Escalation Floors and Caps: Who They Actually Protect

Most CPI escalation clauses are bounded. A floor protects the landlord; a cap protects the tenant. Together they convert an index-linked escalation into something that behaves like a fixed one most of the time.

The floor

A floor sets a minimum annual increase regardless of what the index does. A 2% floor means rent rises at least 2% even if CPI comes in at 0.4%.

This protects the landlord. It guarantees nominal growth and eliminates the deflation scenario entirely.

Floors are near-universal, and tenants rarely resist them hard, because a period of genuine deflation is rare enough that the floor feels theoretical at signing. In practice, floors bind more often than expected — CPI came in below 2% in a majority of years between 2009 and 2020.

The cap

A cap sets a maximum. A 4% cap means rent rises no more than 4% even if CPI runs at 7%.

This protects the tenant. It bounds the worst case and makes long-term occupancy cost predictable.

Caps are the more heavily negotiated of the two, and their value became extremely visible in 2021-2023. Tenants with 3% caps in an environment where CPI exceeded 7% were substantially protected. Landlords who had granted those caps recovered far less than inflation for several consecutive years — and because escalations compound, that shortfall persists for the remainder of the term.

Neither is free

Both are priced into the base rate, whether or not anyone says so explicitly.

A landlord granting a tight cap should be compensated with a higher base rate or a higher floor. A tenant accepting a high floor should get a correspondingly lower cap or a lower starting rate.

Where this goes wrong is when one side treats a bound as boilerplate. A cap agreed without adjusting the base rate is a concession given away, and it can be worth several percent of total rent over a long term.

The narrow band problem

A 2.5% floor with a 3.5% cap is administratively a CPI clause and economically a 3% fixed escalation.

Consider what that costs to run. Every year, someone must identify the correct index series, find the right comparison periods, compute the change, apply the bounds, apply the rounding rule, and record the inputs. In the overwhelming majority of years, the result will be 2.5% or 3.5% — one of the two bounds.

If the band is that tight, take the fixed escalation. You get the same economics without the annual lookup and without the ambiguity about which index and which periods.

A band is worth administering when it is wide enough that the index genuinely determines the outcome most years. A 2% floor with a 6% cap is a real CPI clause. A 2.5% to 3.5% band is a fixed escalation wearing a costume.

Cumulative bounds

Some clauses make the cap cumulative, letting unused headroom carry forward — the same mechanism as a cumulative CAM cap.

Under a 4% cumulative cap, a year where CPI came in at 2% banks 2% of headroom. A subsequent year with 7% CPI can then bill up to 6%.

This substantially weakens the cap protection over a long term, because low-inflation years fund high-inflation ones. Tenants should understand that a cumulative cap is meaningfully less protective than a non-cumulative one of the same rate, and price it accordingly.

Recording them

Floor, cap, cumulative behavior, and rounding all belong in the lease abstract as structured fields — not as a note saying “CPI with floor and cap.”

Every one of them changes the number, and every one of them will be questioned eventually. See how escalation terms are modeled and applied.