The shape of South Africa’s telecoms network infrastructure has undergone significant changes in the past twelve months, with further shake-ups on the horizon. From operator-led towercos and Cell C’s plans to rebuild their tower portfolio, to American Tower’s acquisition of Eaton Towers and the Government’s proposal to create a Wholesale Open Access Network. Such developments made for interesting debate at TowerXchange’s 2016 South Africa roundtable as participants voiced their concerns, ambitions and forecasts for the year ahead.
The current shape of the South African mobile market and tower industry
South Africa has four MNOs, with Vodacom and MTN leading the market with 38% and 36% of mobile subscribers respectively, Cell C possessing 23% market share and Telkom accounting for just 3%. There are also a number of MVNOs which hold less than 1% of the subscriber base (figure one).
The country has an installed base of just under 30,000 towers with each of the MNOs, bar Cell C, retaining their tower portfolios. Whilst Cell C had sold their portfolio of 1,400 sites to American Tower back in 2010, the operator is now in the process of rebuilding their own network (more of which later).
Having entered the market following the acquisition of the Cell C towers, American Tower are the largest towerco in the South African market with a portfolio of 2,309 sites after their recent acquisition of Eaton Towers’ 300 South African sites absorbed their largest competitor. In addition to American Tower there is a long tail of other independent towercos in the country, with fast-growing Atlas Tower and their 171 sites heading the group. In addition to the MNOs and towercos there are approximately 7,500 sites managed by broadcast, web and other industries (figure two).
Figure one: MNO market share of South Africa’s 85.1mn subscribers (FY2015)
The rise of the operator-led towerco
Much discussion on the South African table this year centred around the role for towercos in the South African market, particularly in light of the fact that each of the South African MNOs are now actively pursuing co-locations on a commercial basis between themselves; effectively functioning as operator-led towercos.
The success of the towerco business model, referenced one of the MNO representatives present, has demonstrated to operators that there are decent revenues to be made in the space and as such, leasing up their towers is becoming an increasingly important strategy. What’s more, South Africa’s robust and extensive electricity grid means that the energy challenge and costs faced by MNOs across much of sub-Saharan Africa are of less concern to South African operators and as such, much of the value that a towerco can add is removed.
Figure two: Ownership of South Africa’s ~29,500 towers
The balance of towercos and operator-led new build of macro sites
Whilst the capability of MNOs to build, manage and lease up their own tower portfolios is increasing, participants at the table raised the fact that operators had undergone a massive rationalisation of staff in recent years, through which much of their expertise in rolling out new sites had been lost. It was also observed that historically MNOs had been given large reserves of capital with which to acquire new sites but that the funds had often not been well used, often sitting untouched. As a specialist skill set in high demand, some participants felt that it would take operators a while to build up highly effective teams in order to manage rollout more effectively.
Acquiring sites in key locations represents a major challenge for operators and towercos alike. It was suggested that American Tower’s recent acquisition of Eaton Towers was as much about the latter’s portfolio of 1,000 sites under development as it was motivated by the acquisition of their existing portfolio of 300 sites, and that generally towercos had been more successful in acquiring highly desirable sites than had the MNOs. Operators at the table agreed that this was one of the key strengths of towercos in the country.
Whilst towercos argued their ability to rollout sites minimised capex spend for MNOs, one of the operators explained that their own ability to access cheap capital, combined with the high towerco lease prices in South Africa, meant that in current market conditions it sometimes made more financial sense for MNOs to build their own sites. They did however explain that there are certain times of year when they are more capex constrained and in these instances it would then make sense to work with towercos more closely to manage network expansion.
Reflecting on whether new site build was best managed by towercos or operators however, MNOs agreed that the level of demand for new sites over the next two years would mean that both would have an important role to play. The volume of new site build required far outstrips the capacity of MNOs to deliver such numbers alone and as such, towercos will be key partners in the process.
Whilst towercos are a useful partner, one area where network planners would like to see an improvement was on the visibility of sites that towercos were able to offer. It would be highly valuable for network planners to be able to automatically view the stage of development or construction of each of the towerco sites, enabling them to weigh up their options more efficiently when designing the timely rollout of their networks.
Ground leases aggregators and property concerns
Escalating ground lease rates are causing significant challenges, as the price of real estate continues to rise exponentially in South Africa. One participant referenced how it was hard for the sector to keep pace with the amount of money that is filtering into real estate in the country and expected the problem to become more acute.
When questioned on their appetite to own versus lease land, it was referenced that towercos will aim to buy the land when they can or aim for long term (~40 year) leases. One of the challenges in purchasing land, referenced one of the MNOs present, is that you only want a small area, however the effort that goes into portioning off a section of land often doesn’t make the exercise worth it. One of the smaller independent towercos referenced that this is where their strength really comes in. As a smaller, more nimble company, they can look at alternative use of the land that is not being used by the tower - for example, a small factory could be built on the land, thus finding further ways to monetise their investment. A towerco’s ability to do this is one of its key strengths above an MNO; flexibility is their niche.
Participants at the table reported seeing an increasing number of approaches from ground lease aggregators in the market, whose involvement will further intensify property management challenges that MNOs and towercos are facing.
Speculation surrounding an MTN tower sale
It would not be a South African roundtable discussion without questions surrounding the likelihood of an MTN tower sale being raised. With the company’s primary focus being on opex reduction at present, a tower sale was not something that participants expected to happen imminently, although of course there were plenty of interested parties present.
There had been rumours circulating in the public domain that IHS had been in discussion with MTN regarding a tower sale, with the towerco having acquired the operator’s sites in Cameroon and Cote d’Ivoire and the two parties having entered into a joint venture in Nigeria. It was referenced however that the South African market was very different to these countries and as such there would be different motivations and considerations when looking at a deal.
