The Mexican 700MHz shared network: what’s in it for towercos?

red-feature.jpg

TowerXchange roundtable reveals optimism (tempered by ongoing concerns) about Red Compartida

Wholesale LTE networks have been mooted in many countries, deployed in few, and never before deployed in a country with as many towercos as Mexico. Shared 700MHz network Red Compartida is being rolled out by ALTÁN Redes, and TowerXchange revealed several practical insights into the implications of the rollout for towercos at a roundtable at the TowerXchange Meetup Americas 2017. The following article respects the Chatham House Rule – none of the opinions nor quotes in this article should be attributed to roundtable moderator René Espinosa, VP at Mexico Tower Partners, nor any of the other 14 participants.

Red Compartida: saviour of the stuttering Mexican tower market?

It’s been a slow couple of years for Mexico’s tower companies. The Mexican Peso was trading around 0.75 to the USD between 2009 and 2014 when most towercos entered Mexico, but declined to 0.50 from 2015, recovering to around 0.55 more recently. In the mobile market tariff wars have driven penetration at the expense of margins, leaving little capex to be invested into Mexico’s overloaded network.

AT&T’s market entry has not heralded the network investment many hoped for. Similarly Telefónica has added few new towers or co-locations, and is considering exiting the Mexican market. Dominant market leader Telcel remains the only company significantly investing in their network, yet most of their new build has been funnelled to América Móvil’s carve out towerco Telesites. TowerXchange estimate that in the last two years less than 1,000 new towers have been built by independent towercos (excluding Telesites) in Mexico. Yet Mexico’s many independent towercos see light at the end of the tunnel in the form of co-locations, and eventually new build, from Mexico’s new wholesale LTE network, Red Compartida.

ALTÁN Redes emerged victorious from the tender to rollout Red Compartida. The first phase of ALTÁN Redes’ rollout targets 30% population coverage by Q118, and is likely to be concentrated on the country’s largest cities including Mexico City and Monterrey. This constitutes the main project of the year for towercos such as Mexico Tower Partners, Uniti Towers, Torrecom and American Tower, and roundtable participants suggested the first phase would require co-locations on ~2,500 existing towers, of which Telesites were expected to provide ~1,000. Master lease terms were signed by six towercos in late 2016 and early 2017, and equipment has already been installed on some sites.

“Mexico’s MNOs have slowed their investments and deployments, so Red Compartida represents the best opportunity to reach and exceed our goals on co-location,” said one tower company.

“ALTÁN Redes is working faster than any other operator,” added another towerco.

The second phase of ALTÁN Redes’ rollout targets 40% population coverage by Q119, extending the footprint of Red Compartida to larger inhabited areas, but remaining focused on relatively high population density areas.

In three to five years, when the rollout extends beyond the 82% existing coverage and ALTÁN Redes starts connecting settlements with populations of 5,000, new sites will be necessary, consisting of less capitally intensive 40-60m tall towers, and more smaller, lightweight structures. Will Mexico’s towercos want to build what may well remain single tenant towers, given that Red Compartida has an explicit goal to sublease to the companies who would have been towercos’ second and third tenants?

“The towers built in the latter phases of the rollout may be beyond the reach of the existing transport and electricity infrastructure, but if ALTÁN Redes agrees to pay a premium we could build those single tenant towers,” said one towerco. “We have already built some single tenant towers in Mexico at a premium. ALTÁN Redes could use electricity towers in some instances, but they’re high tension towers and capex would need to be invested to make them low tension.”

To achieve the government’s ambitious target of 92.2% population coverage within seven years, one roundtable participant suggested Red Compartida might ultimately require as many as 12,000 sites, comprising around 60% co-location on existing towers, 40% new sites.

Red Compartida might ultimately require as many as 12,000 sites, comprising around 60% co-location on existing towers, 40% new sites

What is Red Compartida?

Red Compartida a public private partnership (PPP) project to create and share a 700MHz LTE network, with the objective of bringing high speed mobile broadband to at least 92.2% of Mexico’s population within seven years.

