Leveraging TowerXchange’s new towerco league table business intelligence database, soon to be made available to subscribers, we reveal that between Q115 and Q117, over 22,000 towers were added to CALA towerco portfolios, 39% of which were acquired through large sale and leasebacks, 61% added through new build or through the rollup of existing, small private portfolios. This study reveals the share of organic growth captured by different towercos; which towercos and which CALA tower markets are growing fastest.
TowerXchange have been compiling a global database of tower ownership, now updated quarterly and stretching back two years – readers will soon be able to subscribe to this data set in our new premium business intelligence service. I have used that database to answer a set of simple but important questions:
- Which towercos are rolling up and building the most towers in CALA?
- In which CALA countries are the most towers being acquired and built?
- And in which countries is the total inventory of towers growing fastest?
Before I reveal the results of the study, allow me to explain my methodology. Apart from the public disclosures of listed towercos, this analysis is based on qualitative market research; specifically the disclosures of towerco leaders, and it should be noted that there can be no guarantees as to the accuracy of those disclosures – TowerXchange ask for a count of completed sites, but some subjects inflate their figures by including works in progress. Given that private towercos probably inflated their figures by a similar magnitude two years ago, the margin for error is small, albeit not insignificant!
The effect of large scale sale and leasebacks have been excluded until the figure five and the associated commentary in this analysis, enabling us to focus on organic growth. However, a significant proportion of the portfolio additions we can count are not new tower builds, rather they come from the roll up of smaller private tower portfolios in transactions too small to be disclosed. These ‘rollups’ could represent 40-60% of the ‘organic’ growth identified.
In any analysis of build volumes, there is a time lag; a significant proportion of the towers built between Q115 and Q117 represent the execution of contracts signed in 2014 before the worst of the economic slowdown and FX devaluation which affected many CALA tower markets. However, with the recent first green shoots of recovery, and anecdotal reports suggesting the volume of new search rings coming from MNOs is picking up, the relative buoyance of the tower markets prior to the period studied is comparable to the current state of the market.
Figure one: Sites added by CALA towercos, Q115-17
13,621 CALA towers were been built or rolled up by towercos between Q115-17, good for a CAGR in tower count of 10.77%. We don’t have the total built by carriers for comparison, but with towercos deploying the majority of sites in CALA’s largest tower markets, and with Telcel and Telefónica having carved out their towers into captive towercos Telesites and Telxius respectively, it’s safe to assume the majority of new sites are being added to towerco rather than MNO balance sheets.
Figure two: Share of new build and rollup growth by towerco
Which towercos added the most sites to their portfolios, Q115-17?
Referring to figure one, Telesites leads the organic growth chart having added an impressive 4,342 sites at a CAGR over 18%. Two caveats; Telesites’ organic growth numbers are inflated by the transfer of later tranches of existing Telcel sites – they are not all new build; and two, while organic growth has been undeniably impressive, lease up of Telesites towers to third parties has reportedly been sluggish. If one is inclined to remove the Telesites growth from the analysis, the site count CAGR Q115-17 drops from 10.77% to 7.45%.
Telxius, which added 27 towers in Q117, has been excluded from this analysis in the absence of a baseline Q115 count (in Q115 the towers remained on the balance sheets of Telefónica’s local opcos, and counts were not disclosed).
Stripping out the large scale sale and leasebacks, American Tower added 1,744 sites and SBA Communications added 1,104 between Q115 and Q117– although as much as half of that was rollup rather than new build.
In terms of sticks in the ground, Telesites, American Tower and SBA Communications may head the table, but in terms of growth rate Phoenix Tower International is CALA’s fastest growing towerco having built 607 towers, more than half of which were in Brazil, good for CAGR in excess of 200%. Again, it should be noted that Phoenix Tower International rollup a lot of private portfolios, so that 607 probably becomes around 300 new builds if you exclude the rollup.
Next in the growth rankings, Innovattel / Torresec added 621 new sites at close to 100% CAGR, while another rollup specialist, Digital Bridge added 1,103 sites across their Mexico Tower Partners and Andean Tower Partners subsidiaries, giving them a CAGR of around 50%, similar to BTC (+450 sites, mostly new build) and Torrecom (+405, also mostly new build).
From an admittedly much lower base than the publics, CALA’s private towercos are growing rapidly, indeed 47% of the towers added outside of sale and leasebacks in CALA in the period studied have been built by towercos other than the three publics; Telesites, American and SBA Communications.
Firms not singled out in this analysis as they have not provided comprehensive enough tower counts, but which seem to be growing fast, include Cell Site Solutions, Torres Unidas and Highline Do Brasil – we feel we have been conservative in modelling that these and other CALA towercos added a total of 2,230 sites between Q115-Q117.
