AGB 2021.4 - American Tower (AMT)
The Tower of Power
Subscribe to AGB - One analysis of a good business every two weeks.
American Tower is the largest owner and operator of wireless towers and other communications real estate globally. The company mainly earns monthly recurring revenues by leasing space on its tower sites to wireless carriers (e.g. AT&T and Verizon in the U.S.). The carriers install and maintain their own radio equipment in leased vertical space on the towers and enter long-term agreements with American Tower with annual price escalators (3% in the U.S. and inflation protected in international markets). These contracts are typically 5-10 years in length with multiple renewals set in place and importantly, are non-cancellable. The company experiences low average net churn in the range of 1%-2% annually (though it’s been higher in recent years due to India carrier consolidation), which makes the business very predictable.
Tenancy is an important metric for tower operators because the business is a high fixed cost, low variable cost model. Towers require very little maintenance capex and the tenants are responsible for their own equipment, even the cabling that that connects the wireless equipment to the edge network. To add an additional tenant requires very little additional operating expenses, and typically the 2nd or 3rd carrier on a tower results in very high margin dollars falling to the bottom line.
The tailwinds and the runway for growth for the tower leasing business are large. The wireless industry is in the beginning stages of upgrading to the next standard of wireless technology (4G to 5G) within most developed markets. With this new standard, users will be able to access the internet from their mobile devices at much quicker speeds (10x-100x) than the previous generation with lower latency and more spectrum efficiency.
With these new capabilities, data usage over wireless networks should continue to grow exponentially as they have in the past. From its FY19 annual report, SBA Communications estimates that Global mobile data traffic will reach 160 exabytes per month by 2025, an increase of 321% over 2019. In the U.S., data usage has grown 30%-40% annually and wireless carriers have had to put up more and better equipment to handle the data loads. New and heavier equipment typically results in amendment revenues for the tower companies.
In many developing countries, consumers are still migrating from basic handsets and smartphones that are capable of 3G/4G wireless data speeds. This implies that there is a longer runway for growth in these markets. American Tower understands long-term opportunity in developing markets and has invested in tower sites and fiber in these countries in anticipation of that growth.
The three largest U.S. based tower companies are American Tower, SBA Communications and Crown Castle. If you look back to 20 years ago, these businesses were fairly similar in makeup, but they are very different today. American Tower now has much higher international exposure than the other two tower companies as the result of M&A and expansion into many of the major international geographies. In FY11, almost half of American Tower’s properties were based in the U.S., with no presence in Europe. Fast forward to FY19 and 41% of its properties are in Asia while the U.S. only accounts for 23%.
SBA has small international exposure in Latin America and Africa but lacks presence in Europe and Asia. And Crown Castle has elected to invest in fiber and small cell antennas (as opposed to towers that host macro cells), which many believe to be a much lower returning business. American Tower’s U.S. properties are similar to SBA’s in that the towers cover mostly rural and suburban areas, whereas Crown Castle has much higher exposure to suburban and urban areas.
Why is it a good business?
American Tower benefits from scale economies and switching costs. Scale mainly helps when spreading certain fixed costs across more tower sites and negotiating contracts with carriers (the more coverage a carrier can get with one contract, the more power in the negotiations). This also helps with increasing tenancy at the company’s tower properties and as we’ve discussed before, increasing tenancy is material to improving margins.
The company also benefits from the switching costs of its customers. From an economic standpoint, carriers should be reluctant to switch tower companies unless they absolutely have to because removing and moving equipment from one company’s tower properties to another’s is the responsibility of the carriers. Crown Castle has stated that it costs $40k to remove equipment from a tower. With annual rents ranging $20k-$30k, it’s easy to see why switching would be cost prohibitive. Furthermore, switching could also lead to potential downtimes in the wireless network within certain coverage areas, which would be bad for business.
As mentioned above, mobile data usage is expected to grow exponentially, which will lead to eventual upgrades of the wireless equipment on tower sites. These upgrades typically mean heavier and bulkier equipment, which leads to amendment revenues for the tower companies. With 5G, that means the installation of massive MIMO antennas, which require heavier mounts and larger pipes to transmit data. This trend can already be seen at SBA with amendment revenues driving a larger part of the growth in the leasing dollars at its U.S. based towers. Since F18Q4, amendments have increased from 53% of increased leasing activity to 84% in F19Q3. American Tower has also commented at a sell-side conference in August 2019 that the average price per amendment has gone up from $400-$600 range historically to $800.
From a carrier’s perspective the economics are just better to lease communications sites from tower companies, rather than build and operate their own. A lease scenario vs. a build scenario results in a present value savings of over $200k after a 20 year period. This assumes $275k construction cost with $1,250 in monthly operating expenses that increase at 3% per year.
