The lease escalator math you are describing is real, and the gap you are noticing is something analysts have been debating for a couple of years. Cell tower REITs like AMT and CCI carry contractual rent escalators, typically around 3 percent annually on domestic leases, baked into long-term agreements with carriers who have no practical exit. The bear case is that the market is pricing both names primarily on interest rate sensitivity, treating the dividend yield as a bond substitute and moving the share price with rate expectations rather than with the compounding value of those escalators. If that is right, the contractual rent growth is being discounted more than the fundamentals justify, and the spread you are seeing is at least partly real.
The things that complicate the model are worth naming directly. Small cell buildout costs are genuinely capital-heavy, and AMT in particular has been working through losses on its international portfolio that have pressured reported earnings. Churn assumptions matter because carrier consolidation has already produced some lease non-renewals, and the model is sensitive to what you assume about that going forward. The implied cap rate you are calculating also depends heavily on which earnings figure you use, FFO (funds from operations, the standard REIT profitability measure) versus AFFO (adjusted FFO, which strips out capital expenditure), and those two numbers can look quite different for tower companies.
I am not in a position to tell you whether your spread analysis is correct, and I would want to flag that anyone giving you a confident answer on individual REIT valuation is working from the same incomplete information you are.
For questions about how the escalator structure affects taxable income and how that flows through to the dividend, a tax professional familiar with REIT structures is the right call.
Which figure are you using as the earnings base when you back into the implied cap rate?
