Even where regulators are encouraging new players, risks remain

There are exceptions to the gloomy picture outlined in the previous item. As noted, Ireland – whose regulator, ComReg, has often awarded short term or low cost licences – awarded a 3.5 GHz licence to Airspan and another to fixed wireless provider Imagine. And it split the airwaves between rural and urban allocations to ensure 5G would reach country areas too.

In total, in that auction (concluded in May 2017), 350 MHz of spectrum in the 3.6 GHz band were sold for more than €78m, split between upfront fees and usage charges.

Vodafone Ireland won the biggest chunk, claiming 85 MHz for use in rural regions and 105 MHz for the cities, at a cost of €22.8m. Meteor, a subsidiary of telco Eir, won 80 MHz for use in the countryside and 85 MHz for cities. It paid €15.6m. Three Ireland bought 100 MHz for use nationwide, spending €20.3m.

Airspan got rather less, but still a useful 85 MHz (60 MHz urban), for which it paid €9.6m. It is focused on deploying networks for vertical markets, especially utilities, transportation and public safety. And Imagine spent €9.7m on 60 MHz of rural to add to its existing holdings in 2.5 GHz and 3.5 GHz, which it used for WiMAX and, more recently, TD-LTE.

The licences will run for 15 years. ComReg said the sale would increase the amount of harmonised spectrum available by 86%, highlighting the importance of relatively high frequency bands to add capacity to the mobile market, especially for local urban or rural networks based on small cells.

ComReg chairperson Gerry Fahy said: “Continuity for existing services has been underpinned and the possibility of new services has been significantly enhanced.”

That balance between enabling the MNOs to continue to invest in existing and new business models, while encouraging new players, is also being sought by Canada’s government, which says it will exclude the country’s three national operators – Bell BCE, Telus and Rogers – from bidding for 43% of the spectrum in its forthcoming 600 MHz auction.

This is notable because the sub-1 GHz bands are usually the least likely to include allocations for new entrants, because of the limited amount of spectrum available, and because operators regard these bands as the most valuable, and so will pay large sums per MHz. But Canada’s federal innovation minister, Navdeep Bains, wants to see more competition and told local reporters: “The fact that we have an auction that will set aside 43% of the spectrum demonstrates our commitment to competition, which will provide more options to consumers which ultimately means better price points.”

The auction is set to take place in March 2019.

Of course, allowing new entrants to secure spectrum at reasonable prices is only the first step to opening the market for innovative services and increased competition. The winners of the licences need to have the financing, expertise and business model to deploy and monetize a mobile network, and in the past, this has certainly not always proved the case. Indeed, a national or large regional, sub-1 GHz licence – best suited to wide area mobile connectivity, in direct competition with established MNO networks – is a high risk way to encourage new players. Deploying such a network is a daunting undertaking when starting from scratch, and the competitive landscape is tough in most developed markets.

That is why assigning spectrum on a more localized or vertically specific basis – which favors the higher bands where there is more capacity to share around – is a less risky way forward, and leaves the MNOs with the incentive to continue to enhance the national mobile networks, since they will continue to control and monetize those.

Canada has been one of the most eager countries to support new entrants, but has had limited success in terms of a more diverse market, since many of those spectrum winners of the past have been acquired by one of the big three or have failed commercially. In 2015, Rogers refused to participate in the AWS-3 auction because of the regulator’s insistence on reserving spectrum for new bidders. Eastlink, Videotron and Freedom Mobile all bought spectrum, but have not made a significant impact on the Canadian mobile landscape. In 2008, three companies – Wind Mobile, the forerunner of Freedom; plus Mobilicity and Public Mobile – all won AWS-1 licences, but Mobilicity was acquired by Rogers and Public by Telus.

The USA has also sought, periodically, to encourage new players with reserved spectrum, or by offering regional or local parcels of spectrum. However its trading laws mean that the spectrum is often acquired as a tradeable asset, to be sold on to the major operators later; or even by companies fronting for the big MNOs. And the failure of would-be new operators like NextWave has shown how hard it is to get the balance right between supporting low risk network deployers, and spurring innovation.

All regulators have to strive for that balance between seemingly incompatible goals – to encourage competition and guard against monopolies; and to ensure rapid roll-out of the best services for consumers.

The challenge has got harder with mobile broadband – the services everyone wants to support their rising data addiction require hugely expensive networks, and the cost of the infrastructure can rarely be borne by a new entrant.

One solution is to split the infrastructure from the service providers, and this may well be the end game. Steps along the way, like MVNOs and RAN sharing, and eventually slicing, are gradually being approved by the world’s telecoms watchdogs, while other steps, such as over-the-top services, are happening without any real regulation at all. But in the short term, most countries still have a system of several competing carriers which have to invest in their own infrastructure, which requires the kind of scale that may work against innovation and entrepreneurialism.

The difficult balance has always plagued US mobile telecoms, where there have been successive disasters surrounding the bid to support counterweights to AT&T and Verizon (NextWave, LightSquared and so on, not to mention MVNOs like Helio). Scale is everything.

Of course, there have been successful and disruptive new entrants in recent years, most famously Free Mobile in France and Reliance Jio in India. Both of these were backed by deep pocketed parents, and in Free’s case one with a broadband network already in place to reduce the costs of backhaul. In both cases, they have largely undercut the existing MNOs and focused on offering the same kind of services at more competitive rates. This has been, in the short term at least, good for consumers, but has led to financial problems, and a wave of consolidation, among the existing operators.

The end result has been a reduced overall ability to invest in new vertical and enterprise services, which require a different kind of network to consumer broadband, and in many cases, a very high quality one. The existing MNOs have less revenue and profit, while the newcomers have focused on the same services.

This is why regulators need to start thinking about new approaches to licensing which lower the barriers to entry, but help encourage new players which will support new classes of service, especially those which MNOs are reluctant to offer. Saturated markets cannot support larger numbers of national operators providing mobile broadband, but they badly need a new spectrum and financing model to make it easier to device new types of connected services, and so achieve 5G’s hoped-for impact on the broader social and economic landscape.