RANsharing is always a controversial topic at TowerXchange Meetups. The RANsharing roundtable at the TowerXchange Meetup Europe 2016 was tremendously useful because MNOs and towercos both put their cards on the table, the discussion got a little tense, but ultimately all parties left better understanding a little more about what was at stake for one another. We summarise the debate here…
Introduction
If RANsharing is an exciting potential source of efficiencies for MNOs, it’s an equally daunting dampener of revenue potential for towercos. With so much at stake, it is perhaps unsurprising that negotiations between MNOs and towercos can get a little heated where RANsharing is concerned. Provision for the treatment of RANsharing must be negotiated into towercos’ Master Lease Agreements, so what is the magic number – how should RANsharing be priced, if at all?
RANsharing 101
There are already several permutations of RANsharing, with the potential for still more to evolve in the future. The foremost of these are:
- MORAN (Multi Operator RAN where antennae are shared but not spectrum)
- MOCN (Multi Operator Core Network, in which both antennae and spectrum are shared)
- GWCN (Gateway Core Network, where both RAN and core network are shared)
With so many brands of RANsharing, it is difficult for a towerco to know what it is seeking to protect itself against.
RANsharing joint ventures in Europe
Europe has one of the world’s least mature tower industries in part because instead of passive infrastructure sharing, in many instances Europe’s MNOs dived straight into active infrastructure sharing joint ventures. The continent is home to the world’s most mature RANsharing ecosystems, with several deep RANsharing partnerships in prominent markets. These RANsharing partnerships have generally been successful for MNOs, driven by a principle that if you’re an MNO with your own estate into which you invest €100mn you get €100mn back, whereas if sharing the same network costs two parties €50mn each (or you still put in €100mn each and rollout faster). The benefits are obvious before one even considers the savings from sharing opex.
In order for independent towercos to take root in markets where RANsharing is taking place, or may take place in the future, towercos must adapt their contracts, business models and expectations. When the independent towerco business model took root in the U.S., India and even more recently in SSA, RANsharing was not prevalent. The balance of power is different in Europe, and the contractual provisions concerning RANsharing can be critical to ensuring a win-win outcome for MNOs and towercos.
Let’s compare three examples; from a market where RANsharing is widespread and where towercos are active; another market with a RANsharing joint venture (JV) where towercos may enter in the medium term; and a market where there is currently little or no RANsharing but where several MNOs are seeking to monetise their towers.
Figure one: RANsharing models as identified by Analysys Mason
RANsharing and towercos co-existing: the UK example
The management and operation of UK telecom networks is unlike anywhere else in the world, largely thanks to two deep infrastructure sharing joint ventures, CTIL (Vodafone+O2) and MBNL (EE + Three) which between them operate over half of the country’s ~50,000 towers.
CTIL owns and operates the networks of Vodafone and O2 (Telefonica) in the UK. Structured along an East-West divide, CTIL is programme managing Beacon, Vodafone and O2’s joint 4G rollout, underpinned by RANsharing.
Another JV, MBNL is one of the deepest integrated RANshares in the world, wherein EE and Three share sites, power and security under a mixed MORAN business model – sharing everything except spectrum. MBNL manages RAN for EE’s 2G and 4G network, and for Three’s 4G, and also runs a MORAN equivalent model for transmission. The JV was envisaged in 2007 when the strength of market leaders Vodafone and Cellnet (now O2) forced Three and T-Mobile (now EE, and being acquired by BT) to consolidate assets, decommission parallel infrastructure, and reinvest savings and rollout to achieve 99%+ coverage. An estimated ~£1bn has been saved over each of the initial seven years of the joint venture.
“RANsharing looks scary,” said one UK towerco, “but the worst thing a towerco can do is fight its customers. We’ve got to get in front of these changes and consider how to create value for our customers in ways which also creates value for your towerco. For example, look at some of the non-core assets in shared networks; perhaps the RANsharing joint venture identifies parallel infrastructure they no longer need but on which there are still third party tenants – perhaps a towerco can acquire or manage such sites?”
“We restructured relationships with the UK’s JV RANsharing companies, and were able to protect our business,” continued the towerco. “Remember that network sharing isn’t about reducing footprint, it’s ultimately about increasing points of presence (PoPs). Through the creation of MBNL, for example, EE would have increased from around 8,000 to 18-19,000 PoPs.”
