Man on a mission: Akhil Gupta on the disarmament of operators

Akhil Gupta is truly one of the founding fathers of the tower industry as we know it today. He was responsible for the separation of Bharti Airtel’s passive infrastructure and subsequent founding of Bharti Infratel, one of the world’s largest and most respected towercos. He is also a co-founder of Indus Towers, a joint venture between Bharti Airtel, Vodafone and Idea Cellular, which has become the largest tower company in the world. Akhil is Vice Chairman of Bharti Group and Chairman of Bharti Infratel. He is also Chairman of TAIPA (India’s Tower and Infrastructure Providers Association) and President of TSSC (Telecom Sector Skill Council of India).

TowerXchange: Can you provide a brief introduction to Bharti Infratel and give our readers an idea of your history and footprint?

Bharti Infratel has about 35,000 towers in 11 of the 22 telecoms circles in India, plus a 42% share in Indus Towers, which operates in 16 circles, giving us a total of 85,000 towers. One of the world’s largest towercos, Bharti Infratel is a listed entity with a US$12bn market cap today. We maintain complete neutrality and are non-discriminatory vis-à-vis all Indian MNOs; all tariffs are transparent and are the same for all clients.

TowerXchange: Thinking back on the founding of Bharti Infratel, what motivated the carve-out of tower assets?

Many people think it was just to raise funds but this isn’t the case; I know because I was part of the top team that developed the strategy behind this move! We realised that operators were fundamentally ill-suited to manage passive infrastructure efficiently. They are by very nature marketeers and thus best suited to do marketing, sales, brand-building and advertising. Managing passive infrastructure is a low technology engineering job; refilling diesel on-site is critical but can seem mundane; it was often not appropriately prioritised and operators were never able to optimise opex and maintain the best uptime. At Airtel, we had earlier outsourced IT and network management on the principle that our partners had better domain knowledge, better economies of scale and could attract better human capital vis-à-vis the activity in question. We felt we must do the same for passive infrastructure (i.e. towers) since we as an operator did not have the requisite skill set and mindset for the same. The other reason of course was that we felt that this infrastructure would need to be shared extensively amongst operators to reduce cost, especially in low tariff countries like India, which was possible only with independent towercos. However, we realised that there were no companies with those skill sets and so we decided to create a company and outsource tower maintenance to it. This is how Bharti Infratel was formed. The basic principles for the towerco were to share towers on a non-discriminatory basis amongst all operators and to have a business model that would create a win-win scenario between MNOs and towercos. Accordingly, we designed a Master Service Agreement (MSA) whereby with the addition of a second or third tenant, it would result in a lower charge for everyone. Towercos make a lot more money with a second tenant, but the anchor tenant also gets approximately 20% relief on their lease rate and energy charge. This made it a true win-win situation for both and resulted in a unique situation where a company makes more money when its existing customers start paying less than before. The principle of sharing benefits between MNOs as more tenants come on board is fundamental to this industry, and it also ensures operators will never find it worthwhile to build their own towers. We’re on a mission of operator disarmament – to disarm their manpower and make it economically unfeasible for them to build their own towers. I’m pleased to say that today virtually no operator in India builds its own towers.

TowerXchange: Reflecting back on the restructuring of MNO licenses in 2012, how did that change the way the tower business was run in India?  

Since the tower business has evolved on this basis of a strong, deep-rooted win-win business model, licensing and regulatory changes have made little difference.

TowerXchange: It seems like the Indian tower market has restructured to the extent that M&A deal flow is recommencing - do you expect to see consolidation among the Indian towercos in the coming year?  

Personally, I feel there are not too many towercos in India: Infratel, Indus, Viom and Reliance Infratel  have sizeable portfolios. ATC has a little over 12,000 towers, and there are a few other small tower companies. Over 85% of the total towerco-owned towers are with top three towercos. Consolidation in the tower industry is not an important change. Having too many operators in a market causes havoc, but in the tower business it doesn’t. If I don’t have a tower at a location and say Viom does, they can offer a second or third tenant rate much cheaper than what I could offer with a new tower; this fosters the automatic co-existence of towercos. Demand is high and there are long-term contracts with our clients with pricing protection. As a result, no towerco can offer much lower rentals for incremental business, with the result that towercos do not typically witness price wars.

