Lessons learned from the MNO panel session at the TowerXchange Meetup Africa

Prakash Ranjalkar has been fostering Airtel’s passive infrastructure assets in SSA until their imminent transfer to the towercos; he currently serves as CEO of Airtel’s Africa Towers entity. TowerXchange ‘Inner Circle’ advisory board member Michel Faivre has led Orange’s infrastructure sharing programme for several years. Tim Knowles represented both an M&A perspective and that of Etisalat, who of course recently sold 2,136 towers in Nigeria. Riana Donaldson was a key part of the team that outsourced towers to Eaton at Vodafone Ghana – she now serves as Manager: Network International for Vodacom. And Sudhir Chopra, Group CTO of Smile, represented the tenant’s point of view as the foremost of Africa’s many new entrant 4G plays. The panel was ably moderated by Gulfraz Qayyum, Managing Director of TMT Investment Banking at Citi.

How each operator introduced their tower strategy:

Etisalat is present in 19 countries, with a leadership position or strong challenger status in most markets. As such, they are considered a highly desirable counterparty in any tower transaction. During 2014, Etisalat’s West African business was restructured through the sale of the Atlantique Telecom / Moov assets in six countries to Maroc Telecom and Etisalat’s acquisition of a 53% stake in Maroc.

Etisalat’s Head of M&A Tim Knowles revealed there was no group wide strategy toward towers. Indeed, Andrew Kemp CFO of Etisalat Nigeria, championed the sale of their Nigerian towers in 2014, a market in which Etisalat needed to raise funding and reduce opex, but also needed to bring their towers to market in a timely manner in to get best price, the Airtel and MTN processes having triggered a scramble to acquire Nigerian towers.

Etisalat doesn’t see tower sales as the sole means of generating efficiencies; they have also outsourced managed services and created 3G joint ventures in selected markets.

Airtel has a stated strategy to ensure their towers are managed in a focused manner – a large capital investment has been deployed, and they see passive infrastructure as critical to operations.

The sheer number of towers and the complexity of operations in India prompted the creation of Bharti Infrastructure, within which Bharti manages their own Indian towerco Bharti Infratel and their stake in joint venture towerco Indus Towers. Airtel’s tower strategy has played out successfully in India, reducing costs and enabling expansion while driving tenancy ratios over two.

From 2010 it was Bharti Airtel’s intention to share towers in Africa but, given the smaller individual tower portfolios across such a diverse SSA footprint, they found they couldn’t replicate the carve out towerco model they used in India. Nor did efforts to attract partners in MNO joint venture towercos achieve much traction in SSA. With the growing wave of tower transactions in SSA, Airtel felt it was better to consolidate and bring their African towers into the hands of expert independent towercos. Thus 15 months ago Airtel commenced an Africa wide tower sale process, described (in October 2014) as 70% complete, and anticipating that by Q1-2 2015 their partners will have been selected in each of Airtel’s SSA countries.

By offering up Airtel’s “entire platter” of African towers, whilst being prepared to carving up the portfolio and allow towercos to mix and match to select the countries which best matched their investment thesis, Airtel has been able to offload towers in their smaller perhaps less attractive markets, bundled with the assets that more obviously met towercos’ criteria.

The drivers of Vodacom’s tower strategy are the stabilisation of cost and optimisation of QoS.

Vodacom’s only tower transaction to date has been in Tanzania, where they had previously outsourced the management of passive infrastructure to Nokia. They see the Mozambique market somewhat driven by coverage, creating a specific niche opportunity for a more rurally focused tower venture that could focus on supporting coverage objectives. Vodacom undertakes substantial bi-lateral tower sharing in DRC, but retains their passive infrastructure assets in the country.

Orange’s strategy is to share the network as much as possible to improve QoS and optimise rollout capex, particularly in the context of the difficulty securing authorisation to permit new sites.

Orange adopts a different tower strategy model from one market to the next, according to market need. Orange has agreed managed services deals with IHS in Cameroon and Cote d’Ivoire, and had sold towers to Eaton Towers in Uganda prior to the sale of their license to Africell.

Smile considers outsourcing and infrastructure sharing as two pillars of their strategy – they believe they can achieve better performance, better KPIs at a lower cost by sharing towers. Smile has evolved to focus on mobile broadband internet using 4G LTE technology – they launched Africa’s first 4G LTE service in Tanzania, in March 2012, and now also provide 4G LTE services in several cities in Uganda and Nigeria. Sudhir noted that as recently as three years ago there was hardly any sharing in Nigeria, yet today over two thirds of towers are shared.

“We’ve tried sale and leasebacks, we’ve tried to create MNO joint ventures to which we then add a towerco partner – the right deal structure depends on local market requirements,” said one operator.

Tower strategy isn’t just about the management of existing towers, it’s also about getting new towers built. Build to suit programmes are typically bundled with tower divestitures. Tower strategy is also a larger corporate finance consideration. Few African opcos are 100% owned by their parent companies, and if local shareholders aren’t prepared to contribute further capital, monetising passive infrastructure is an attractive alternate way to raise capital. Every shareholder has their own view of the potential conflicts of interest between MNO and towerco, which is why stakeholder engagement is critical and can affect how empowered the local management team is to manage balance sheet considerations such as divesting their tower portfolio. Sometimes it is simply not possible to sell the towers.

Orange’s Michel Faivre called attention to the many efficient joint venture infrastructure sharing business models working very effectively in Europe, particularly in fiercely competitive mobile markets, and to the potential for such models to be implemented in Africa.