So why would it be more profitable for a real estate company like American Tower to build or buy communications sites and lease them out to carriers? The answer comes down to average tenancy. A carrier that owns its own tower assets will likely have a much lower average tenancy than an independent tower company would. That’s because carriers that own the tower assets would be hesitant to lease their best tower sites to their competitors and the competitors would be wary of this exact thing happening to them.
This is evidenced by the fact that average tenancy tends to increase after American Tower purchases tower assets from carriers. In 2012, American Tower purchased the Telefonica sites which only had a 1.1x average tenancy. Similarly, the Verizon sites that American Tower acquired in 2016 only had an average tenancy of 1.4x. This goes for Vodafone in 2017 at 1.5x and Eaton in 2019 at 1.5x. American Tower has an average tenancy greater than 2x.
Returns on capital?
Over the past 10 years, American Tower has spent 22% of its capital on capex and 78% on acquisitions. Within capex, the company spends most of its capital on discretionary capital projects (33%) and redevelopment projects (23%). Ground lease purchases, capital improvements and start-up capital projects make up the remainder of capex spend.
The unit economics of a tower build are quite good, especially compared to other real estate leasing projects. The construction costs for a tower range from $250k-$300k in the U.S. The tenant revenue is assumed to be $20k (with 3% annual escalators), operating expenses to be $12k, which implies a gross margin of $8k. The ROI for a single tenant tower in the U.S. is just 3%. However, with two tenants, the ROI increases to 13% and with three tenants, the ROI increases to 24%. This is because of the aforementioned operating leverage that a tower site has.
The returns internationally are even better than that of the U.S. because construction costs are much lower. Construction of a tower costs $85k-$170k in Latin America, $60k-$120k in Africa and $20k-$30k in Asia. As you can see from the chart below, the returns are much higher as well. It’s worth noting that there are higher risks associated with the international business. First, there is higher customer risk as we’ve seen in India over the past 3+ years (more on that near the end). Second, there is regulatory risk and third, there is foreign currency translation risk.
In terms of acquisitions, American Tower mostly acquires international tower assets, with the exception of the Verizon assets in FY16. For $5.1B, the company acquired 11k communications sites and locked Verizon into a 28 year deal at a rate of $1,900/month with 2% annual escalators. This deal made American Tower the largest communications site company within the U.S. and increased the company’s coverage of rural and suburban areas.
Historical deal valuations have ranged 10x-20x EBITDA, which is not ideal (since it implies a 5%-10% return), but we have to remember than tenancy tends to increase after acquisition. For example, in the recent acquisition of Telxius which was announced in January 2021, American Tower paid 25x EBITDA which would imply low 4% returns. However, this deal is a bit unique because it gives American Tower scale in certain European markets and it’s expected that average tenancy will increase from 1.3x.
We estimate that American Tower has returned 9%-12% on its capital spent over the past 5 years. Returns tend to be lower in years after a large acquisition is made, but in those years, the reinvestment rate is much higher, so the intrinsic value growth is relatively steady.
Similar to other REITs, debt has fueled a lot of the growth at American Tower. While the company is under-levered compared to its peers, high reinvestment rates have been achieved through debt financing. American Tower’s net debt/EBITDA has ranged from 3.4x to 6.4x over the past 10 years and net debt has increased from ~$4.5B in FY10 to almost $22.5B in FY19.
With the backdrop of increasing data traffic and evolving wireless standards, American Tower has ample opportunity to grow for the foreseeable future. The company continues to build out communications sites where its return hurdles are met and acquire existing towers from independent operators or carriers looking to monetize their tower assets through a sale.
Most of the growth going forward will be international as consumers are quickly moving up in mobile data standards from 2G to 3G/4G and soon 5G. This is because data consumption doesn’t trend backwards and many of these countries (India for example) don’t have established wired networks to fall back on. For many in these countries, wireless is the only way to access the Internet.
On top of international growth, American Tower (and the other tower companies) has the opportunity to increase tenancy, even in the most established markets. This is because as data traffic increases, more coverage is necessary within a certain geography to maintain the consumer experience. Carriers will eventually have establish denser coverage areas to provide 5G and eventually 6G wireless technology, which implies that average tenancy across towers should increase over time. Goldman estimates that it will take 4 years for American Tower to achieve tenancy of 3 in the U.S.
With returns on capital of 9%-12% and a reinvestment rate of 75%-180%, American Tower has increased its intrinsic value between 11%-12% over the past 5 years. As discussed above, the high levels of reinvestment comes from years when large acquisitions are made and this is typically funded with debt. The debt levels keep rising, but as more assets are acquired the EBITDA increases to keep its debt service ratios in-line with what’s reasonable.
