How to regulate the telecom tower industry

The IFC and TowerXchange joined forces to host key regulatory stakeholders from four Asian countries, plus industry leaders including TAIPA, American Tower, edotco and PwC, at our inaugural Telecom Tower Regulatory conference, held after the TowerXchange Meetup Asia in December 2016. Discussion points and recommendations from this think tank are summarised in this article.

1. Enabling towercos to acquire MNOs’ towers and to build for MNOs releases capital to be reinvested in spectrum, network extension and enhancement

2. FDI should be encouraged to bring capital, expertise and innovation

3. Taxation of towercos must be fair but, above all, predictable

4. Regulators can play a critical role in accelerating permitting and rollout, the optimal strategy being a time bound regime such as the ‘shot clock’

5. Clarify towercos’ rights of way to accelerate deployment of fibre and infill sites, particularly small cells for Smart Cities and 5G

6. Infrastructure sharing should be encouraged not mandated

7. Regulators may give themselves rights to intervene in dispute resolution, but otherwise should let the market set tower lease rates

8. Where possible, allow lease rate escalators to be linked to inflation

9. Some stakeholders prefer a ‘light touch’ registration rather than licensing regime for towercos

10. A fairly priced licensing regime can qualify the financial and technical capabilities of towercos while providing surety to investors

11. Inconsistencies in towerco licensing are an inhibitor to the development of pan-Asian towercos

12. Towerco licensing should be futureproofed against future involvement in small cells and other alternate site typologies and business models

13. Multiple towercos should be licensed to facilitate the creation of healthy, competitive markets

14. Regulators and government can play a crucial role in promoting EMF awareness and combating the NIMBY mentality

15. Consider allowing towercos and MNOs to utilise State owned structures, potentially creating a valuable new revenue stream

16. Encourage private sector involvement in fibre infracos

17. USFs should be placed under an independent regulator, with the participation of private sector representatives                     

18. Consider allowing towercos to access USFs, and consider allowing USFs to subsidise opex in certain scenarios

19. Consider reducing USF revenues levies as revenues increase, for example from 2% down to 0.5%

20. USFs that don’t disperse 50-75% of funds should stop collecting

TowerXchange opened the inaugural Telecom Tower Regulatory conference by contending that the era of unsightly, inefficient parallel infrastructure was coming to an end. With 68% of the world’s towers and rooftops (hereafter referred to as simply “towers”) now owned or operated by towercos, driving infrastructure sharing, it had become imperative for regulators to develop regulatory regimes to facilitate the investment of towercos, accelerating mobile broadband agendas. Monetising their towers by selling to towercos also enables MNOs to release capital to be reinvested in spectrum, network extension and network modernisation.

While 78% of Asia’s towers are owned by towercos, that statistic is somewhat distorted by the strong positions of towercos in China, India, Indonesia and Myanmar, where towercos own 100%, 68%, 65% and 64% of the country’s towers respectively. In the rest of Southern and Southeast Asia, the percentage of towers owned by towercos is still in the mid thirties, and many countries remain untouched by towercos, with the landscape compromised by duplicate tower networks. The tower industry remains less than three years old in much of Southern and Southeast Asia, with business models still evolving, and regulatory and taxation policies still being drafted.

TowerXchange made a case for a light touch in regulation: towercos manage only passive infrastructure, just ‘steel and grass’ (sometimes including power), and should be thought of as a lean, telecom real estate investor – although TowerXchange cautioned that towercos may also manage antennae, in a near term future of shared small cells. Nonetheless, towercos will never own spectrum nor direct subscriber relationships. Towercos manage infrastructure that is both critical to the nation, but not critical to national security – for example, while towercos can provide the infrastructure for public safety, security and surveillance networks, they won’t actually operate those networks.

Every telecom industry dialogue with regulators reminds participants that a 10% increase in mobile penetration can yield a 1.5% increase in GDP, and when one considers the transformational impact of the towerco-led mobile network rollout in Myanmar, where mobile penetration has increased from 10% to over 70% in the last three years, the potential impact of towercos becomes apparent.