Orange favours an outsourcing model with managed services – very similar to sale and leaseback, but the only difference is that Orange retains ownership of the towers. They still have an SLA to leverage QoS improvements, they can realise good prices as the towerco can still create extra cash flow through co-location sales. Orange typically bundle a build to suit programme, and the towerco still takes over power. While there are exceptions, generally Orange prefer to retain towers.

The question of whether to retain equity in joint venture towercos was raised, although in fairness it should be noted that two of operators that seemed most inclined to retain equity, MTN and Millicom, once again declined to participate in this year’s debate. Vodacom indicated a preference to retain a little equity and thus retain a bit of operational control, for example control over quality of towers.

The other MNOs didn’t seem to feel retaining equity was necessary as some sort of hedge against mispricing of deals. Their feeling was that the independence of the towerco was key – that an anchor tenant’s representative on a towerco board could have minimal influence on capital deployment, opex reduction priorities or on third party pricing without restricting the towerco. Such considerations certainly factored into Airtel’s thinking when they resolved to sell 100% equity in their African towers – with portfolios of 500-2,000 towers in most countries except Nigeria, Airtel didn’t have an appetite to participate in 15 or 16 joint venture towercos.

Of course pricing is a risk – it’s a trade off of cash released and reduced monthly opex bills against the leaseback rate agreed. Ultimately, tower transactions are hugely complicated – sometimes expediency prevails, an attitude of “we’ve got we need to get it done and move on”.

Tower transactions shift capex to opex, which materially affects the EBITDA reported to stakeholders. EBITDA targets have to shift and other KPIs need to be adjusted accordingly, for example benchmarking exercises must take the new structure of the tower-outsourced opco into consideration. Such performance measures aren’t always as aligned as they should be!

Opinions differed in response to this question. One MNO was not concerned about oligopolistic behaviour in the near term, but felt any consolidation among Africa’s ‘Big Four’ towercos would have to be monitored closely.

Another operator had a nearer term concern that a dominant towerco may have a temptation to use their position to keep prices high, and may have less incentive to optimise services. The term of towerco contracts extends beyond several generations of technology, which itself may have implications for the operating costs for towercos and therefore for the lease rates charged to tenants.

A third operator felt that under a regime of passive infrastructure management by towercos, costs had already been proved to come down, with MNOs and towercos striking partnerships where is one was successful the other would share in that success. Fear of loss of control, quality degradation, and loss of pricing power had all been overcome, he felt.

The growth of independent towercos has improved the business case for newer operators like Smile Telecom, who originally felt they’d have to build a significant portion of their own towers, but find that 100% of their current towers are shared. Even though Smile was not an anchor tenant, they didn’t feel there was any overt favoritism, and they were particularly impressed that even in cases where the towers they wanted were idle, the towerco invested to deliver QoS at the right price.

A note of caution was shared by one operator, who expressed concern about the difficulties getting rural sites (often destined to become single tenant towers) built in a market where one towerco was dominant. The same operator also expressed reservations about fair pricing of towers to be transferred to a dominant towerco after the majority of towers in a given market had been transferred.

When seeking to extend coverage or improve capacity in a given location, MNOs will always check whether there is an existing tower nearby that they can share first. Failing that, a new tower may be “built to suit”. Co-location on an existing tower can take as little as a month, a tower built to suit may take around 120 days, depending how long it takes to secure a permit. Build to suit (BTS) is still quicker to market than waiting for towercos to acquire and integrate towers.

Airtel have retained flexibility in their BTS strategy through the divestiture of their African towers, ensuring that if the prevailing lease rate doesn’t match their requirements, they can build their own or ask other companies to build towers for them. Despite the sale of towers, strong network planning and project management capabilities still exist within Airtel – those competencies will not be fully transferred to their towerco partners – Airtel’s tower operating company still exists and will be responsible for SLA management and governance. The reality is that today most of the rollout work for MNOs in SSA is done by third parties – turnkey contractors – so while MNOs that divest passive infrastructure to towercos may end up with a few stray BTS towers which towercos refuse to build, the ecosystem is big enough that an alternate service provider can readily be secured.

There were mixed views on active infrastructure sharing both within the panel and within the represented MNOs’ board rooms! Whilst in Europe infrastructure sharing has included both passive and active infrastructure, opinion was divided about the timeline and benefits of active infrastructure sharing in SSA. While some were wary of the potential implications active infrastructure sharing may have in terms of restricting future consolidation, others felt active infrastructure sharing was critical to the implementation of new technologies.

Master Lease Agreements between anchor tenant and towerco are increasingly drafted with multiple future scenarios in mind. The implications of shared antenna for lease pricing may be covered alongside clauses defining the specific space available made available on a tower and the kWh of power included in the service. So if the MNO tenant increases capacity, whether space or power, there are provisions built into the contract but it still provides enough flexibility for the opco to manage their capacity.

Several members of the panel called on towercos to consider providing shared solutions for the last mile. One panelist suggested towercos would be unlikely to diversify beyond “hard infrastructure” (fibre, backhaul, IBS and DAS) unless the model changes from a terrestrial to a completely different model.

In most SSA markets, coverage is no longer a differentiator. QoS, products and services and the Customer Experience differentiate MNOs – the customer doesn’t care whose tower they get their signal from, they just care about tariffs, they care that the call doesn’t drop, and they’re starting to care about mobile broadband services. If the era of parallel infrastructure in SSA is coming to an end, 2015 has demonstrated that there is first mover advantage in bringing towers to market before the competition, securing the best pricing and preferred terms as anchor tenant.