Towercos own around one in three of the world’s ~3mn telecom towers. When China’s legacy towers are injected into CTC, that proportion will rise to two in three; MNO-captive towers will be in the minority. One struggles to imagine a parallel wherein an infrastructure asset class has been so radically transformed in such a short space of time. TowerXchange estimate that the tower industry currently owns 971,696 towers worldwide, assets worth ~US$100bn, but if smartly managed the aggregate market cap of companies in this asset class could be in excess of US$200bn.
After three years of studying the telecom tower markets on every continent except North America (which TowerXchange intend to start covering in Q2-3 2016), we can now make an educated estimate that there are around 3mn telecom towers worldwide. We’ve seen very few forecasts that low, which gives us even more confidence in the number – in every market we study, the actual number of sticks in the ground is usually significantly less than most assume – distorted by the more common count of base stations, and by the grey area separating a tower from a rooftops.
We call them ‘tower counts’ but what TowerXchange actually endeavors to count are shareable / investible structures – thus a robust rooftop suitable for co-location is more interesting than a lightweight pole mounted on a building which lacks the structural capacity to mount anything heavier up there. The imprecision of the differentiation serves to illustrate the approximate nature of ours and anyone else’s tower count: when is a tower not a tower? With MNO asset registers seldom accurate, and towerco asset registers subject to occasional inflation by sales teams keen to advertise tenancies on towers that are not quite finished, counting towers is an imprecise science.
Figure one: Comparing towerco penetration worldwide
Those caveats notwithstanding, as we roll up tower counts in each region, we see the ‘old growth’ tower markets of USA (initiated in the mid 1990s) and India (initiated over a decade later) approaching saturation with towercos owning 82% and 76% of the towers in each market respectively. One effect of this is that investors who achieved good returns from the asset class, particularly from the U.S., are seeking new investible platforms in less penetrated tower markets. Thus one seldom hears of a tower auction with less then ten well capitalised bidders. Capital is no longer scarce in this asset class – proven management teams are less common!
I’ve undertaken a crude comparison of average cost per tower in figure two. While cost per tower is a horribly inaccurate measure of valuation, lease rates are seldom in the public domains so multiples of tower cash flow (TCF), which would be far more useful, are not accessible. While flawed, this analysis still reveals the premium necessary to pay to participate in two of the next most penetrated tower markets; CALA and Indonesia, reflecting the increasing scarcity of remaining investible assets, the competitiveness of tower auctions, and the reduced risk of investing in markets with embedded cultures of infrastructure sharing.
As an increasing proportion of the world’s most investible towers have been acquired, towercos and their investors may find themselves having to deviate further from the established business model to extend their growth narrative. We have already seen the likes of American Tower extending beyond the limit of the ‘steel and grass’ business model to offer ‘power as a service’ in emerging markets - indeed, the more operationally complex the market, the deeper the service proposition should be; just look at the end to end services edotco is providing in parts of Asia as an example.
Figure two: Average cost per tower comparisons
The ‘rules’ governing deal structure may have to be rewritten too, if the tower industry is to penetrate beyond owning two in three of the world’s towers. We’ve seen the likes of IHS deviating from established wisdom which dictates that towercos must own at least 51% equity - MTN retained 51% in Nigeria. There are countries where foreign direct investment is limited to 49% - if the international tower industry wanted to participate in opportunities which could come to market in the near term in Algeria or Thailand, for example, then investors would have to content themselves with a 49% stake. Such restrictions have not prevented Tower Bersama and Protelindo from prospering in Indonesia! A different flavor of deal structure predicated on shared risk partnerships may be required if towercos are to absorb tier two operators’ assets onto their balance sheets.
Finally, the macro network-centric tower industry of today may evolve substantially in the heterogeneous network era. We have already seen the likes of Crown Castle, Protelindo and STP investing in small cells and micro cells, while an increasing proportion of conversations TowerXchange is having in Europe acknowledge that towercos on that continent may have to manage active as well as passive equipment.
