Tower Companies

(Q3 2024 Letter)

Tower Companies (American Tower Corp (AMT), Crown Castle Inc (CCI), SBA Communications Corp (SBAC))

Tower companies own and operate telecommunication towers that they then lease to wireless service providers for the buildout and maintenance of their wireless networks. With long term (10+ year) contracts with blue chip customers supported by annual price escalators (3%), tower companies are among the highest quality businesses out there. The industry structure is attractive, with local politics making new builds difficult, and wireless service providers preferring to lease space on towers as opposed to owning towers themselves because sharing the fixed costs provides significant cost savings. Tower companies benefit from meaningful operating leverage manifested in the first tenant covering the tower build cost and the second tenant providing near complete revenue to margin conversion. With ten-year leases and annual price

escalators, any equipment upgrades and incremental tenants directly adds to the earnings power.

In the third quarter we invested in the tower companies at near historic low valuations. We believe that the reason for the compressed valuations; an inflation and interest rate spike coinciding with the cadence of the wireless carriers 5G buildout, is temporary. We see three main paths to value creation at the tower companies.

One, the pullback in capital expenditures from the wireless carriers will reverse course and lead to high margin revenue growth for the tower companies. A perfect storm of an interest rate spike elevating financing costs, excess network capacity from the first phase of the 5G network rollout, and the lack of a “killer application” for 5G to act as a forcing function for subscriber and network competition gave the wireless carriers cover to pull back on their spending in favor of focusing on their balance sheets. Two years of curtailed capital expenditures are set to catch up to the wireless carriers as an accelerating treadmill of data consumption is set to force capital expenditures on the networks. This accelerating treadmill of data consumption is fed by both annual mobile data consumption growth of between twenty and thirty percent, and twelve million fixed wireless subscriber additions that use upwards of twenty times the bandwidth that typical wireless subscribers use. Coinciding with data usage growth acting as a forcing function on capital expenditure growth, wireless carriers have regulatory coverage requirements from past spectrum auctions. As wireless carriers are forced to add capacity through denser infrastructure build outs and expansion into rural areas not previously prioritized, tower companies benefit from high revenue to margin flow through from their ownership of the proverbial toll road on wireless network spending.

Two, tower companies are set to accrue high margin revenue growth as the wireless service provider industry shifts from a four to three player transition to a three to four player transition. Following the TMobile- Sprint merger in April of 2020, duplicative tower footprints were rationalized, causing uncharacteristic tenant churn for the tower companies. Coinciding with the ending of T-Mobile-Sprint tower churn in early 2025, a September merger and restructuring of Dish and Echostar has created an increasingly positive industry structure. With Echostar as a newly viable fourth player with valuable high propagation spectrum, we see a base case for increasing tower co-tenancy, and the potential for this technologically enabled upstart to force real network competition with the big three wireless carriers.

Three, the tower companies offer a free option on widespread adoption of artificial intelligence. Large technology companies are in the midst of an unprecedented amount of capital expenditures to enable artificial intelligence. While there has yet to be a “killer application” that causes widespread consumer and enterprise adoption of artificial intelligence, recent advancements have pushed the marginal cost of content creation near zero. As we have seen with past historical fact patterns from computing to the internet, when marginal costs go to zero, there is a creative explosion in use cases. Currently artificial intelligence is primarily expressed in text or photo mediums, which lack bandwidth intensity. As we have seen with social media,

when applications shift from text or photo to real time video, data usage explodes upwards. We could theorize ad infinitum about artificial intelligence creating customized high quality real time video streams, runtime libraries of data built into every application at the consumer and enterprise level, and exploding data sets used to train models and inference in real time, but the common theme is that there is a step change in the amount of data that needs to be moved, stored, computed, and moved again. With the mobile phone (and possibly Apple Vision Pro or Meta Orion) acting as the avenue for expression of artificial intelligence given its role as the remote control for the physical world, it is hard to see how exploding data usage will not necessitate lower latency, higher bandwidth wireless infrastructure networks. With wireless

service companies forced to compete based on their networks, the tower companies collect the toll from more equipment upgrades and increasing tower co-tenancies. Additionally, tower companies stand to benefit through their ability to use their secure, electrified, centralized land and space footprint to create their own latency sensitive artificial intelligence and cloud hubs for mobile edge computing. While this fact pattern remains firmly in our “free option” bucket, we are happy to own the tower companies for the two near term catalysts as we watch the third path to value creation unfold.