TowerXchange: What are the prevailing lease rates in India?

Prevailing lease rates are around 32,000 rupees per calendar month (approximately US$500) for a single tenant on a ground based tower, on top of which each state has a different energy charge depending on grid availability. This is discounted to 28,000 rupees with a second tenant and 24-25,000 with a third tenant. The rates for rooftop sites are lower but with the same construct.

TowerXchange: Are there any differences in network coverage and density between the different Circles, or is it a fairly mature network across India?

The whole country is pretty well-served now; Airtel for instance covers about 83% of the geography and towers are deeply distributed across every state. There is more density in some Circles than others, but there are no vast areas without towers. However, many more towers will be needed for the continuing increase in data demand.

TowerXchange: What do you think of the rumored towerco launch by BSNL?

This would be a welcome step and it has been in the works for some time now. The sooner they can launch a towerco and make all of those towers available to operators, the sooner we can stop building parallel infrastructure. We as towercos are not in this business to put a tower in a place where we might have a single tenant for five years; the return on capital invested and return on equity wouldn’t make sense. We do not want to put up single tenant towers unless we absolutely have to. As such, I’m happy if BSNL has towers that they make available to Bharti Airtel and other operators. TAIPA has offered to co-operate with BSNL and help them bring their 35,000 to 45,000 towers to market sooner rather than later.

TowerXchange: Tell us about Bharti Infratel’s Green Towers P7 programme, the role RESCOs are playing and could play in the future?

One of the major costs to operators is energy. From our point of view this isn’t a part of the business we have to make a lot of money from; we’d love to work with partners to provide the lowest cost energy by putting capex into renewable energy. Indian grid power is not always the best, which means our networks have a historical dependence on diesel generators. We are driven by environmental issues and cost reduction to speed up the P7 initiatives and reduce energy costs below that of diesel generator-powered units. Ultimately a towerco is designed to run passive infrastructure; we are not a power company and have no aspiration to be a power company. If Renewal Energy Service Companies (RESCOs) can provide an electricity connection 24 hours a day 365 days a year at a reasonable price or if our clients can put in their own connections and have their own meters, that’s the ideal scenario. Until then, power is an integral part of what we do,  and we’ll do what we can to make energy more efficient, and to achieve uptime SLAs with our customers. The RESCOs in India are currently very small; a lot of companies are thinking about it and trying to raise the required capital. In the meantime our aim is to make as many towers diesel-free as we can. Currently, 23-24% of our towers are diesel-free. We manage sites with 16+ hours of grid with large battery banks; we make extensive use of efficient lithium-ion batteries that charge very quickly and have deep discharge. We also use solar energy where we can and where the conditions are suitable.

TowerXchange: How achievable are the Indian government’s green energy targets for 2015 and 2020?  

Green energy targets are under discussion, but our stance has been that it’s the government’s job to provide electricity. It’s unreasonable to force the telecom industry to build out renewable generation capacity beyond the reach of the grid when the telecom industry consumes a small proportion of India’s diesel. We’re happy to install renewables on a voluntary basis, but if the government is not mandating electric cars for the automotive industry why should they mandate solar energy for cell sites? The best way to reduce carbon footprints is to provide reliable grid power and that is the responsibility of the ministry of power. It is important to note that as per an independent study by Nielsen, the telecom industry in India only accounts for less than 2% of total diesel consumed in India.

TowerXchange: Some analysts have suggested that India will need twice as many towers for the 4G era, increasing from 400,000 today to 800,000. Do you agree with such a forecast? 

Indian MNOs have overall invested close to US$50bn in spectrum auctions to date. Most of this spectrum will be used for data networks. As more investment goes into 3G and 4G using a lot of higher frequency bands, there will be demand for cell site densification. I’m less concerned about the increasing number of towers, but more interested in the increasing number of tenancies. There are about 600,000 tenancies in India today, and we definitely expect that to double with the continuing 3G and new 4G rollout. In our view it is not a race to build more towers, but a race to increase the number of tenancies.

TowerXchange: How will this infrastructure growth be financed? What role will the towercos continue to play?

