Mott MacDonald’s Digital Infrastructure team have tried and tested models for evaluating tower investments in existing and greenfield markets. Michel Grech and colleague Frederico Oliveira created hypothetical scenario in which a new entrant towerco acquired 20% of the assets in each of nine different South American markets, ran those scenarios through their model, and came up with tower and LUR forecasts, plus financial KPIs like NPV, return on asset, EBITDA and enterprise value. TowerXchange persuaded Michel and Frederico to share a few hints from this unique comparative study.
TowerXchange: Please re-introduce Mott MacDonald and the Digital Infrastructure team for any readers not familiar with your work on towers.
Michel Grech, Principal Consultant, Mott MacDonald:
Mott MacDonald is a world-class employee-owned management, engineering and development consultancy with global experience, local insight, and a multi-sector perspective. Our 14,000 staff operate in 140 countries through our global network of offices – with various in Africa including Johannesburg, Durban, Cape Town, Gabarone, Kampala, Lagos and Lusaka. In 2014 we were named overall Global Technical Advisor of the Year for the fifth time in six years by the Infrastructure Journal.
Our Digital Infrastructure team offers a unique blend of business and technology expertise; providing an independent perspective on many markets and technologies and successfully delivering projects to leading financial advisors, lenders, operators, equipment vendors, governments and regulators.
Our towerco expertise spans, in particular, strategy development, due diligence (commercial and technical) and operational assessment – and we have recently delivered various assignments in these areas. As an example, we have been engaged as commercial and technical due diligence providers for tower asset transactions involving a total of over 25,000 towers across more than 20 countries in Africa, Latin America and South East Asia, working with most of the major towercos and operators – providing us with unrivalled knowledge and experience.
TowerXchange: What is the methodology behind your evaluation of any given tower opportunity?
Michel Grech, Principal Consultant, Mott MacDonald:
There are three major components to our methodology.
1. Extrapolating subscriber forecasts and projections per technology to a ten year horizon. Industry sources are readily available which quantify and forecast subscriber growth over the next two to three years. Given tower transactions are effectively investments in long-term, ten plus year contracts, a longer forecast is required both in terms of subscriber numbers and substitution of 2G by 3G and eventually LTE.
2. Next we need to divide those current and forecast subscriber numbers by the number of BTS, NodeB and eNodeB deployed in a market to establish the current and future industry benchmarks. For the purposes of the model we’re about to speak about we used an aggregate of BTS at a country level, but we can drill down to an operator level. BTS growth in terms of 2G tends to be driven by coverage, at 3G and 4G by capacity, which gives us an estimated total number of base stations required by technology.
3. Finally we evaluate the degree of sharing already taking place – are there a lot of bi-lateral swaps between operators? Is the regulator promoting or mandating infrastructure sharing? By comparing the distribution of BTSs across existing and future sharing, we can arrive at a forecast lease up rate (LUR, or tenancy ratio).
We leverage these data points and our tried and tested models to generate a pessimistic forecast for the number of Points of Service (PoSs) required in any given tower market across three categories; anchor tenant upgrades, supplemental tenancies and build-to-suit.
Figure 1: Overall methodology
Michel Grech, Principal Consultant, Mott MacDonald:
The aforementioned modelling gives us a starting point for the key financial metrics; revenue as a function of lease rates and tenancy ratios, opex costs (ground lease, maintenance, insurance et cetera), and the investment required to acquire towers. Note that the investment required to acquire towers, and the average price per tower, is not just defined by market factors but also by the structure of the deal, for example the leaseback rate the seller is prepared to take on.
Our models are adaptable to evaluate tower transactions using multiple different business models – sale and leaseback, joint ventures where the seller retains equity, manage with license to lease deals – and can also be adapted according to whether the towerco provides space only or if they provide a full power service. For the purposes of the model we’re about to talk about, we’re assuming the towercos are using a space only business model, which is most prevalent in South America.
We bring all these technical and financial metrics together in a full P&L model looking at NPV, NPV per tower and payback periods. Once we have the full model, we can test the influence of modifying individual starting parameters, financial or technical, and see how that affects overall NPV.
TowerXchange: Moving on to look at the hypothetical model you’ve built to compare tower markets in South America, what assumptions did you have to make to run the model?
Michel Grech, Principal Consultant, Mott MacDonald:
We’ve modelled a scenario in which a new market entrant acquires 20% of the towers in each market. We’ve estimated how many acquired sites would typically be shareable, and the proportion that might not be shareable for structural reasons. Our models have some baseline data on 2G-3G and, where applicable 4G substitution, but we’ve often had to make an informed approximation of the starting point in terms of the number of towers in the market. We’ve modelled ground lease, maintenance and insurance costs based on benchmarks from our work on tower transactions worldwide and in the South American region, but these costs vary significantly according to local market conditions.
We’ve modelled a prospective lease up rate (tenancy ratio) – there’s a delta between the starting point and end point of the lease up rate (tenancy ratio), so we haven’t always used the same starting point. For example, the prevailing tenancy ratio in a market like Brazil, where infrastructure sharing is relatively widespread, might be 1.3-1.4, whereas in a greenfield tower market it is likely to be nearer 1.1.
We’ve used a standard lease rate from the outset, taking revenue per tower at a mid-point to reflect anchor tenants, incremental tenants, non-traditional operator tenants who might require less space, and in cases where the anchor tenant has negotiated a discount if other tenants are added.
Finally, we’ve taken into account inflation and discount rates to reflect country risk in markets such as Argentina.
Figure 2: Ten year NPV by country
Figure 3: Key sensitivities to the NPV results based on a +/- 5% on the input assumptions, example shown here for Colombia
Figure 4: Return on the investment
Michel Grech, Principal Consultant, Mott MacDonald:
Colombia, Chile and Peru offer rich pickings but require the highest investment given the sheer size of the markets.
The tower forecast for Argentina is the highest if it is able to reach the full potential, however NPV is significantly reduced as a result of the perceived country risk and accounted for through the discount rate applied.
Colombia, Chile and Peru offer rich, pickings but require the highest investment given the sheer size of the markets. The tower forecast for Argentina is the highest if it is able to reach the full potential, however NPV is significantly reduced as a result of the peceived country risk
Of the greenfield markets, Bolivia appears to offer a very good return based on the calculated NPV as the investment required is much smaller than other big opportunities.
Revenue per tenant, initial lease up rate (LUR, or tenancy ratio) and acquisition costs are, unsurprisingly the highest sensitivities to the overall NPV and are the top three key sensitivities for all the countries. The other sensitivities depend on the characteristics of the country, which are mostly the forecasted growth in 3G/UMTS subscribers, the sharing profile evolution of BTS on towercos over time, and the cell radii of the various radio technologies.
TowerXchange: If readers want to learn more about the model, how can they contact you?
Michel Grech, Principal Consultant, Mott MacDonald:
For further information please contact me at Michel.Grech@mottmac.com
If you’re interested in investing in, acquiring, operating or building telecom towers in any of the nine markets mentioned in this article, or in Brazil, Mexcio and Central America, then don’t miss the TowerXchange Meetup Americas, coming up on April 28 and 29 in Hollywood, FL!