However, every Asian country is struggling with a degree of infrastructure deficiency. Leading towerco SBA Communications uses two measures of infrastructure deficiency: the technical deficit – the deficiency of infrastructure, in this case towers, needed to unlock pent up demand and maximise mobile penetration in a county. And the economic deficiency – a measure of private and public telecom operators’ capacity to invest in infrastructure as a function of the revenues they collect, itself a function of the disposable income subscribers have to spend on telephony. MNOs are unlikely to be able to collect enough money from subscribers to bridge the gap between the economic deficiency and the technical deficiency of infrastructure, and this is where towercos can come in as a source of capital and an accelerator of mobile broadband initiatives.

Foreign direct investment in towers should be encouraged, not just to bring capital but also proven expertise and innovation. Towercos invest to improve uptime and QoS, and international investment brings with it lessons learned from regions such as sub-Saharan Africa, where MNOs have been impressed by towercos’ ability to improve energy efficiency and reduce downtime.

Lessons learned in mature tower markets such as the USA include:

- The need for fair, but above all predictable, taxation regimes at both federal and local level

- Accelerated permitting attracts investment and accelerates rollouts. For example, under the ‘shot clock’ regime, if a towerco (or MNO) submits a compliant application for a site permit and does not receive a response from a local authority within 30 days, the towerco can commence the build

- Clarifying rights of way will become even more important for 4G and eventually 5G, which requires the use of street furniture to deploy small cell networks – towercos can help with network densification

- Let the market set the lease price. The tower industry is naturally self-governing: if towercos over-charge, MNOs will revert to building their own sites

TowerXchange concluded by debating whether towercos should be licensed. While many tower industry stakeholders prefer that towercos be registered as opposed to formally licensed, exemplified by towercos’ registration as IP-1s in India, a fairly priced, swiftly executed licensing regime can serve to qualify the financial and technical capabilities of towercos, restricting the entrance of inexperienced developers who might not build infrastructure of a professional standard.

While the licensing question may be uncertain, TowerXchange were certain about their call for multiple towercos to be allowed in any countrty: the artificial restriction of a tower market to one or two license holders limits the number of prospective bidders for MNOs’ existing towers, suppressing their value to an extent that may be prohibitive to MNOs being able to get full value for their towers, while multiple licenses encourage a competitive tower market and can facilitate the development of regional or subject matter specific expertise, such as urban infill and in building solutions, or low cost rural sites.

Some of the more mature tower regulatory regimes have turned their attention to simplifying and accelerating the permitting of cell sites. A telecom tower often needs a dozen or more permits, often from multiple municipal, State, federal, aviation and environment agencies. At best this slows the deployment of towers, at worst it can encourage tower builders to build illegally.

---

“Regulators have often treated telecom towers like a building, needing only municipal permission,” commented one regulator. “When we had only one MNO, that approach worked. When the second and third MNOs entered our market, the incumbent would try to slow their competitors’ rollout. There was no incentive to share, and certainly not at an efficient price. So we saw a lot of overlapping sites built: we now have almost three times as many rooftop sites and towers as are needed to efficiently provide geographical coverage. When citizens see two or three parallel sites it provokes the Not In My Back Yard (NIMBY) mentality. So we stopped issuing permits! But as a result the MNOs simply rolled out without permits!”

“To rectify the situation, we established a new agency as a single point of contact to distribute documentation to all the authorities from which permits are required. Instead of going to multiple authorities, MNOs or towercos now come only to one, and we’ll be automating business processes. We will be recovering costs through fees, which will be clear up front.”

“In 2012 we did a feasibility study with a consultancy, based on which we encouraged infrastructure sharing joint ventures between the MNOs but they refused. Now each MNO may have to spend US$50mn to rectify the situation, which could represent an opportunity for towercos. We will not license towercos, because they manage only passive infrastructure, but they will register as commercial companies.”