The Mexican government, via the IFT, has made available to the project 90MHz of coveted contiguous 700MHz “digital dividend” spectrum at a discounted price reported to be just US$30mn (for comparison, the auction of similar spectrum bands raised US$2.3bn in Brazil!)

While the focus of Red Compartida will be on broadband, voice will be possible through VoLTE.

While the project is part of President Enrique Peña Nieto’s Reforma de Telecomunicaciones, and is coordinated by the Ministry of Communications and Transportation and the Federal Institute of Telecommunications, the rollout will be led by ALTÁN Redes. ALTÁN Redes will invest US$7bn over the next ten years, drawing upon a multinational consortium led by Spanish cable giant Multitel and a selection of financial investors including funds managed by Morgan Stanley, FFLATAM-15-2, CDPQ, CKD-IM, Hansam, the IFC, and Multitel Chairman Eugenio Galdón’s Isla Guadalupe Investments. Telcos Axtel and Megacable each have non-voting shares.

Red Compartida has an agreement to use the fibre of CFE (the Federal Electricity Commission, the government owned powerco) for fibre backhaul.

The objective of Red Compartida is to improve coverage (currently just over 80%, with significant areas of the country receiving service only from Telcel), to improve QoS, to drive down tariffs by enabling competition and creating efficiencies, and thus to stimulate economic productivity and industrial competitiveness.

Red Compartida will own and deploy the base stations (reportedly using equipment provided by Huawei and Nokia), but will have no direct subscriber relationships, instead using a wholesale model to sell capacity to licensed Communication Service Providers. Red Compartida is expected to accelerate rollout, and leverage efficiencies, by co-locating on third party towerco sites where possible, and may outsource much of the build of new sites to towercos.


The rural coverage conundrum

The rural coverage debate is not unique to Mexico. While the economics are challenging for MNOs to extend coverage into areas where population density and ARPU may not be sufficient to generate return on investment within an acceptable timeline, MNO investment in rural coverage can be stimulated by regulatory requirements. Towercos have no such obligations, and have business models and investment criteria attuned to tower cash flow (TCF) generation rather than ARPU – thus rural sites remain unattractive as the first tenant may struggle to afford a premium lease rate and a second tenant may be a distant pipedream, and because opex is often higher. The cost of energy grid extensions, or of distributed generation, is also a concern in Mexico, although there is an electricity distribution network extension plan.

Could Red Compartida solve the rural coverage conundrum? Could multiple operators leverage one shared rural network? And will ALTÁN Redes have to build their own rural sites, or will towercos rise to the challenge? The debate came back to one simple response by towercos: “towercos will build the rural sites, but only if ALTÁN Redes is prepared to pay a premium for rural sites such that the TCF is investible.”

One roundtable participant suggested that one of Mexico’s incumbent MNOs had unsuccessfully piloted a revenue share business model to serve rural communities. Another participant called attention to a successful community-led network deployed in Oaxaca by Rhizomatica as illustrative that low cost, open source connectivity could be economically viable in regions with a population under 10,000.

In order to extend population coverage from 82-92%, ALTÁN Redes will have to connect small communities where ARPU might be US$1 or less; it is in these use cases that Mexico’s incumbent MNOs are perhaps most likely to use the ALTÁN Redes network. Ultimately, ALTÁN Redes must make enough money wholesaling coverage in Mexico’s big cities to subsidise rural coverage.

Who will use Red Compartida?

While Telcel, Telefónica and AT&T all have coverage in Mexico City and Monterrey, given the capacity of Red Compartida’s 700MHz band, with one Red Compartida site the MNOs could achieve 2-3x the capacity as when using their own spectrum bands. So as networks fill with data, the three MNOs could turn to Red Compartida to offload and improve QoS, thus reducing churn. ALTÁN Redes has to meet KPIs on cell throughput on the network edge, and claim they will be able to offer a wholesale service cheaper than Telcel. But one roundtable participant said they had explicitly asked Mexico’s incumbent MNOs if they planned to use Red Compartida in the big cities: “They said no, because they have plenty of spectrum available themselves.”