How do MNOs compete when RANsharing?
Competition is through a raft of factors beyond the network: brand, handsets, price, channels, distribution and the customer experience all create a brand halo, and those are todays competitive differentators. However RANsharing doesn’t prevent competition at the network level: if one MNO has more spectrum they still have more capacity. Sharing partner MNOs can still deploy at totally different rates.
RANsharing in a market towercos may enter soon: the Greek example
Economic turbulence has created a scenario where at least one Greek MNO may monetise their towers in the next 12-24 months. Towercos interested in this 12,000 tower market will have to decode the implications of the current RANsharing agreement in the country between Vodafone Greece and Wind Hellas, operated Victus Networks, which manages around 7,000 of the country’s towers.
Victus Networks is a 50/50 JV formed in 2014 by Vodafone and Wind to help them challenge the dominant market share and larger spectrum holdings of Cosmote, 100% owned by OTE, itself now majority owned by Deutsche Telekom. Victus Networks manages the Radio Access and Transmission Networks of its parent companies and, in parallel, is implementing a partial active radio network sharing (MORAN) for 2G and 3G technologies in rural and selected urban areas of Greece. Victus networks claims to have delivered ~€100mn in savings across capex and opex.
“If the network is a factory, our objective is to reduce unit production cost and enable reinvestment,” said one of the stakeholders in the Greek RANsharing venture.
When inaugurating a RANsharing JV, MNOs need to consider a “checklist of soft issues,” continued the same Greek stakeholder. “The corporate matchup (people and management) is important – in this instance it was easier for two challengers to agree terms. It requires a lot of management effort to make these partnerships happen, and the first year is tough – you need commitment, focus and drive, but we’ve ultimately found it to be a fruitful journey.”
Preparing for RANsharing in a market where none exists, and where MNOs are seeking to monetise their towers: an Asian example
There is currently little or no RANsharing between the MNOs in this next example country, but negotiations are ongoing for the sale of towers by multiple MNOs.
The MNOs’ proposed terms and conditions leave the door wide open for RANsharing, while the interested towercos would prefer to see RANsharing prohibited, or at least priced in the MLA. The issue has become contentious as it is potentially highly value destructive for the towerco at a time when the MNO is seeking to realise a valuation equivalent of 8-12x EBITDA when selling their towers.
“I need to achieve a tenancy ratio of 1.7 to build an investible business case to acquire the towers. Given that if two parties engage in RANsharing it could eliminate 60% of potential tenancy ratio growth, I need to negotiate some form of contractual protection against RANsharing,” said one of the towercos bidding for the assets.
Mitigating the risk of RANsharing by extending the towerco business model
“If you cannot fight against RANsharing you have to lead it,” suggested one towerco. “Vendors forecast network densification may require 10x as many base stations as we have today, so perhaps it is incumbent upon towercos to build sites designed for RANsharing, and to build value for MNOs. We’re building 1,000 light poles per year, on which we own the base stations. Outdoor DAS is a similar opportunity.”
Another towerco suggested: “we’re talking about telecom infrastructure sharing at a country level rather than speaking to MNOs. When smaller countries prove the model of sharing networks from the outset, that will change the model.”
“There’s only one power distribution network, and one rail network, in many countries,” suggested an MNO. “Deep network sharing is a natural extension of the dynamics of infrastructure consolidation.”
A UK towerco disagreed: “The UK regulator clearly wants to see infrastructure competition. I feel we’re decades away from a single network. We still need differentiation on quality of network to incentivise innovation and investment – we’re a long way away from telecoms networks becoming a utility.”
The balance of power depends largely on how embedded active infrastructure sharing is within the network at the time a towerco enters
These three examples illustrate a simple truth: if RANsharing already exists, towercos have to live with it. If RANsharing doesn’t exist, MNOs need to be mindful of the potential of RANsharing to lower the glass ceiling on tenancy ratios, so they need to manage their valuation expectations accordingly – or afford the towerco some degree of contractual protection against the potential future impact of RANsharing.
Contrasting perspectives on RANsharing
Contractual protection: what price RANsharing?