---

One of the key influencers of permitting policy are EMF (Electro Magnetic Field) radiation guidelines. There is often a substantial gap between public concerns about EMF radiation from cell sites and the actual risks which, for example, resulted in an enormous number of Public Interest Litigations (PILs) in India, forcing the removal of a number of towers. With the support of the Ministry of Communications, and with a nationwide TV show promoting EMF awareness, India’s EMF programme has reduced NIMBY concerns, and judgements have found a lot less radiation hazards in the last two years. As a result, over a hundred thousand new base stations were added to Indian mobile networks in 2016 which has enabled a significant reduction in call drops.

In Indonesia the government regulates towercos with a light touch, only in terms of guidance on quality and safety. A cautionary tale comes from Bali, where new local regulation forced MNOs to tear down many sites, resulting in the Supreme Court having to suspend the regulation to enable service continuity, ultimately restricting the applicability of the new regulation only to new towers.

Most towercos don’t want to be licensed, contending that they are just like any other real estate business, and are already controlled by the permitting regime.

If a towerco license is to be introduced, it prompts the question: how are bidders to be evaluated?

One driver in favour of licensing towercos comes from efforts to close the Digital Divide. Given the acute sensitivity of the towerco business model to the potential for co-location, will unlicensed towercos be adequately motivated to extend networks into rural areas? “If you haven’t licensed towercos then it’s a dilemma to hold them accountable if they don’t meet your National Broadband plan,” said one regulator.

Regulators should approach the issue of licensing towercos as a distinct category from retail telcos as towercos have no subscribers and do not use finite spectrum. As such, the same magnitude of fees and revenue sharing may not be applicable as to MNOs.

One of the inhibitors which might explain why there are less multi-country towercos in Asia than in some other continents, is the lack of consistency of licensing regime and fees. Even among workshop participants, license fees, where existent, ranged from 0.5% to 2% of adjusted gross revenue.

Towerco licensing regimes in Asia vary from the mature, converged facilities and network services license in Malaysia, issued as long ago as 1998, to the new towerco licensing regime in Myanmar. While towercos in Myanmar would have liked their licenses issued sooner – they had to trade under a letter of non-objection for over a year, which made raising capital challenging – in the end Myanmar’s licensing regime was seen as fairly priced and a deterrent to less credible and well-capitalised prospective towerco startups who might have lowered standards. There are towerco licenses in Pakistan and Cambodia, but towercos are not yet licensed in Sri Lanka (although infrastructure sharing guidelines have been issued and are expected to be amended in 2017). India has eschewed licensing towercos in favour of the lighter touch IP-1 registration scheme, which still affords the government some control.

There is also no towerco licensing regime in Bangladesh. The BTRC recently consulted industry stakeholders for comment on a draft licensing regime, prompting MNOs and towercos to advocate a simple licensing regime that opens the market to the innovation, expertise and capital of foreign direct investors, while allowing licensed MNOs to partner in towerco ventures. 51% of the world’s towers are owned by towercos that are themselves majority owned by MNOs. Another sticking point in Bangladesh has been the number of towerco licenses to be issued, with proponents suggesting the more licensed, the more competitive and efficient the market.

Significant, long term investments are being made by towercos and their investors in telecoms infrastructure – licensing secures investors’ long term interests.

The introduction of licensing may not be bad news for towercos, but limiting the number of licenses can inhibit the monetisation of tower assets, and hinder the emergence of a competitive tower landscape.

---

Experiences from the mature Indian regulatory environment were shared by TR Dua, Director General of the Tower And Infrastructure Providers Association of India (TAIPA), whose members include Bharti Infratel, Indus Towers, American Tower, GTL Infrastructure, Tower Vision and Reliance Infratel (now Towercom). TAIPA advocated the creation of similar associations for infrastructure providers across Asia – and recommended leveraging the TowerXchange community to facilitate such initiatives. An infrastructure provider association maintains a dialogue with relevant Ministries including those for Telecom and Finance, with regulators and with local authorities.

Until 2008 India’s MNOs managed their own infrastructure, and mobile market growth was slow, inhibited by a reluctance to share towers. In 2006-7 the Indian Ministry of Communications encouraged tower sharing, together with the Ministry of Urban Development, as most the issues arose locally, through a Project MOST (Multi Operator Shared Towers). This government endorsed programme created a lot of excitement about independently owned infrastructure companies, including supporting the carve out of the Indus Towers joint venture.

