Excluding Russia and the CIS, which we include in our European analysis, TowerXchange estimate that there are 2,205,708 telecom towers in Asia, 53% of which are in China and 20% in India. Towercos own 71% of Asia’s towers, but the picture is significantly distorted by the high towerco penetration in China (100%) and India (68%). In this analysis we look at the relatively untouched tower markets in Asia, finding some virgin towerco markets that might be a good fit for the independent towerco business model.
Afghanistan: There are five licensed 3G MNOs in Afghanistan; Etisalat, MTN, Roshan, Aftel and AWCC. Security obviously remains a key issue for local towerco Frontier Tower Solutions, who have built or manage around 1,200 towers for AWCC, while there were rumors that MTN and Etisalat approached IHS about operating a shared towerco in Afghanistan, but the deal fell outside IHS’s investors’ remit. There are over 20mn subscribers and over 90% population coverage in Afghanistan. The local MCIT shares annual tower count data as seen in figure two.
Figure two: Afghanistan tower counts, 2002-15
Bhutan: GSMA Intelligence report that Bhutan has 676,400 SIMs among a population of 779,500, which still represents formidable growth for a market only liberalised in 2008. The mobile market is a duopoly between market leaders B-Mobile (Bhutan Telecom) and TashiCell, which launched in 2008. A 3G rollout has been ongoing nationwide since 2008, with LTE being deployed in Thimpu by B-Mobile. Infrastructure sharing has been mandated by the Bhutan InfoComm and Media Authority (BICMA) since September 2015, although some limited tower sharing had taken place prior to the regulator getting involved.
Brunei: Another very small market with a duopoly, this time dominated by DST with Progresif Cellular having acquired B-Mobile in 2014. There were 521,000 connections and a SIM penetration rate of 122% at the end of 2015 according to GSMA Intelligence. The Brunei Times quoted Brunei’s Minister of Communications stating that there were 235 network towers in the country in March 2015, 151 belonging to DST and 84 belonging to Progresif Cellular, which is believed to be building relatively aggressively, hence our current estimate of 250 towers in Brunei.
Japan: Japan is one of the most sophisticated mobile markets in the world. Yet towers are still seen as a source of competitive differentiation, which perhaps explains why initial interest in carving out a towerco a few years ago seems to have tailed off, and why tower count data is so hard to find – readers should consider our estimate a very rough guide. Japan is famous for having the fewest number of subscribers per tower in the world – reportedly around 500 – suggesting a staggering tower count of around 220,000 for a nation of 127mn people and a landmass of just 378,000 sq km. LTE was launched as long ago as 2011 by former State owned monopoly NTT DOCOMO and in 2012 by the nation’s other MNOs, SoftBank and KDDI (au). DOCOMO has already started rolling out LTE-A. Japan’s three leading MNOs are believed to have each added up to 30,000 microcells and small cells as infill sites. TowerXchange understand several tower companies are trying to establish themselves in the Japanese market, but that their penetration remains negligible.
Figure one: Estimated tower counts, Asian tower markets where towercos are less active
Laos: According to the World Bank, Laos has the lowest population density in the ASEAN region (29 per sq km) and a notably mountainous terrain, which presents obvious challenges for mobile network planning. Market leader Unitel, backed by Viettel, claims to have 3,100 2G and 3G base stations, but other clues to Laos’ tower count are few and far between. The country is home to four MNOs (Unitel, ETL, Lao Telecom’s M-Phone, and Beeline, from which Vimpelcom are seeking to exit) who between them have achieved 75% population coverage and a little over 5mn subscribers among a population of 6.8mn. 3G coverage is fairly widespread, 4G largely limited to capital city Vientiane.
Mongolia: Infamously the most sparsely populated country in the world, yet almost half of Mongolia’s 2.9mn citizens live in Ulaanbaatar, where the network is sufficiently developed that small cells were installed as long ago as 2013. ARPU is around US$6. Outside of the capital, Mongolia is home to a dispersed in some cases nomadic population with typical rural coverage challenges. Telecoms were liberalised in the mid nineties and the country is now home to a competitive mobile market led by Mobicom (39%) and Unitel (36%), followed by Skytel (14%) and G-Mobile (11%) with 34% of subscribers on smartphones and 26% of subscribers on 3G (statistics courtesy of the Communications Regulatory Commission of Mongolia, 2014). 3.5G is available in most provincial capitals, and LTE spectrum auctions are expected within a year. With 4.5mn subscribers, subscriber penetration is around 150%.
Nepal: Ncell and Nepal Telecom have 46% market share each (NTA, August 2015) in a five MNO market with a well funded sixth MNO launch by the Chaudhary Group imminent. TeliaSonera has reached agreement to sell their majority stake in Ncell to Axiata, which may herald the entry of Axiata’s fast-growing towerco edotco into Nepal within the next year. SIM penetration is 93% in this 28.7mn population country (statistics again courtesy of GSMA Intelligence, from Q4 2015). TowerXchange estimate there are around 6,000 towers in Nepal. The GSMA’s Mobile for Development study for Ncell (http://www.gsma.com/mobilefordevelopment/wp-content/uploads/2015/10/NCell-report-October-2015.pdf) revealed 90% population coverage but only ~40% geographical coverage in Nepal. The same report called attention to Nepal’s unreliable grid, with nearly a third of the country’s 5.4mn households off grid. Of Ncell’s 2,378 towers, 156 were off-grid at the time of the report (September 2015).