It’s a simple question with an important answer: how many tenancies does a towerco lose when their clients agree to share RAN?
If a towerco is well protected by their contracts, they may lose very few tenancies. “When two of our clients agreed to 2G, 3G and 4G RANsharing in rural areas only, representing around 60% of territory, we were barely affected,” said one towerco. “However, another towerco in the same country lost almost 10% of their tenancies.”
“There’s virtually no active infrastructure sharing in Africa,” said another towerco. “But the biggest argument we have with our clients today still concerns RANsharing because we sign 10-15 year agreements. All our agreements have active sharing provisions, and we generally agree an extra fee to share RAN, but it’s difficult to come up with one economic model to cover all the different RANsharing models.”
All our agreements have active sharing provisions, and we generally agree an extra fee to share RAN, but it’s difficult to come up with one economic model to cover all the different RANsharing models
An MNO countered: “but it is better to share our networks – there are efficiency gains through sharing costs – this is the main argument you use to make us partner with towercos: the more you share, the more you gain. There are economics if you can share passive infrastructure, even more if you can share active equipment or MORAN.”
“MNOs are generating economics from RANsharing, the towerco needs to have some too – they are our sites after all!” Said another towerco. “We might charge €10,000 per year without RANsharing, or €16,000 with RANsharing. Alternatively RANsharing might simply be forbidden to force a discussion it if and when it appears.”
Another towerco took a different stance: “It can’t always be just about the tenant paying more money – the tariff might be defined by additional capacity, additional radio units, additional ground space, more activity on site. It’s not like nothing’s changing onsite; there is more infrastructure on site for RANsharing.”
“It seems some towercos are seeking a mutually beneficial, long term partnership, while others are talking about protecting their business in the short term by forbidding RANsharing or charging incremental fees,” said one MNO representative. “You can’t stand in the way of efficiencies; you need to diversify into small cells and FTTT – seek new sources of value.”
“MNOs need RANsharing to protect against escalating lease prices and the associated risk to EBITDA,” added another MNO, as the discussion got a little tense!
“We have not intention of forbidding RANsharing; we’re advocating creating clauses in the contract which create a strong position from which to negotiate,” interjected a towerco. “In some cases we won’t increase prices where the RAN is shared, we might get a BTS contract instead, for example.”
“When we are buying towers from an MNO, their valuation is derived from the anchor lease fee plus a premium for the potential incremental value from additional tenancies,” commented another towerco. “If the book value versus decommissioning is €40,000 and the MNO is selling for over €100,000 per tower, it’s not reasonable for the seller to then get in bed with their competitor, agree to share RAN, and take away 60-80% of that potential lease up premium. Most anchor tenants don’t pay enough for the towerco to breakeven – the value is in the lease up. There has to be a magic number for the value and price of RANsharing where it works for both sides. If MNOs insist on sharing RAN, they need to pay the lost lease up potential, or at least respect that their towerco partners need to retain the right to negotiate that if it happens. If towercos’ capital providers see too much risk of RANsharing, the contract becomes unbankable and MNOs will lose access to towercos as a source of capital.”
“We’re seeing the difference between RANsharing in a mature tower market when the towers were transferred off MNO balance sheets years ago, and a market when you’re at the point of handing over the cheque,” said another towerco. “The UK, for example, is a heavily network shared market – we know the ‘crossed arms’ position (several towercos and MNOs round the table were sat with crossed arms at this point!), and we’ve changed the language we use completely. That was the only choice we had, if we were to believe in a bigger long term story.”
Conclusion
One or more permutations of RANsharing are inevitable in most tower markets over a long enough time horizon. The question isn’t whether RANsharing will happen, the question is what are the risks and what are the fair protections a towerco can ask for?
If the anchor tenant was expected to pay so much that all the return on capital invested was recovered, then the lease rate would be unpalatable, so MNOs have to accept that some degree of lease up is necessary, and therefore some degree of control over network sharing must be agreed. Towercos don’t necessarily want to enforce premium lease prices for RANsharing, and they don’t want to forbid it entirely, but they do need contractual provision to guarantee their place at the negotiating table. If not, the business model ceases to be investible and we all go home.
