Mott MacDonald: South American cell tower market investment analysis

Frederico Guedes de Oliveira and Michel Grech of Mott MacDonald offer an in-depth analysis of the opportunities for investment in South America's burgeoning towers market

Introduction

Across the globe, operators are spinning off their tower infrastructure in search of growth, investment capital, shareholder value and better margins. Independent tower companies (TowerCos) are emerging with a strong focus on operational excellence and a drive to new business models. Across Asia, Africa and North and Central America, TowerCos now own significant amounts of tower infrastructure. However, South America has been slower in migrating towards a TowerCo model – with the exception of Brazil, and more recently Chile and Colombia.

South American countries, like many others, are seeing acceleration towards mobile internet growth driving operators’ plans to extend 3G networks and deploy more efficient 4G data networks. We believe that these drivers are likely to create significant divesture of mobile operators’ tower infrastructure. For many years, tenancy ratios have governed the independent TowerCo industry; however, there is a growing shift towards metrics like individual tower profitability, meaning that any new tower acquisition opportunities place more emphasis on overall financial viability.  This paper highlights the result of extensive modelling analysis undertaken by Mott MacDonald to investigate the investment opportunity in several South American countries through a hypothetical market entry by a TowerCo.

Our analysis combines a tower demand forecast with financial projections. It calculates potential market share, and therefore potential lease up rates for the (theoretical) portfolio of interest in each South American country market studied. It also produces important financial metrics such net presence value (NPV), net profit margin, earnings before interest, taxes, depreciation, and amortisation (EBTIDA) and return on asset on a hypothetical market entrant. Our demand forecast is driven by expected subscriber growth (including migration to 3G/4G), the evolution of sharing in the country and the impact of small cells. Central to the analysis is detailed sensitivity analysis for all our assumptions.

South American market overview

Brazil is now mostly completing a migration from operator owned towers to TowerCos. Most of the towers have been acquired by TowerCos and the remaining ones are owned by América Móvil, which currently appears to have no inclination. However, we believe that the aggressive pricing offered by middle market TowerCos will lead to some consolidation.

There is some TowerCo market activity in Colombia. American Towers (AMT) is the market leader and companies such Centennial, Continental and NMS Towers are also present. The Colombian and Chilean markets have many positive characteristics with regard to their potential for TowerCos and the significant potential for subscriber growth. 

With the new tower regulation in Chile and recent investments in LTE rollout, as well as a price war led by MVNOs, and with more than 8,000 towers spread over the country, it represents an attractive market for TowerCos. Although there are some concerns with government policy and macro-economic turbulence in Bolivia, Venezuela and Argentina, most of the towers in those countries remain carrier-owned and this could represent new investment opportunities for established as well as new tower operating companies. 

Investment analysis 

Figure 1 shows Mott MacDonald’s forecast subscriber numbers in 2023 against overall 10 year NPV for our theoretical market entrant.  This highlights that the three largest markets forecast by NPV are Colombia, Peru and Argentina. However, the return on investment for Argentina is significantly lower than for Colombia, despite being the larger of the two markets.

Figure 1 Summary of NPV and forecasted subscribers by 2023

Our analysis indicates that Colombia and Peru appears the most investable opportunities with the highest potential NPVs. However, in terms of return per tower, the capital required for Colombia is substantial given the large size of the market. Normalized NPV per tower is higher than Paraguay or Uruguay, mostly due to the low subscriber per tower ratio but also due to the high number of towers forecast by 2023.

Despite being the two largest markets, the normalised return on investment per tower for Argentina is significantly lower than Colombia, predominantly attributed to the discount rate (weighted average cost of capital) applied to the cash flows to account for the higher perceived investment risk in Argentina.

In Venezuela and Ecuador, although the large size of the countries brings good perspective for tower growth, there is a limited return on investment. This is driven by the lower tower per subscriber ratio and low 3G and 4G subscriber numbers. In addition we have also assumed a higher WACC due to the high risk profile for private sector investments.  

The low NPV for Bolivia, Uruguay and Paraguay are primarily driven by the slow increase in the incremental tenancies and the size of the market. However, the investment required for those countries is considerably lower than countries such as Venezuela, Ecuador or Chile.  

Sensitivity analysis 

Figure 2 shows an example of the type of sensitivity analysis that can be completed. It illustrates a spider chart representation showing overall NPV for a particular country and for three different scenarios. In this example, we can see the impact on NPV as each of assumption is changed by -5% (low) to +5% (high) from the base value. Essentially, the steeper the slope of a particular parameter, the higher its impact on the NPV. Note that the sensitivity analysis only shows the impact of one parameter whilst all others are kept at their base value.

From this, the key factors that have highest impact on NPV are the monthly revenue per anchor tenant, the initial LUR and the average acquisition cost per Tower. With limited TowerCos in these markets and limited public information, our starting point assumptions are based on prevailing rates in Brazil and other Central American countries. For example, in brownfield markets, current tenancy ratios are likely to be 1.3/1.4 but in green field markets, where no TowerCos are present, these are likely to be closer to 1.1. 

The average cost per tower in Central America and Brazil has risen from US$128k in 2010 to around US$213k per tower in 2014, however we recognise that the cost per tower is highly dependent on the structure of the particular deal. Our assumption for average acquisition cost per tower is therefore derived from recent transactions in the region (in particular Brazil) coupled with our assessment of the likely tenancy rates in the respective countries. 

Certainly the top sensitivities change depending on the characteristics of the countries, or more precisely, the assumptions about the specific assumptions made about the countries. 

Figure 2 An example of the key sensitivities to the NPV result assuming a +/- 5% variability on the input assumptions 

The spider chart provides an initial first pass on key parameters. The graph appears linear as it extrapolates the end points over a short range. However, for a wider range in some parameters, there is a non-linear relationship. Figure 3 displays the impact on normalized NPV per tower as well as the return on asset (ROA) per tower as the initial investment assumption is changed.

Figure 3 Impact of average investment per tower on financial results

Concluding remarks

South America is a nascent market from a tower company perspective and is embracing the tower co model. We believe that Mobile network operators will be looking to divest their tower portfolios to raise capital to fund expansion plans or to return value back to shareholders via dividends or improved margins.

Our analysis shows that there are appear to be rich pickings in South America. However, reality will be different as conditions will change due to competitive forces and regulatory evolution. 

Methodology

Our analysis combines national tower demand forecast and respective lease up rates with a financial forecast to determine the economic viability of market entry, through the acquisition of a proportion of the existing tower portfolio. The methodology is shown in Figure 4 below.

Figure 4 Overall Methodology diagram

Our technical model uses a combination of a top down approach and a bottom up analysis to determine expected base station growth by technology and how these base stations will be distributed across existing towers, newly built towers or shared towers. The financial model then uses this tower demand forecast and, assumes market through the acquisition of a certain percentage of towers to and gets a market share of the forecasted tenancies, leading to a forecast lease up rate (LUR) on the acquired portfolio. The operating business model we have assumed is a simple one, where the TowerCo only provides passive infrastructure and is not responsible for power or active equipment maintenance.

Our model considers 120+ independent variables for each country to reflect each unique market’s conditions. The detailed sensitivity analysis allows us to focus on the key assumptions that have the largest influence on the results. 

 

Frederico Guedes de Oliveira, 
Mott MacDonald, London
Email: frederico.oliveira@mottmac.com

 

Michel Grech
Mott MacDonald, London