Cell tower lease abstraction that reads the whole chain

The hard case

Why tower leases are the hardest abstraction case

Most commercial abstraction is one document per tenancy. A tower site is a bundle. On the cost side sits a ground lease with a landowner, or a rooftop agreement with a building owner when the structure stands on a roof. On the revenue side sit colocation leases with each carrier on the structure. The forms, counterparties and amendment clocks all differ.

The terms run long. The standard US ground lease runs a five-year initial term plus a stack of automatic five-year renewals — 25 to 30 years in total, often more — so which option period applies, and what rent is today, are arithmetic across the whole history, not a single signature page.

Escalations multiply across borders, and the regimes move. A fixed annual percentage — 3% is the long-standing US benchmark — sits next to RPI- or CPI-linked reviews on UK mast agreements, French rents indexed by statute to INSEE’s ILAT or ILC, and German clauses tied to the VPI under the Preisklauselgesetz, which bars upwards-only clauses. Since May 2026, France’s Economic Simplification Act lets leases collar ILC indexation symmetrically. MYX Analyzer supports 40+ languages, because the leases never arrive in one.

And every mature site drags an amendment chain, each document silently overriding the last; the economically true lease is the end state, which no single document states.

Generic schemas underperform on all of this. As MYX CEO Yavor Mihailov told Inside Towers: “We have a specific workflow for ground leases. We have a specific workflow for tenant leases.” For the broader document problem — drawings, diligence reports, legacy site files — see the telecom towers page.

Amendment chains

One row, one lease package

Take one site: a base ground lease, a first amendment extending the term, a second rewriting the rent exhibit, and a notice of assignment changing the payee. Ask for the current rent and each document alone answers wrong; only the chain’s end state is true.

MYX Analyzer treats the chain as the unit of work. A site’s documents load into one row of the grid as a single lease package — base lease plus amendments together — and the MYX Pro engine reads the package in one pass. Current monthly rent comes back as what the chain actually produces, with the document reference pointing at the controlling amendment and page, not the superseded clause.

Review follows the same logic. Confidence indicators steer attention to the values worth checking, the reference opens the exact page, and when a prompt was underspecified, rerun the cell, row or column — the package is re-read without restarting the project.

In an acquisition, this is the difference between sampling a data room and reading it — the rent roll gets rebuilt from the chains themselves. The tower due diligence page covers that use in depth.

Escalations

Escalations as data, not prose

An escalation clause abstracts poorly as a text blob. The Analyzer captures it as fields: escalation type, index, uplift over the index, collar, frequency, base period. A clause like INSEE CPI + 1.5% — a synthetic demo value, shown as an illustration — becomes data a rent model can consume.

Modeling comes next. The free lease escalation calculator applies official index series to a clause, to check one site’s trajectory before running the portfolio, and the rent calculation page covers turning extracted clauses into rent schedules at scale.