Infrastructure providers, registered as IP-1’s in India, have been granted infrastructure status in the country, although it has yet to be fully implemented.

More recently, TAIPA has worked with government stakeholders to clarify IP-1’s Rights of Way and to accelerate permitting. Time bound clearance, differentiating overground and underground, is critical to accelerating time to market, as is a permitting regime with fair fees, clearly laid out.

Today more than two thirds of India’s towers are owned by IP-1s, with an average tenancy ratio around two. Tens of thousands of India’s towers are green sites, which means they consumer less than a litre of diesel per day. And towercos play a critical role in the realisation of the country’s national broadband plan, Digital India, and in the creation of Smart Cities. Dua concluded by echoing TAIPA Chairman Akhil Gupta’s vision for the ‘disarmament of MNOs’, calling for towercos to build all new sites and acquire legacy sites.

Tower industry stakeholders are unanimous in their call that regulators not get involved in lease pricing. Regulators may need to reserve the right to intervene in dispute resolution, but should let the market set the lease price.

However most tower leases will have some sort of escalation policy in place, and it is important that lease rate escalation be allowed to be linked to inflation. If escalators cannot be linked to inflation, international investors in mobile networks amplify their exposure to the performance of the economy and to country risk in general. More risk means less investment.

“Towercos are needed to support national broadband initiatives, especially in remote rural areas, to ensure delivery of high speed data,” appealed one regulator.

“There’s no lack of intent to support what the ITU calls second wave reform,” responded a towerco. “If regulators and governments need to license an independent towerco to meet these targets, then we will consider those opportunities, but we’re mainly driven by business imperatives. Broadband policy remains an MNO-centric issue, but towercos need to take more guidance from policy intent.”

TowerXchange note that while the majority of towercos worldwide have not yet diversified into fibre, there are interesting exceptions, such as Crown Castle selling towers in Australia ostensibly to re-invest capital in U.S. fibre, and IHS Towers securing a license to deploy fibre in Nigeria.

“Fibreisation of towers is already under way. If a tower is fibreised, then aesthetics can be improved – we don’t need big microwave dishes, height can be reduced,” contributed a towerco. “But we need the regulator to legislate in a market like Pakistan which is over fibreised in urban areas, but lacking fibre in rural areas where 70% of the community lives. Towercos can invest in fibre but rights of way and permissions are hard to secure.”

“Our National Telecom Policy 2012 laid out targets for Smart Cities and Digital India, generating a paradigm shift and a much bigger role to play for towercos,” added TAIPA’s TR Dua. “Internet connectivity and accessibility is one of the nine pillars of Digital India, but access, clearances, documentation for sites and rights of way remain critical. The rules under the Indian Telegraph Act now provide clarity, and compliance is mandatory at State level, giving 30 days to permit a new site. If the permit is not received, the towerco or MNO can take it as a deemed clearance. The regulator has also recommended, not mandated, provisioning of ducts to facilitate easy broadband service provision.”

The achievement of national broadband targets, motivated by the imperative to close the digital divide, can be inhibited in markets where only the leading MNO is making a profit. In such circumstances, the regulator might encourage network sharing in rural areas where the market leader is the only service provider and infrastructure owner, with no inclination to share. The #2, #3 and #4 operators could then build their own shared rural towers.

Our conference co-hosts concluded: “The IFC have been advocates of infrastructure sharing. And your countries are our shareholders! The tower companies we finance can help: towers have an important role to play in mobile broadband, which cannot be seen in isolation – everything is converging. If a 10% increase in mobile penetration can yield a 1.5% increase in GDP, then countries should promote open access, carrier neutral open broadband networks. This may involve persuading MNOs spinning off towers and fixed line operators spinning off the backbone, but beware the ecosystem ending up a captive network which won’t provide services to parties outside the consortium. While there is no one size fits all solution, we see towercos as part of infrastructure sharing solution to deliver the best possible broadband access for the population.”

