Network sharing is one of those topics that makes economic sense, but often falters when it comes to adoption and implementation. Recently, I was asked about the prospects for network sharing given 5G capital investment requirements at a time of economic crisis. This is, of course, related to active sharing models where operators share the radio access network and spectrum. From financial perspective, service providers should be inclined to adopt some form of network sharing to reduce cost and improve their financial performance. Regulators who approve network sharing agreements are eager to encourage 5G adoption and reduce barriers to deployments. But in practice, it’s not that straight forward.
Network sharing has multiple complex dimensions: what is logical may not be practical. We have been in the same situation during the 2008 financial crisis when LTE was about 1 year from commercial deployments. This is roughly analogous to the situation today when we are still in the early phase of 5G deployments. I believe the 2008 financial crisis did help bring about active network sharing in particular, but it was limited to a few markets. Rather, the dynamics in each market plays the dominant role in determining the type and prospects for network sharing.
Basis of Competition Inhibits Sharing
Despite the promise of 5G enabling services and applications to consumers and enterprises, operators compete on the basis of coverage and speed of connectivity. Nowhere is this more evident in than in the US with ads extolling the performance of a certain service provider [I find billboard ads most interesting!]. The US operators are incidentally the most hostile to network sharing. This would be a natural position for operators except when external force – economic or regulatory – demand otherwise. Competing on coverage and speed is appealing because of it simple to understand and measurable. Competing on services is more challenging and harder to measure, not to mention operators have failed at providing meaningful services above connectivity.
The Economic Factor for Sharing
Financial competitiveness is the single largest driver for network sharing. Europe provides a good example where 4 service providers compete in relatively small markets. Scandinavian countries which are large and have small populations were first to share networks going back to 3G deployments in 2001 in Sweden where Telenor and 3 setup a joint venture mirrored by another between TeliaSonera and Tele2. Many other active shared networks followed in Denmark, Finland, Denmark, Hungary, Poland, the UK, and France (excludes urban areas).
Economics will play a role in 5G shared networks but I expect it will remain regionally selective. China Telecom and China Unicom are sharing a network, but in Korea, despite regulatory encouragement for shared networks, operators are deploying independently. In the US, operators have a high debt load – like AT&T at over $150 billion. They are not inclined to share networks for competitive reasons. European operators are shedding assets and retrenching their operations: one would expect network sharing to continue despite past challenges in execution such as the case between Telefonica and Vodafone in the UK. Operators in emerging markets in Asia, Africa and Latin America are relatively small and dependent on foreign capital: they could benefit from sharing.
The Regulatory Factor
Regulators look to enable competition are hesitant to mandate network sharing. Service providers on the other hand use the financial argument to forestall new entrants and competition. Basically, the cost argument allows service providers to share networks, cut costs and prevent competition from coming in! An example of this is Canada, home to a shared RAN network between Bell and Telus. Together with Rogers, all three incumbent operators argue against opening the market to MVNOs which the regulator is considering as a way to stimulate competition. History showed operators will not enter into MVNO relationships unless mandated by regulators.
Most regulators worldwide mandate roaming services among operators as a way to equalize coverage which is particularly important for new entrants. Mandatory roaming is often considered a form of network sharing. Conversely, in markets that allow network sharing, regulators put measures to enable other service providers including MVNOs, as is the case in many European countries.
Potential Applications of Network Sharing
There are three areas where network sharing makes much sense, and to date relatively little network sharing has happened.
The first is rural area deployments where low subscriber density makes it uneconomical for any single service provider. In Turkey, for example, Vodafone, Turkcell and Turk Telecom share a network in rural areas with population under 500 inhabitants. A recent example is the UK’s Shared Rural Network between Vodafone, O2, EE and Three UK to meet new 4G coverage obligations with £500 million matching government funding.
The second are the nascent enterprise and neutral host models which leverage CBRS and other enterprise spectrum bands. These markets have their own subtleties that differentiate them from the common definition of network sharing. Nevertheless, they are not areas that one could ignore.
The third area is core sharing. While neutral host provides RAN-sharing, a cloud-hosted core could be shared among multiple enterprise users. Following the acquisition of Affirmed Networks, Microsoft could provide such a model; several other players are contemplating similar initiatives.
Business Models for Network Sharing
There are different ways to share networks among operators. Each business model has its pros and cons, so it needs to best match the objectives of the service providers. Sharing a new 5G network roll out is easier than sharing an established 4G network or a legacy 3G network planned for sunset. Similarly, sharing networks in rural areas is different from that in urban areas. Hence, selecting the right business model is critical to ensure efficiency of operation and avoid disputes. A few examples include:
- Partnership model: One of the operators takes the lead in managing a geographic area. The lead operator has responsibility for operation and support. Operators can work out offsets as needed. Network deployments and services can be fast using this model.
- Joint venture model: A joint venture acts as an operating company with each operator taking an ownership stake in the company. The JV will be responsible for coordination, procurement, planning, design, operational and management activities. Regulators typically have requirements on the extent of assets deposited in the JV.
- Wholesale model: Typically happens when a third party plans to sell bandwidth to existing operators. I am not aware of a successful model, while a few unsuccessful models come to mind such as Red Compartida in Mexico, Yota in Russia and a failed attempt in Australia.
Balancing Costs and Benefits
Active network sharing projects savings between 33% – 45% in capital expenses and 25% – 33% in operational expenses. Additionally, sharing can provide higher throughput and wider coverage in shorter time.
On the downside, sharing could reduce service differentiation and innovation. It could negatively impact competition if not handled properly. Sharing could require longer period of planning and execution, and make reaction to technical failures harder.
Form this perspective, the business model needs to balance costs while optimizing operators’ cost synergies. This is especially challenging when more than 2 operators are involved. The presence of a 3rd operator reduces efficiency and increases complexity.
From a regulatory perspective, the business model needs to address concerns related to collusion, effect on spectrum rewards (e.g. circumvention of caps and collusion in spectrum awards), and reduction of competition on parameters such as coverage or deployment of new technologies.
Final Thoughts
Network sharing evolved from the days when it referred to tower and antenna sharing to include many different possibilities. One of the most interesting development in my opinion is the potential to share the core infrastructure to enable multiple enterprises. This activity stand at the intersect of telecom and cloud technologies. While this does not follow the traditional concept and objective for network sharing among service providers, it creates the potential for new services which is exciting.