But Red Compartida is designed to do more than just enable the densification of incumbent MNOs’ networks; the government’s hope is that the network will serve new entrant MVNOs, ISPs and enable mobile money and m-healthcare services. “If MVNOs will rent capacity from Red Compartida and enable Mexicans to get a 3G / 4G phone with a cheaper tariff than with Telcel, I can see this working,” suggested one roundtable participant.

“But at present less than 1% of Mexican subscribers use an MVNO,” countered another.

The game-changer would be the entrance of a major new national MNO – particularly if Telefónica does indeed exit Mexico - conversation in the corridors at the TowerXchange Meetup Americas suggested Verizon or Sprint / Softbank could be interested in entering the market.

A network for tomorrow… but not for today?

One of the principal challenges facing ALTÁN Redes and their partners is the lack of compatible devices. “No devices, no customers. Given that Mexico’s existing three MNOs already provide coverage in urban areas, I’m sceptical it will work!” remarked one participant. As more 700MHz spectrum is auctioned across CALA, availability of compatible devices may become less of an inhibitor.

Given the risks inherent in the Red Compartida business model, how comfortable are Mexico’s towercos with the credit risk of their new tenant? ALTÁN Redes has already raised several billion dollars, supplemented by access to vendor credit. It’s already a large operation, with close to 200 employees, although it is not expected to be revenue generating until March 2018. One towerco suggested that Red Compartida would not be allowed to fail because there was too much at stake politically, while another reminded participants that towercos remain insulated from risk by the 10-year term on their co-location agreements. “We can distribute any risk across our large portfolio – and Red Compartida are paying on time to date!” added a third.

How Red Compartida selects and adds sites to its network

Red Compartida’s opco ALTÁN Redes typically chooses the closest existing tower to a required location. Where there are two or three competing locations, they usually go with line of sight. ALTÁN Redes typically prefers to lease space at 30m height, which can mean their ideal elevation on a tower is not available, thus shaking loose a handful of BTS projects in phase one, which could be fulfilled by adding rooftop sites to government buildings, where permitting can be fast-tracked.

ALTÁN Redes reportedly has MSAs with six towercos, so pricing is pre-negotiated and is provided on a tower lease plus rent pass through basis.

ALTÁN Redes undertakes their own civil works, so towerco capex is negligible, with Nokia and Huawei also involved in the Construction Programme.

Conclusions

Red Compartida has provided a welcome influx of new business to the Mexican tower sector. “2017 has been a good year for co-location sales because of ALTÁN Redes,” concluded one towerco “We already have three people implanted with ALTÁN Redes.”

If we make the assumptions that Red Compartida is successful enough to complete its seven year rollout, that it does so on time, and that Mexico’s towercos can agree an investible MLA enabling them to build the rural sites, what could the Mexican tower market look like in 2024?

TowerXchange subscribe to the view that Mexico’s incumbent MNOs, whoever they are in seven years time, will use Red Compartida at least for rural network extensions, but this in turn will mean much of towerco’s organic growth comes from ALTÁN Redes rather than from the MNOs. Mexico is home to more than ten independent towercos – if Red Compartida works, expect this to stimulate consolidation, led by American Tower and Mexico Tower Partners, but don’t rule out Torrecom and Uniti Towers as prospective buyers rather than sellers if the market prospers.

In Q117 TowerXchange estimated there were 36,500 tenancies on 29,320 towers in Mexico, with a prevailing tenancy ratio around 1.24x.

Estimated tower count for Mexico, Q117

mexico-towercount-q117.png

Estimated tower count for Mexico, Q124

mexico-tower-count-q124.png

Taking into account the aforementioned assumptions, by Q124 we forecast there could be 65,200 tenancies on 41,765 towers in Mexico, with a prevailing tenancy ratio of 1.56x. If an MLA cannot be agreed between the towercos and ALTÁN Redes for the rural sites, expect 3-4,000 towers to be moved from towerco balance sheets to ALTÁN Redes themselves in our Q124 forecast.

Gift this article