New Zealand: We’ve heard two credible reports of prospective towerco activity in New Zealand, so don’t expect this ~4,000 tower, 6mn subscriber market to be dormant for long. 2degrees (Trilogy) has around 1,000 of those towers, Vodafone New Zealand and Spark about half each of the remainder. LTE is offered by all three MNOs. New Zealand’s US$175mn Rural Broadband Initiative (RBI) has accelerated investment in towers, targeting coverage in all communities with greater than 300 inhabitants. Under the RBI Vodafone had added 104 new towers and upgraded 305 others by Q2 2015.
North Korea: Mobile telephony remains limited to affluent members of society, with penetration at around 11%. Orascom Holdings has a 75% stake in Koryolink, which launched in 2008 and which seemed to be the sole MNO in in North Korea until the recent licensing of ISP Byol, with which Orascom is attempting to merge. The complexity of that proposed transaction recently prompted Orascom to deconsolidate Koryolink, making it an associate rather than a subsidiary. Telegeography suggested in 2014 that Koryolink’s 3G network had 90% population coverage, but mid-teens geographical coverage. There remains almost no prospect of international tower company investment given sanctions and the near impossibility of repatriating funds. TowerXchange estimate there may be just 600 towers in North Korea.
Papua New Guinea: Mobile telephony is a luxury almost two thirds of PNG citizens cannot access or afford. SIM penetration is still around 50% in a population of 7.6mn. Digicel competes with state owned operators bmobile and Telikom PNG, who shared a 3G/4G rollout. NICTA, the local regulator, is considering mandating infrastructure sharing. PNG’s Universal Access Service is active, with a project to build 59 towers in remote locations contracted to Digicel and completed in Q2 2015. Market leader Digicel has around 800 towers.
Philippines: The world’s eighth biggest mobile market, with around 100mn subscribers, the Philippines became a virtual duopoly in 2011 with PLDT’s acquisition of Digitel, merging it with their Smart mobile operation which then had around 13,000 base stations. A third MNO may soon enter the market: Aussie market leader Telstra in a JV with San Miguel Corp’s (SMC) Vega Telecom, which has been rolling up telecom assets in the Philippines. The opportunity to accelerate time to market for a new entrant would amplify the attractiveness of the Philippines to build-to-suit-centric towercos, but family ownership and fierce competition between incumbent MNOs means their towers are unlikely to come to market.
A third MNO may soon enter the market: Aussie market leader Telstra in a JV with San Miguel Corp’s (SMC) Vega Telecom, which has been rolling up telecom assets in the Philippines. The opportunity to accelerate time to market for a new entrant would amplify the attractiveness of the Philippines to towercos
We haven’t found any hard data on PLDT / Smart’s tower count, but bearing in mind that the entity is the market leader by a significant margin, combines two networks, has achieved almost 90% 4G coverage with LTE-A being selectively overlaid since 2014, we conservatively estimate Smart’s tower count at 11,000. Globe are more transparent about tower counts, stating that they had 6,227 cell sites in their most recent annual report, with a reported 4,300 new cell sites currently being deployed for LTE. The currently dormant Express Telecoms and Liberty Telecoms’ Wi-Tribe Asia, both controlled by SMC, may account for several hundred more towers, indeed Liberty Telecoms themselves had considered leasing out their 500 base stations in metro Manila.
Singapore: Singapore’s macro tower network is supplemented by a substantial IBS and small cell deployments to boost LTE network capacity. According to the iDA there were 8,166,100 mobile subscriptions in the city-State in August 2015, representing 149.3% penetration. Outdoor 4G coverage is approaching or above 99%, while all three of Singapore’s existing MNOs – SingTel, M1 and StarHub – are making rapid progress upgrading to LTE-A. There is stiff competition for Singapore’s fourth MNO license, expected to be awarded in Q1 2016: ISP MyRepublic is trialling 4G and promising unlimited data plans, while facing competition from Circles Asia and OMGtel. A public sector network could also provide opportunity to add tenancies to Singapore’s existing towers, although TowerXchange are not aware of any commercial towercos operating in the city-State.
South Korea: According to GSMA Intelligence, SIM penetration was at 113% among a population of 50.4mn in Q4 2015. South Korea boasts one of the most sophisticated telecommunications infrastructures in the world, cultivating an insatiable demand for high speed mobile broadband among its citizens. Mobile broadband penetration in South Korea is above 99% and fibre has been widely deployed. South Korea is a three operator market featuring SK Telecom, KT and LG Uplus, while the Ministry of Science, ICT and Future Planning (MSIP) has announced intention to license a fourth MNO. South Korea was the first market in the world to migrate the majority of users to LTE, with LTE-A rollout now well under way. TowerXchange are starting to pick up the first faint signals that towerco activity may be emerging in South Korea.
Tower count methodology
Counting towers is an imprecise science. Tower counts are often considered competitively sensitive, and it’s often easy to confuse tower counts with base station counts which is NOT what we’re counting (base stations can be mounted on all sorts of points of service, and of course multiple base stations can be mounted on a single tower). It’s also very difficult to distinguish between a live ground based tower (GBT), an incomplete or a dormant GBT, a rooftop, a microcell and even small cells. TowerXchange aims to count investible, shareable sites – so we would count robust rooftops and microcells suitable for multiple tenants for example, but not single operator small cells. Our sources often don’t disclose the exact nature of their site counts, while asset registers are notoriously inaccurate.
TowerXchange’s usual methodology is to verify country tower counts through multiple sources – MNOs, towercos, regulators, site builders – with a strong emphasis on person to person research. As there is little towerco activity in the markets covered in this article, our research has been largely desk based and draws on less sources than usual, thus these tower counts have a higher margin for error
If you wish to share any refinements and corrections, please email TowerXchange’s CEO at kosmotherly@towerxchange.com.