What else is important?
Carrier consolidation in India
If you’ve followed American Tower, you’d know that growth and margins in Asia have been hampered over the past three years. This is because all of the company’s Asia exposure is in India and there has been significant carrier consolidation in the region. Reliance Jio (the largest carrier by number of subscribers) made heavy investments into its 4G network, while offering very compelling prices, so that many of the smaller wireless carriers just weren’t able to compete. In 2016, there were 13 wireless carriers operating in India and that number shrunk to 6 in just over a year.
For American Tower this meant increased churn for the leases on their towers, which resulted in negative organic tenants billings growth in both FY18 and FY19. FY20 will likely show a similar decline as another issue related to a recent ruling by the Indian Telecom authority. In FY19, the Indian Supreme Court ruled that wireless carriers are required to make certain payments that are past due related to use of spectrum and licenses, which means that capex spend for these carriers will likely be much lower than originally anticipated. This implies lower average tenancy growth.
However, the long-term trends in India are very positive. There are over 1.3 billion people and more people are crossing the poverty line each year. Usage of basic handsets on 2G technology is still prevalent but more consumers are upgrading to 4G/5G enabled smartphones. And like other developing regions, there isn’t a thoroughly built out wired network to fall back on.
Small cell vs macro cell
One of the debates related tower companies with respect to the transition from 4G to 5G wireless technology has been whether small cell or macro cell antennas would be the dominant access point used. Small cells are wireless antennas on lower architecture like rooftops of small buildings and signs/lampposts. Macro cells are wireless antennas on tower and rooftops of large buildings. American Tower thinks that the bulk of U.S. based 5G wireless will be delivered on macro cell towers because (1) 85% of the population lives in an area less than 5,000 people per square mile, (2) and it’s the most cost effective way for carriers to provide wide area 5G coverage to their customers.
Small cell antennas are likely to make more sense in dense urban areas where there needs to be better coverage in buildings and high traffic areas. Furthermore, higher frequency spectrum on small cells can provide faster data speeds but the signal doesn’t travel as far, making it better suited for denser environments.
American Tower doesn’t have a small cell strategy in the U.S. The company has stated that the risk adjusted returns for small cells in the U.S. are just not comparable to that of towers. Tower leasing typically produces low double digit returns on capital, but American Tower estimates that small cells in the U.S. are mid-single digits at best, with some projects below 5%. Small cell tenancy (remember that higher tenancy has a huge impact to margins) tends to be much lower than towers and competition is much higher, resulting in pricing pressure. Furthermore, the average contracts lengths tend to be much shorter.
T-Mobile/Sprint merger and Dish
With the merger between Sprint and T-Mobile complete, there are some moving parts in the U.S. based tower business. First, it looks like the combined entity will likely decommission most of the Sprint towers but continue to build out their 5G coverage. This will likely result in a net negative for the number of towers leased, but T-Mobile will continue to lease new towers as the company is required to deploy its 5G network that covers 97% of the U.S. population by 2023. T-Mobile signed a 15 year agreement with American Tower that will likely result in $17B of revenues during that period.
Dish is also in the works to become the 4th largest wireless carrier after the Boost Mobile acquisition and winning spectrum. Dish must deploy a nationwide 5G network covering 20% of the population by 2022 and 70% by 2023. Dish has signed an agreement to lease 20k tower sites from Crown Castle, likely because of the exposure to more dense urban populations, which can really move the needle towards hitting those coverage requirements.
Overall this may mean that the U.S. organic tenants billings growth number will be lower going forward, but American Tower seems well positioned to take advantage of the transition from 4G to 5G through its strategy of owning macro cell towers exposure to rural and suburban areas.
Going forward, American Tower will have the most optionality with large international acquisitions. The company just announced to acquire Telxius to become one of the top two tower companies in Europe and the leader in Germany and Spain. American Tower will continue to leverage the company’s balance sheet to make large acquisitions where the returns on capital meet their hurdle rates.
Data center assets
Interestingly, the two highest quality REITs, data centers and wireless towers may see convergence in the future. Equinix has mentioned that the company could potentially look to acquire tower assets where it makes sense to gain exposure to the mobile edge.
Similarly, American Tower, through its innovation program, is experimenting with data center assets. In FY19, American Tower acquired Colo ATL for $70M, or 15x EBITDA, as an entryway into the data center colocation market. It remains to be seen whether this becomes anything larger, but it’s interesting to see the potential convergence of these companies.
If you made it this far, I hope you received some value from reading our analysis. Please subscribe to the free newsletter and share with anyone that would find it valuable. Thank you for your support!