TIP826: American Tower (AMT): The Wide Moat Business Your Phone Can't Live Without w/ Kyle Grieve & Shawn O'Malley

American Tower participated in one of the most remarkable long-term holding stories in recent investing history. Chuck Acre bought shares at the 1998 IPO at roughly 80 cents and achieved approximately 28,000% returns, a 280x multi-bagger, before reducing his position to 0.14% of his portfolio around 2020. The stock is now in a nearly 40% drawdown from its 2021 peak, making it the first REIT covered on the show and a timely subject for reexamination.

The core competitive advantage is deceptively simple. American Tower owns the specific land parcels beneath its nearly 150,000 towers globally, and no competitor can replicate those exact locations. Zoning restrictions and surrounding residential development make assembling an equivalent portfolio practically impossible, a dynamic Kyle Grieve compares to Copart's land holdings. A single new tower costs between $250,000 and $350,000 to build, and replicating AMT's portfolio would require billions of dollars and decades of effort.

The tower economics illustrate extreme operating leverage. With one tenant, a tower earns roughly $20,000 in revenue against $12,000 in operating expenses, producing a 40% gross margin and a 3% return on investment. With three tenants, revenue rises to approximately $80,000 while operating expenses increase only to $14,000, pushing gross margins to 83% and ROI to 24%. Additional tenants pay a 50% premium over the anchor rate. Leases are non-cancellable for five to ten years, US contracts carry a fixed 3% annual escalation clause, and international escalators are tied to inflation indexes. The 2025 churn rate is approximately 2%, but Grieve notes the 1% margin between the escalator and churn is slim enough to make churn a meaningful ongoing risk.

The primary churn driver has been carrier consolidation rather than ordinary customer dissatisfaction. In India, the Vodafone-Idea merger, Reliance Jio's disruptive entry, and the collapse of Tata Tele Services caused severe churn, and AMT fully exited India in 2024. In the United States, T-Mobile's absorption of Sprint between 2021 and 2024 wound down redundant leases over multiple years under a negotiated master lease agreement. Grieve notes that even through these large events, AMT's overall churn remained manageable. Switching costs reinforce the moat because a carrier decommissioning an AMT site must reinstall equipment elsewhere, retest coverage, and absorb downtime, while AT&T, Verizon, and T-Mobile collectively generate over a quarter trillion dollars in revenue, making tower lease costs a rounding error that reduces price sensitivity further.

The balance sheet has gotten significantly uglier over the years. AMT carries approximately $37.3 billion in debt against roughly $7.2 billion in adjusted EBITDA, a net leverage ratio of about five times compared to three times in 2017. Grieve argues this is less alarming than it appears because the majority of debt extends to 2051, the weighted average interest rate is only 3.5%, and the company holds $54 billion in non-cancellable future lease obligations from customers. Revenue grew approximately 52% over the same period leverage nearly doubled, and AMT generates more than $5 billion per year in operating cash flow. Shawn O'Malley adds that borrowing at low long-term rates and redeploying capital at higher returns can be accretive to shareholders, but also that heavy debt makes equity value less robust because creditors have first claim on assets in a distress scenario.

The two largest recent acquisitions raise questions about capital allocation discipline. The Telxius acquisition cost $9.6 billion and doubled AMT's European footprint. The CoreSight data center acquisition cost $10.4 billion at a 27 times EBITDA entry multiple, added about $2.5 billion in goodwill, and was completed just before interest rates rose sharply. Since acquisition, CoreSight revenue has grown at roughly an 8% CAGR and operating margins have expanded from 46% to 53%, but Grieve argues that growth rate does not justify the entry multiple. The data center segment now represents about 10.5% of total revenue and is currently valued at roughly 17 times EBITDA.

AMT's stock has compounded at approximately 6.3% per year since 2016, and with a current dividend yield of roughly 3.7% the total estimated return is around 10%, consistent with the company's average ROIC of approximately 10% over the last decade. As a REIT, AMT avoids the standard 21% federal corporate income tax on distributed income but must pay out at least 90% of REIT taxable income, forcing reliance on outside financing for growth. Since 2016 the dividend has grown at 15% per year, though buybacks totaling only about $1.7 billion over that period are, in Grieve's words, quite small. AFFO represents a ceiling on recurring cash distributable to common shareholders rather than the actual dividend paid, and AMT depreciates and amortizes about half a billion dollars per quarter, which inflates adjusted EBITDA relative to free cash flow.