Towering sharing the way forward for Bangladesh’s MNOs to drive efficiency and customer experience

In a country where SIM penetration is high and 3G adoption is on the rise, yet data and voice tariffs remain amongst the lowest in the world, operators in Bangladesh are under pressure to develop sustainable business models. While the MNOs have historically built their own towers, they are becoming more cognizant that tower sharing is the way forward and towers are no longer strategic assets. The regulator is compelling commercial tower sharing and MNOs need assets everywhere to fulfill their service obligations. With an eye to the future, VimpelCom’s Banglalink talks to TowerXchange.

TowerXchange: Please introduce yourself and your role.

Peter Dindial, Programme Director, Banglalink: At the moment I am the programme director responsible for the carve out of the Banglalink infrastructure assets. My ambition is to establish the portfolio for the future. Once the tower company is formed I’ll take the lead and take on the financial accountability.

I’ve been in telecom for 30 years, been a CTO and done a lot of M&A, in particular for Telenor. I’ve been in Asia and Europe, and just come out of running the Jawwy personal mobile experience programme for STC in the Middle East. I’ve been in post since September 2016 with Banglalink.

TowerXchange: What can you tell us about Banglalink’s network infrastructure in terms of number of sites and typology (ground based towers versus rooftops)?

Currently there are 5,890 assets, excluding in-building solutions (IBS), which we may add to the process, bringing it up to about 6,000 total. Out of that, about 50% are green field sites, which predominantly service suburb and rural areas, with the other 50% as urban rooftops.

Most of the sites are about 10-12 years old as the big push on network growth occurred around 2005. Average construction age is 2009. The majority of the underlying ground leases are quite long, with three to ten years left on most. We’re currently extending 129 leases with less than three years remaining. The landlords here are quite savvy as to the value of their assets to telecom operators, but contractually we are entitled to co-locate.

TowerXchange: To what extent are Banglalink’s sites shared with other MNOs and on what basis (e.g. barters versus commercial leases)?

Money is changing hands, as regulated under the tower sharing framework. There is no bartering; there are different commercial agreements between different partners. For Banglalink, around 20% of our towers have tenancies, and a handful have three or four tenants.

TowerXchange: Please put that into a broader context of the tower market in Bangladesh as a whole: roughly how many towers are there? Are the other MNOs sharing sites? We’ve estimated 24,000 sites between the three big MNOs, with just under 30,000 in total in the country.

Yes, TowerXchange’s numbers sound about right.

This year has been quite interesting, with conversations between MNOs on providing services to each other to fulfill network needs. The desire to build our own or build to suit doesn’t seem to be there, rather there’s a very active desire to increase sharing relationships with everyone.

The economics of the market is that the MNOs have been in business for many years and the market has matured to 82% SIM penetration. However, the market rates for voice and data are amongst the lowest in the world. Investments in 3G and LTE will be a significant drain on the MNOs’ balance sheets, as they look to reduce costs, raise capital, and have sustainable business models that continue to service the existing population. MNOs are under financial pressure to sweat their assets.

TowerXchange: Are tower lease rates at commercial levels a towerco would charge?

There seemed to be a point where pricing was being pushed down by some volume co-location deals, however that seems to be at an end and MNOs generally recognise pricing needs to stay at current, rational levels. There are still operational costs to cover and some margin needs to be retained for this activity for towercos to cover overhead and remain investible.

TowerXchange: How would you characterise population and geographic coverage? And the progress of 3G roll out? What still needs to be built and overlaid?

From my experience since I’ve been here and the conversations with other MNOs, I see a big distinction between urban and rural growth in 3G. Early estimates of 3G coverage were under-estimated, so tower sharing is imperative to enable cell splitting and to fulfill capacity requirements in urban areas. There is a lot of pressure to fulfill capacity requirements so that’s driving a lot of co-location requests in urban areas.

Broadly speaking the country is well covered. It is a dense population, with 160mn people packed into a small area. Different MNOs are stronger in some areas than others, so there are holistic synergies between, for example, Robi and Banglalink.

In terms of 3G rollout and capex of deployment, it hasn’t peaked yet. There is a lot of demand and a lot more to put out. The 4G spectrum auction is also still in the cards and the Bangladesh Telecommunication Regulatory Commission (BTRC) is in the process of determining when to hold it. The spectrum issues do need to be addressed, but LTE is coming and 3G is still growing.

The country is early in terms of 3G but mobile broadband handset penetration is increasing, with low-end 3G handsets permeating the marketplace. Sub US$50 handsets are selling fast and feature phones are virtually non-existent in the stores now. This is very good for the market as 3G-enabled handsets become the norm and pricing becomes more affordable for consumers.

TowerXchange: Do the economics of coverage pitter out at any point? As we know, there isn’t always the business case for an MNO or towerco to undertake rural coverage. Is there anything else to fill in the gaps?

There is a social obligation fund that all MNOs contribute to, 5.5% of turnover, plus 1% into a universal service fund (USF) which has not been significantly deployed yet.