Should MTN’s towers come to the table, however, it was observed that a deal would be complicated by the fact that around 1,500-1,800 sites are shared with other MNOs, with some of them having been paired off with other operator sites. Uncoupling these sites and starting to invoice each other for space used would present an added level of complexity to the execution of such a tower transaction.
As to how such a transaction would impact on smaller towercos in the market, representatives from such companies felt that there would be a positive impact. When a major transaction occurs, one participant observed, a vacuum is created in the country whereby the other MNOs come rushing to the smaller towercos.
The American Tower - Eaton Towers transaction
One of the biggest talking points in the South African tower industry in the past year was American Tower’s acquisition of Eaton Towers, which had been met by opposition from the MNOs in the market. With American Tower possessing just over 2,000 of almost 30,000 sites in the country however, South Africa’s competition commission ruled that there was no case to answer and as such the merger was cleared.
Participants observed that it was no secret that American Tower’s lease rates were considerably higher than those of other towercos in the market (with one party suggesting rates were almost double), as function at least in part of the high leaseback rate agreed with Cell C, a deal which was structured to maximise cash released at the time of sale. One MNO suggested they had a moratorium on the use of American Tower sites, at least until their lease rates came into line. A merger between the American Tower and Eaton Towers has caused concern from MNOs using Eaton’s sites, however measures have been put in place to protect existing contracts.
On the subject of the transaction, the smaller towercos present felt that American Tower’s acquisition of Eaton has had a positive effect on their business. Should the Eaton portfolio have been acquired by a new entrant to the market it would have caused concern, but the consolidation of the number one and number two towerco has driven further business towards the smaller independent players.
Evolving pricing mechanisms and business models
One operator at the table said that they had been undertaking benchmarking of towerco pricing, finding considerable variation in the structure of towerco pricing models. At one end of the spectrum, American Tower’s pricing is pretty much all inclusive, whilst at the other end of the spectrum some towercos charge extras for absolutely everything.
MNOs questioned towerco appetites to offer discounts to anchor tenants as additional tenants are added to sites. One towerco referenced this was a model that worked well in other markets and is something that they were open to discussing.
One participant at the table questioned towerco appetite to get into fibre as a means of offering additional services in the market. Whilst some towercos are starting to assess this, one company referenced that the adding fibre to the towerco business model hasn’t always been effective in the US, as the anchor tenant usually wants to control the fibre and make money from it. What’s more, with so many fibre players in the South African market already participants questioned whether there was a role for a towerco to play.
One area where participants did see towercos expanding successfully into was small cells and indoor DAS. Whilst in the instance of the Mall of Africa the role for an independent infraco was bypassed with the four MNOs collaborating on an extensive DAS installation, instances where not all MNOs are keen to invest make sense for an infraco to get involved. The proposal by one towerco that MNOs could opt for coverage in just select areas and pay for the service they use was met with approval from MNOs present.
The move to 5G and the role of disruptive technologies
Participants at the table forecasted that 5G would explode in the US and Europe in 2019 and 2020, with this filtering through to South Africa around 2022/23. Most at the table felt, however that there was still some way to go in understanding how the deployment of 5G should be executed, with one person commenting that RF planners and real estate teams often have very different ideas.
All were in agreement that macro sites wouldn’t go away, rather you will see 4G and 5G equipment hung on towers in place of the 2G and 3G that is there today. When it came to deploying small cells, it was commented that you wouldn’t always need permits, all you might require was a fibre connection with one participant commenting that simply installing a network of routers in residential properties would create a small cell network, whilst others talked about the Bluetooth experiment that had been conducted in Boston to create a communications network.
On the subject of whether investing in street furniture constituted a sensible decision for towercos, some felt that it constituted a very high capex investment for very little return whilst others observed that this had been big in Europe since the 1990s.
Creating new synergies
With increasingly thin margins across the industry, discussion moved to the synergies that can be created by a single party managing maintenance of both active and passive equipment, with operators referencing that this would be a desirable outcome. What’s more, reducing the number of people with access to a site also reduces risk.
One participant at the table suggested they felt there needed to be closer interaction between MNOs. Rather than each operating, rolling out and building their own network of sites and creating parallel infrastructure, there should be improved communication so that the rollout costs can be shared. If the government 2030 targets are to be met, the participant continued, it makes sense for MNOs to consolidate budgets. With MNOs typically reluctant to discuss their network planning with their competitors, one participant questioned how such a model could be executed, whilst others raised concern over when this would constitute anti-competitive practice. A potential solution, one person suggested, would be the creation of a neutral independent party to oversee the process.
The government’s draft ICT white paper and the spectrum auction
Discussions around pooled resources led to another big topic of the day, the South African Department of Telecommunications and Postal Services’ National Integrated ICT Policy White Paper.
The paper calls for radical changes in the telecommunications market, including the creation of a Wireless Open Access Network (OAN). All unassigned high-demand spectrum (essentially LTE spectrum) would be set aside for the OAN and what’s more, the paper also puts in place the suggestion that the government can take back spectrum previously assigned to the MNOs. The white paper proposes that competition will only happen at the service level and that the OAN will be managed by a consortium which will operate with ‘competitive neutrality’.
The white paper has been met by opposition particularly amongst the larger operators who see their network as a competitive differentiator, and who have concerns about spectrum not being allocated in proportion to the number of subscribers each operator has, a move they feel which will compromise their quality of service. The announcement of the white paper on spectrum caused MNO share prices to fall, a downward trend which has persisted.
The white paper has also held up the recent spectrum auction in South Africa, with the Independent Communications Authority of South Africa (ICASA) having made the decision to proceed with the auction, only for this to be blocked by the High Court following a challenge from the Department of Telecommunications and Postal Services.