Figure three: Towerco site count additions by country, Q115-17
In which CALA countries have most sites been added by towercos?
Referring to figure three, unsurprisingly Brazil, with 3,326 towers added, sits atop a graphic listing new towers per country. The total is once again an aggregation of new build plus rollup, but we would still estimate that well over 2,000 new build are included in the total.
Next is Mexico, where I prefer to exclude the Telesites addition of 4,117 sites because I figure other towercos may have not had much opportunity to secure the Telcel search rings, thus I count 1,464 sites added in Mexico. While again the majority will be new build rather than organic growth, in addition to the usual levels of roll up M&A, we are aware of one acquisition with a triple digit tower count by Mexico Tower Partners from a private towerco we’ve been asked not to name, so the total new build in Mexico from Q115-Q117 was less than 1,000.
995 sites were added in by Colombian towercos during the period studied – a significant total, but not necessarily sufficient to sustain the gold rush of new entrant towercos noted during the period. While there wasn’t much rollup M&A in Colombia during the period studied, a number of those new entrants will become stranded with portfolios generating insufficient tower cash flow to attract further financing, thus becoming rollup targets for the country’s larger towercos.
Figure four: Towerco site growth by country as a % of Q115 total country stock
Referring to figure four, when we analyse site growth as a percentage of the baseline stock in Q115, the rate of growth in Central America is startling, headlined by Nicaragua which has seen the country’s inventory of towers increase 55.9% during the two years studied. Indeed the 1,770 sites added in Central America aggregate to 16.1% CAGR, compared to 6.3-6.8% in the larger tower markets of Mexcio, Colombia and Brazil. Ecuador and Argentina only rank so low because towerco activity is so recent; expect Argentina in particular to rise to 3% CAGR in 2017 and above 6% in 2018.
Paper is at least as important, if not more important, than tower count
Given the lack of disclosure of lease rates and contract terms, our analysis by necessity must focus on tower counts, rather than on the profitability of those towers, so TowerXchange must emphasise that focusing on build volumes would be to the detriment of a more appropriate focus on tower cash flows and risks to future cash flows. The value of a tower portfolio, whether in CALA or anywhere else for that matter, is derived more from the margin between the ground lease and the lease rate, and the number of tenants on the tower, than from the sheer asset count. Success comes from building and buying good structures, in good locations, with good permits and good lease agreements.
Tower entrepreneurs risk creating uninvestible portfolios if they win BTS contracts and drive site count growth by offering deeply discounted lease rates, discounted (if any) escalators, by granting concessions in cancellation and renewal terms, and concessions in the governance of reserve space and RANsharing. The publicly listed towercos and Grupo TorreSur might not have matched the growth rate of the CALA’s independent developers, but they would contend that they are happy to pick and choose the best sites on the best terms, and prefer not to get drawn into price wars.
SBA Communications President Jeff Stoops said of their strategic siting strategy it “doesn’t always yield high volume, but it is good for return on capital invested.”
Conclusion
CALA needs towers to be built at a faster rate than the ~5,000 new towers built per year by towercos in 2015 and 2016. The period surveyed is universally acknowledged to represent ‘down years’ for new build volume. The problem wasn’t a lack of towercos with capital to deploy – if anything, there was an oversupply of towercos during the period studied. The problem was underinvestment in passive infrastructure by carriers under increasing pressure from declining ARPUs, market consolidation and FX depreciation.
While CALA’s tower network was somewhat densified and expanded, the increase in total stock of towers was less than 10%, and data usage and subscriber growth have outstripped that network investment by a significant margin. The region has an ample supply of well capitalised towercos ready to invest and close the widening infrastructure deficit in CALA. Encouraging signs are emerging – in late July TowerXchange spoke to one of Brazil’s leading carriers who said they had signed agreements with towercos to build more than 2,000 new sites. TowerXchange anticipate a significant acceleration of both new tower build, and rollup of existing towers, in CALA in 2017 and 2018.
How does the analysis change if you put the sale and leasebacks back in?
Inclusive of organic growth, rollup and large scale sale and leasebacks, CALA towercos added 22,323 sites to their portfolios between Q115-17, a 17.13% CAGR.
American Tower heads the graphic with 9,627 net additions, including acquisitions from TIM Brasil announced prior to the period studied, but closed within it. We also include announced but then not closed deals to acquire 1,400 sites in Paraguay and 1,000 in Argentina, although we exclude acquisitions announced after the period studied: 1,200 from Millicom in Colombia, 142 towers from Axel in Mexico, and a final 56 TIM Brasil sites.
Between Telesites and SBA Communications, Phoenix Tower International sits third, albeit with a stunning CAGR in excess of 500%!
Figure five: Towerco site count growth Q115-Q117 inclusive of organic and inorganic growth
