A new network sourcing model can free up billions for CellCos
How the Radio-as-a-Service model can help achieve that
Disclaimer: All thoughts expressed below, unless quoted, are solely mine. I have co-authored two articles on RaaS (Radio-as-a-Service) in Detecon’s journal, which provide the main thesis on that topic. However, some thoughts had not made it to the final publication. Here, I would like to elaborate on my personal views and ideas that were left undiscussed in the publications.
A brief thesis outline:
- Telcos’ CAPEX has grown but revenues did not follow.
- 40–50% of Capex goes to RAN, but RAN is a poor ROA asset and better procured as a service (RaaS model)
- Industry landscape grows supportive for RaaS: TowerCos as the key provider
- RaaS unlocks 20–30% of legacy model costs — billions for value-creating activities
CellCos Capex does not generate more revenue
The telecom industry had been stalled for the last decade. It was a depressing decade for telecom shareholders too. I wrote about that in the “Boring Telco”-series extensively, but it’s best described by the quote below.
Between 2010 and 2018 industry revenues were down $27 billion, while telcos invested [emphasis mine] $250 billion in the network. Over 2019–2025, another $1.4 trillion would be spent by the industry, 70% of which would be on deploying 5G.
Behind that poor perform lies a fundamental industry dilemma, as we wrote:
The dilemma for mobile operators worldwide is to keep up with traffic growth while maintaining a healthy financial position. The size of mobile networks is a function of data traffic, and data traffic is growing by at least 40 percent a year. As a result, technological progress consumes enormous financial resources.
…
While the price per GB is falling, CellCo pay for this consumer surplus while the vendors benefit.
Exploring the root-cause of low ROIC is beyond this essay (see: “Boring Telco”-series again. Yet the question of CAPEX allocation is of long-term strategic value. If the underlying asset has unique qualities — spend would be somewhat justified. If not, there is a major business risk.
RAN is a poor ROA asset without a USP
The most common debate I faced was around “Network is a unique differentiator”. I argue that “some network parts are enabling unique differentiation”. Core network, for example, or service platforms. But not the Radio Network (or RAN).
Mobile networks used to be the key differentiator for carriers. That was a long time ago when coverage was patchy and the location of a mobile mast mattered. Hence why tower infrastructure is still somewhat of a differentiator. But guess what? Towers were the first assets that got sold, only to be leased back…
We explored more arguments that spoke for change in the legacy RAN model:
- There are 800+ mobile carriers worldwide but only 4 real RAN vendors. There is no differentiation
- Carriers have no say in the RAN product map — the vendor selection process is that a “like you, don’t like you” and price
- OpenRAN will not save the situation. Not because it’s not mature, but because CellCos are not willing to manage the growing number of supplier
- Carriers actually buying RAN as a service already — turnkey contracts prevail
- Vendors are in control and they have the bargaining power
(details are in the first article)
The unspoken truth is that RAN is not an asset, but a service already. Given the current -as-a-Service approach in many industries, we went on to explore RAN-as-a-Service model (RaaS).
At its core, it is an extended version of managed services. Some operators told us, that they had discussions of such a RAN-an as-a-service approach with vendors but failed. Vendors were not interested, because it was a risky drag for them.
The good news is, it seems their monopoly days are over.
Landscape grows more supportive for RaaS
We crystallized that the current RAN model is a dead-end cash drain. It’s obvious even to the vendors. The key question is if, apart from vendors, there would be other candidates ready to take risks and drag. Turns out that there are.
I will not recite the full discussion on who and why here — encourage you to read the article. In short, TowerCos are interested to extend to Active InfraCo as they need to back their valuations. See Cellnex in Poland — it is happening already.
The key argument for TowerCo is that it is a neutral player. There is a lower conflict of interest than with an OEM vendor. We conclude that a TowerCo with RAN assets — InfraCo can also solve the challenges for network sharing.
There is another thought left out. If TowerCo could be the facilitator for sharing and managing RAN, it could as well do so with OpenRAN equipment too. This could be a boost for this vendor class too. (Turns out that we are not the only ones assuming this option).
The take is this: RaaS model may be still emerging, but it’s commercial already.
I believe, that actually we will see more actions. The key driver for this model could be the narrative of “de-layering. That is, decoupling the network model from the service platform model´.
RaaS is the real enabler of “delayering” paradigm
The CAPEX-dilemma is long known to the industry. capital efficiency (ROIC) of in Telco industry is barely above the cost of capital. Meaning that there is no value creation happening. Compare that to Technology Sector.
The reason for low ROIC is exactly that network spend is huge but is not generating new revenues. That is because its core function remains unchanged: producing and selling capacity.
Since every CellCo has its own infrastructure, there is a limit to economies of scale. Hence there was always the idea of separating Network (NetCo) from Business (SalesCo) and streamlining the cost base at both. So far there are still too few cases, albeit those we see are successful (CETIN: 1+1=3 or what we at Detecon do for UNN in Brunei). The key driver for value creation is resource focus. NetCo focuses on the “engine”, while SalesCo add value through on-top services. This SalesCo eventually becomes a PlatformCo with higher margins, so the theory.
Separated entities both would generate better value:
- NetCo sets focus on lean operations
- PlatformCo (SalesCo) sets focus on value-creating service development
Clear focus eliminates overhead costs and budget prioritization speculations. It also alings organizational pace — two-tier speed models simply do not work outside of IT.
Yet, despite all benefits, delayering faces many reservations, foremost fear of breaking things. It is a radical move not for every company. And this is exactly why RaaS is the perfect transitional model:
- It creates the necessary impact, due to the high addressable cost base. The synergy potential is high
- It does not yet require to-draw a hard separation line
- Enables phased implementation rather than full-blown change
Benefits outweigh risks
One may argue that switch to -as-a-service model is a mere shift from Capex to Opex. I disagree, as there are transitory efficiencies.
My estimation is that RaaS model can reduce RAN Capex by 20% and even more when Network sharing is used — up to 35%. It will also reduce annual network Opex by at least 15–20%, through lower rent and power (if sharing is installed). Accounted are complexity reduction costs through lower overhead in management. And this is all accounting for the payments for services toward InfraCo. And not even going into the hubris of IFRS 16 and beyond — there are delta savings!
What are the drivers?
- RaaS neutralizes “brand dependence” . It should be irrelevant on which OEM equipment is used by InfraCo. , they should handle the procurement, commissioning, and modernization
- RaaS has lower conflict of interest in terms of upselling, than OEM. The pivot point is the cost of service
- Vendor model hinders Network Sharing implementation. Whereas InfraCo would aim exactly for that to lower cost base
- InfraCo has lower financing costs (WACC) which one would expect to pass onto the price (good for ROIC, as shown before)
Given that no other real consideration remains, it becomes a strategic decision to meet. The decision of whether running the current model is generating shareholder value. For most of the CellCos there (over 800 worldwide), that is not the case. There is a reason why the most successful of them in terms of scaling, Jio, is positioned a platform by its holding company Reliance.
The only risk I see are the trembles of leaving the comfort zone. Yet staying within is not serving the industry and shareholders any good. The current crusade of EU Telcos against BigTech set a wrong management focus. “Litigation instead of innovation” as someone wrote on social media.
Implementing Radio-as-a-Service model frees up billions for the industry to re-align. Telcos need to shift from fading value extraction to value creation. This will benefit both, society and the industry. Business-as-usual is not an option for most CellCos out there.