China has 15-30,000,000km of fibre, Malaysia 500,000km, and Myanmar only 13,000km. Fibre rollout is imperative to the achievement of national broadband plans across Asia and worldwide. Initiatives like the nbn (National Broadband Network) in Australia have a mandate to build national broadband infrastructure. What is the role of towercos in building up infrastructure to meet growth targets and support national broadband plans?

Many countries are just embarking on fibre deployment, and don’t have enough to enable e-government or encourage investment. In some countries the fibre has been separated from the incumbent fixed line operator, like in the UK where Openreach deploys fibre and sells on equal grounds to all operators.

“Our preference is not to permit multiple fibre infrastructure providers,” suggested one regulator. “Other utilities requiring permissions to install infrastructure underground come from a single provider, like water and electricity, but telecom is multiple entities. 60-70% of network cost comes from the laying of fibre – ultimately the investment case falters if this infrastructure and cost is not shared.”

Mongolia has taken a different approach to the separation of the telecom service from telecom infrastructure. The country’s three licensed infracos have no subscribers, functioning solely to provide service to telcos, running everything from fibre and microwave backhaul, to passive and active infrastructure, as well as buildings. However, with the main backbone still government owned within ICNC, they have experienced financing challenges and haven’t been able to provide affordable broadband services as quickly as targeted.

“I would recommend an infraco not be a State owned enterprise,” suggested one of the moderators, “with a rethink toward a consortium model with private sector involvement.”

---

Working with independent towercos has been one of the many recent successful innovations embraced by Cambodia’s regulators, who have successfully fostered a rise in mobile penetration from the lowest in the ASEAN region to the second greatest in 15 years.

With Cambodia granting neutral technology licenses, the market is primed for mobile broadband, which means a change in the profile of new sites from tall green field towers to rooftops, lampposts, indoor and outdoor DAS.

From towers to international gateways, the regulator spoke in encouraging terms about the promotion of a culture of infrastructure sharing engaging telcos, cities, and Ministries to reduce the landscape pollution of towers and cable and provide high speed Internet.

Cambodia’s progressive regulator sees towers, fibre and gateways as intrinsically linked, and had gone so far as to provide tax incentives to attract international towerco investment.

---

“Regulators can focus too much on GBTs (ground based towers), they need to better understand the role of RTTs (rooftop towers), poles, lampposts and other alternate site typologies for infill capacity,” suggested one towerco.

It was also noted that as urban landscapes expand vertically, rooftop towers with blocked lines of sight might need relocating, placing pressure on land management and construction policy.

“Our focus must be on different site typologies for different situations and different markets,” suggested one regulator. “In the end it’s all about antennas – 4G antennas need to be less than 500m apart, that could come down to 200-300m for 5G, and if innovations like Connected Cars are to be a reality, we will need reliable, contiguous coverage.”

We are increasingly seeing towercos diversify into provision of small cells, even Wi-Fi hotspots – towerco licenses must allow them to install this layer of the network. “There are only around 32,000 hotspots in India at present,” suggested one participant. “Yet we may need 800,000 to provide public Wi-Fi. Towercos have an appetite to install the necessary infrastructure.”

“The role of regulator is to reduce disruption, so sharing is the natural way forward, and as we go into the small cell era, the typology of shared infrastructure becomes more diverse,” concluded Hardiman Telecommunications’ Paul Carpenter.

Regulators were encouraged to consider what other structures entities MNOs and towercos could be permitted to use, from the roofs of government, post office and municipal buildings, to power transmission and transport operators’ towers and cable networks, and utility company fibre networks. Some examples:

- Power Grid Corporation of India subsidiary Powertel leases up fibre and tower networks to MNOs

- Russian Towers leveraged a partnership with the Russian Railway to co-locate telecom tenants on their structures

- Towerco IIMT has an agreement with the Mexican Federal Electricity Commission to utilise its structures for telecom purposes – FPS Towers (now American Tower) recently reached a similar deal with the national electricity transmission company in France

- In Nigeria towerco BCTek promotes 700 towers within police compounds to MNOs for co-location

- In Ireland the State Forestry agency sold 113 towers and 400 plots of land to a towerco now known as Cignal – so these assets can have significant value!