The operators have put a lot of money into spectrum and active infrastructure and they do not build towers. Towercos must invest in passive infrastructure going forward. We must make sure that when the operator’s question is build versus buy that they always buy. Towercos do a better job achieving economic and capital efficiency. Do towercos have capital to invest in networks to keep up with data demand?  It would depend on each company. At Bharti Infratel and Indus Towers we have no limitations on capex – Infratel has over US$1bn in cash, and no debt; we have a huge capacity to build whatever is needed. In addition to this we have a host of ways to extend our business model and tap into this capital to do that.

We’re looking at opportunities to put more capex into connecting towers to local fibre and transport networks, as well as replacing single operator microwave networks with a common microwave network.  There is also an opportunity to wire up buildings for IBS and Wi-Fi and offer this to operators as an add-on service. 80% of data is consumed in buildings, and we can provide IBS and Wi-Fi on a white label basis so customers see their operator’s network when they use it. From that building the traffic can then be transmitted over fibre. Projects like these are capital intensive, but towercos are ideally placed to carry them out. We are running some pilot IBS and Wi-Fi projects now and we feel these would have strong potential.

TowerXchange: Does Bharti Infratel have any plans for international growth or will you remain focused on the domestic market?  

We have looked at Bangladesh, but there are some important regulatory issues which need to be sorted out first. There are three main issues; the taxation of tower asset transfers; the current policy whereby the operator must retain at least 51% of the towerco (experiences in other markets have proved that it should be other way round: towercos must own at least 51% and operators a maximum of 49% to make it non-discriminatory and not under control of the operator); in addition to this VAT has to be charged, but there is no pass through. We’re seeking to meet the regulator to discuss these issues.

TowerXchange: Could you reflect on the ongoing process to divest Airtel’s African towers - what was the thinking behind selling those assets as opposed to rolling them into your own towerco as you have done in India?

We seriously looked at taking over Bharti Airtel’s African towers, but we realised that Africa is not a mature enough market to accept a towerco whose group company runs operations in the same country, unlike India where it is accepted that operator-owned towercos are non-discriminatory independent towercos. As a result the decision was made to outsource this to other towercos already operating in SSA. Deals have been agreed in all of Bharti Airtel’s largest African markets, but it’s possible that due to regulatory regimes the towers in two or three countries won’t be sold. That won’t be a problem for us.

TowerXchange: What were your experiences running the African tower sale process in-house?

I felt if the principles and objectives of the tower sale were clear, and the process was simple and clear, we could run it in-house. The moment the second tenant came I wanted a big relief on our rental and energy costs; once that principle was established, the process became relatively simple. We’ve been running a towerco for seven years, so we definitely had the experience to do this in-house! The transaction was not motivated by raising a few billion US$; it was more important to relieve the ongoing burden of passive infrastructure capex and transfer that infrastructure into specialised hands.

TowerXchange: Please conclude by sharing your vision for how infrastructure sharing can create value for telecoms worldwide.

First of all, I think TowerXchange are doing a fantastic job in bringing the efficiencies created by towercos to the forefront of everyone’s agenda. This is one area where telecoms can adopt a similar strategy in both the emerging and developed world, and it is a primary need around the globe. I’m a supporter and proponent of the tower industry and business model. It has now been proven that since the model is based on sweating out existing assets, there is progressively higher growth between number of towers, tenancies, revenues, EBITDA, EBIT and PBT for a tower company. The other striking feature of the towerco model is our long term contracts and consequent predictable revenue stream and strong free cash flows. Our vision for the next three to five years is to get all the passive infrastructure out of the hands of the MNOs. Infrastructure needs to be shared – gone are the days when owning towers conferred a competitive advantage. Nobody has ever been able to stop a competitor putting up a tower next door, but the efficiencies that can be achieved by sharing towers are undeniably compelling. I do believe that outsourcing passive infrastructure to towercos creates significant value for telecom companies. While huge capex on towers by an operator is always a drag on its valuation, a towerco gets a multiple of the same as value due to better capital productivity.

Akhil Gupta is the latest member of the TowerXchange “Inner Circle” Informal Advisory board.