris as who would not want one, but few can afford them which is why they are not clogging up the motorways.

The prevalence of unlimited or “all you can eat” data offerings has resulted in significant growth in traffic but not revenue as the link between demand and the willingness to pay has been broken. Unlimited offers made some economic sense two years ago when adoption of mobile broadband was low and networks had unutilised capacity. Two years ago incremental traffic did not necessarily result in incremental capital expenditure. Networks are now congested and increasing traffic is resulting in increasing capex without increasing revenues which can only depresses returns to shareholders.

Pricing structures need to change if the economics of mobile data are to improve. Within a mobile operators’ customer base there are usually a relatively small proportion of customers who use a very high proportion of network resources. These “bandwidth hogs” should be charged for the data that they use or their usage should be constrained. In fact mobile operators should start to look at mobile data a little more like mustard.

Since 1814 Jeremiah Colman of Colman’s English mustard has been making his money not by how much mustard people eat but how much they left on the side of the plate. If you are going to charge somebody £15 a month for 2GB the customers you actually want are those who use only a quarter of their allowance. Studies show the average smartphone user is using a lot less than 2 GB a month, and perhaps a tenth of what the average dongle user gets through. If operators started to think about mobile data a bit more like mustard we might see an improvement in industry returns.

Heretical thinking: New spectrum may not add value

Hardly a day goes by without a new forecast of mobile broadband demand being published predicting exponential growth in data volumes. This is then translated into a requirement for additional spectrum allocation over and above the digital dividend 700/800 MHz spectrum and the “LTE spectrum” in 2.6 GHz.
Speakers at the 5th Annual EU Spectrum Management Conference in Brussels in June 2010 and the GSMA Spectrum Workshop in the following week articulated a need for more mobile spectrum and urged regulators to make more spectrum available through auctions. Because of the explosive growth in mobile broadband demand, spectrum is deemed to be very “valuable”.

These mobile broadband forecast use the term “demand” when talking about the growth in data volumes. However, incremental demand in the economic sense means incremental revenue, i.e. users must be willing and able to pay for it. Certainly we are not seeing exponential revenue growth. In the mature markets of Europe and North America overall mobile industry revenue is not increasing, and in many cases declining. New revenue from mobile broadband barely covers lower revenues from voice calls, reduced mobile termination rates and lower SMS revenues.

Yet in order to cope with the mobile data tsunami operators buy new spectrum in auctions and subsequently incur further substantial additional capital expenditure (capex) for technology deployment in the new frequency bands and backhaul. In recent years, much to the relief of investors, mobile operator’s ratios of capex to sales had declined and cash flow generation was increasing. Unfortunately, with new spectrum and new technology deployment the trend is again reversing. Shareholders must be asking themselves the question where is the value in this additional spectrum?

On the basis of this analysis Stefan Zehle, the CEO of Coleago Consulting presented a contrarian view at the last GSM Spectrum Workshop, held in London on the 29th of June 2010. In short, once the digital dividend and 2.6GHz spectrum is auctioned, mobile operators should stop pressing for new mobile frequency allocations.

The manner in which mobile operators value spectrum is to work out the additional value that could be generated from owning it. In simple terms it is the difference in the Net Present Value (NPV) of the business with and without the spectrum. If new spectrum is being offered at auction and operator A buys it whereas operator B does not, then operator B would be at a competitive disadvantage in terms of capacity or ease of building mobile broadband coverage. This would translate into reduced cash flow, for example through a loss of market share due to the lack capacity to serve mobile broadband customers. If operator B were to acquire the spectrum, the reduction in revenue would not occur and this “saved” revenue drives the value of the spectrum. Therefore from an operator’s perspective the value of the spectrum is not driven by generating new cash flows but by preventing a reduction in existing cash flows. This leads to the conclusion that if total mobile industry revenues do not increase, new spectrum is a bit like an arms race.
Consider the contrarian case. If no new spectrum is being made available in a country, the industry as a whole in that country may become capacity constrained in terms of mobile broadband. As a result prices for mobile broadband will increase and the return on capital employed will increase. There are no cash outflows to buy spectrum and no new deployment capex. As free cash flow increases, share values increase. Furthermore without new spectrum the threat of new market entry is blunted.

Therefore after the completion of the release of the digital dividend spectrum, the auctioning of 2.6 GHz and re-farming, mobile operators should not press to make more spectrum available for mobile communications. Naturally regulators would be keen to make more spectrum available because this is in the interest of consumers.