While security is often a consideration, selected military structures and fibre may also be made available for sharing by MNOs.

Even if governments are reluctant to allow permanent structures on their property, temporary coverage solutions such as CoWs (Cell on Wheels) should be allowed to provide supplementary coverage, for example during major religious and civil gatherings.

How do we bring towers and a fibre backbone to underserved areas, and what should be the role of USFs?

A Universal Service Fund is typically generated by a levy on operators’ revenue, and is deployed to subsidise the network beyond the reach of economic coverage. Some USFs have been deployed, some are awaiting deployment, and some have been absorbed back into treasury funds.

The Indian government has incentivised IP-1s and MNOs alike with easy rights of way and zero permission fees to extend networks in underserved areas. And under India’s National Optical Fibre Network (NOFN), managed by Bharat Broadband Network (BBNL), a fund has now been made available to towercos for last mile fibre in the most far flung areas, especially in the Northeast of India.

However, it is not always easy for towercos to access funds. One towerco complained that the USF in one country was only accessible to licensed MNOs, and thus towercos were unable to access and deploy those funds. This prompted a debate about whether towercos should contribute to USFs.

If a towerco was to utilise a USF, what mechanism might need to be in place to be clear that funds be allocated where a service provider would actually take a tenancy? While the Universal Service Organisation may identify unmet demand, they must turn over responsibility to industry to determine the economics and feasibility of a site, before resuming engagement to visit the site, inspect the tower, and pay for it. TowerXchange contended that towercos were the ideal entity to determine the economic viability of marginal sites in terms of their potential to attract at least one, preferably two or more tenants, this evaluation being a core competency of a towerco.

“The towerco mindset must change in rural areas,” said one participant. “If they have access to the USF, they must accept they are sometimes being subsidised to run a single tenant tower.”

“If such sites were to work on our balance sheet over a ten plus year horizon,” contended a towerco, “an opex subsidy may be more helpful than capex.”

One consultant present offered a sobering comparison to the attractive idea that a towerco might partner with a USF to deploy capex, build sites, and provide service on an opex model. In one country in Sub-Saharan Africa, 11 rural sites were built for the USF at a staggering US$500,000 each, 100% solar, to meet the anticipated huge power requirement of four MNOs. The sites proved to be over specified. In another country on the same continent, three batches of 30 rural sites were built, one batch managed by each operator but with RAN shared, with a much more cost efficient site design.

One clear recommendation was that the USF be placed under the regulator rather than under than under a Ministry, as the regulator had more powers to enforce collection. It was also suggested that the USF board be chaired by a private sector representative to ensure appropriate independence from the Ministry of Communications and Finance. For example, the Minister of Finance may not always be able to release funds because of economic situations.

In Indonesia the USF is being deployed into the backbone, while in Bangladesh funds have been collected from MNOs but reportedly not significantly deployed. Cambodia has only recently started collecting a USF based on 2% gross revenue. Sri Lanka’s USF is being deployed in underserved areas. In Malaysia over the last six years three phases (Time 1, Time 2 and Time 3) of USF deployment have been undertaken to facilitate broadband rollout in underserved areas. All licensees who contribute could avail themselves of the Malaysian fund, with contracts awarded on the basis of bids for the lowest level of subsidisation.

One of the towercos summarised their recommendations as follows: “What are the reasons for regulatory intervention? Provision of a fair and simple licensing regime for infrastructure providers. Encourage, but don’t mandate, infrastructure sharing. And drive toward achievement of national mobile broadband targets.”

Conference hosts the IFC concluded: “as an investor and lender, the need for clarity is paramount. The regulatory and taxation regime must be predictable – it’s less about value as risk of change.”

Regulations need to be clear in defining the difference between passive and active infrastructure, in recognising the boundaries between wholesale and retail, yet still have appropriate flexibility to accommodate new models, such as towercos operating multi-operator small cells.

The moderator concluded: “With the increasing prevalence of independent towercos in Asia, it is timely for regulators to reassess infrastructure sharing policy and towerco licensing. Regulatory, government and industry stakeholders can learn a lot from each other!” We look forward to future regulatory conferences at TowerXchange!