We’ve done detailed analysis on where all the sites are across the country; all the MNOs share information on the sites. So in terms of geographical coverage, there is a MNO for every part of this country. Holistically there is coverage, however, not all markets are the same for all operators. But this adds to the rationale for infrastructure sharing: the more it is encouraged and supported, the more to everyone’s interest.

TowerXchange: We understand there is another towerco in Bangladesh, edotco, with around 8,000 sites – how is the overlap between your portfolio and their managed?

We are working on a decommissioning and consolidation plan with edotco/Robi. We’re not at a point where we have the final analysis, but we’re looking at taking out up to 20%, which is over 1,000 sites. If you look at a Google Earth map of Bangladesh’s sites, the number of cell sites per square kilometre is quite amazing. With Banglalink and edocto’s assets, there’s more than enough capacity to consolidate with minimum capex downside.

The towers on green field sites are robust and most can take another two tenants. This consolidation could be very positive, helping to drive tenancy ratios up by 0.2 to 0.3 at least.

In terms of cancelling the leases on decommissioned sites, it’s quite manageable as we have non-reciprocal cancellation clauses of six months.

TowerXchange: We understand the Bangladesh Telecommunication Regulatory Commission (BTRC) has drafted a tower license framework, soliciting industry comments last year – what does the framework propose, what was the focus of your feedback?

The framework was an extension of the existing sharing framework, which set the concept of commercial trading for assets. However, there were three major sticking points which the industry gave feedback on:

1) Limiting the number of tower licenses to two in the country. This would not be acceptable as it would create a duopoly, which might not build market confidence in a tower industry.

2) Limitations on foreign ownership to sub-50%. This would restrict the participation of professional towercos who can bring value and innovation to the country.

3) Restrictions on MNOs being shareholders or part owners in any towerco.

The three issues were rejected by the industry overall as they are not conducive to growing the industry, while limiting investments and innovation.

The consultation took place last year and the feedback seemed to be received positively by the regulator. Majority foreign ownership may not be a problem anymore; for example the BTRC has allowed Axiata’s edotco to increase their shareholding to 80%. This also illustrates that the BTRC are softening on MNOs as shareholders. Two out of the three issues have evolved positively for the benefit of the industry. However, we’re still unsure about the limit on licenses, though the BTRC seem to accept the concept put forth.

TowerXchange: At one point there was talk of some sort of state towerco?

It was an idea from 2015, but there is no mention of it in the draft regulation. We don’t think this is a current plan.

TowerXchange: Please tell us about the operational challenges running a tower network in Bangladesh. For example, how reliable is the grid, and how are sites managed and maintained? What initiatives have you undertaken to drive toward operational excellence while reducing operating costs?

The Bangladeshi grid is not that bad in comparison to some of the countries I have been in. The grid is much better, especially in urban areas. There are only about 25% of our sites where we need a DG, the rest are batteries plus grid.

Traffic is challenging in the main cities, but it’s not a big country so drive times and mean-time-to-repair (MTTR) are reduced. It doesn’t take a day-and-a-half to service sites like some sites in Africa, for example

The government to their credit put a lot of efforts into generating more electricity for the national grid. Power is important for the country to go forward so the grid is less of a problem, though reach might still be a challenge from some rural areas.

We have to respect this is not a country full of highways, so you spend time moving around with rural roads. Given the GDP, cost per capita, the cost of services are low, so opex costs are quite benign.

During monsoon it can mean some roads are impassable, but all necessary site autonomy is built into the networks; we don’t see a significant drop in statistics for uptime during monsoon season.

There is redundant capacity on towers. For example, I visited one 50m rural tower and the infrastructure built around it was phenomenal – towers and cabinets on 2.5m plinths to be raised out of any flood water, large landscape, easy access – my predecessors did a great job designing this network!

TowerXchange: Please summarise how the commercial, regulatory, and operational environment affect the tower market in Bangladesh.

Whatever happens as we conclude our negotiations with the BTRC, the tower regime here will be good for the tower owners, for the investors, and for Bangladesh. There is so much opportunity in this that if people are rational and look towards the better outcome, there are great business opportunities for everyone.

In the past, rumours on the future tower licensing legislation scared off a number of prospective international investors, who could bring a lot of value to help advance the marketplace here. As certainty around the tower licensing regime increases, risk reduces.

The MNOs here respect towercos and the value they bring for the customer experience. There is an established infrastructure sharing culture and the economics are being well managed with no loss leaders in any of the portfolios. So it’s a win for everyone.

Our portfolio is strong and robust with great opportunities for growth. There are horizontal moves we can make here. Not just servicing MNOs, but also Wi-Max, ISPs, and other non-traditional MNO tenants. Some of our assets are so well built they could even become localised data centres. There’s so much that could be done with the assets that hasn’t been explored yet.

The growth in this era of telecom will be around the tower business.

The ability of any MNO to unleash its towerco business would allow a welcome injection of capital which in turn will allow MNOs to better participate in a future spectrum auction, which at the end of the day, is good for the consumers as they get better service and experience, and good for the government as MNOs have more capital to bid for spectrum.