Repair clauses are among the oldest provisions in commercial leasing and among the least precisely drafted. The structure is familiar: the landlord takes responsibility for structure, exterior and common parts; the tenant takes the interior of the demise and anything exclusively serving it. The language has barely changed in decades, and it was never written for the buildings it now governs.
Modern commercial property is full of components that do not sit neatly on either side of that line. A rooftop plant unit serving one floor. A chilled water loop shared between two tenants and the landlord's common area. A building management system that controls both demised and common space. Facade glazing that is structural, exterior and also the tenant's window.
None of these is a drafting failure exactly. They are the consequence of applying a two-category model to a building with hundreds of components, most of which did not exist when the model was devised. A lease form that handles a 1970s office adequately will struggle with a building whose services are integrated across the demise boundary by design. The practical effect is that the repair clause does not answer the question anyone actually asks, which is who pays for this specific thing that has just broken. Resolving that before it breaks is an abstraction exercise, and it belongs alongside the other fields captured during CAM reconciliation and lease audit work.
The five questions each element raises
For any given building element, a complete answer needs five pieces of information. Most lease abstractions capture one of them, which is a summary of the repair clause, and that single field is almost never enough to settle an actual question.
- Who repairs it, meaning who must carry out and fund work to restore the element to working condition when it fails.
- Who replaces it, which may be a different party, since an end-of-life replacement is not always a repair obligation at all.
- Is the cost recoverable, through service charge or otherwise, which is independent of who holds the obligation.
- Is it capital or operating, and if capital, whether it is excluded from recovery or amortised over a stated useful life.
- What standard applies, since an obligation to keep in repair differs materially from one to keep in good and substantial repair.
The replacement question is the one that produces the largest single exposures. A party obliged to keep an element in repair is generally not obliged to give back a better element than it received, and a component that has reached the end of its economic life without disrepair may fall outside the clause entirely. Leases that are silent on replacement leave a gap that becomes visible only when a major plant item reaches thirty years old. At that point the landlord argues that the tenant must keep the element in repair and the tenant argues that nothing is in disrepair, it is simply old, and both positions are arguable on the wording.
The cost of resolving an ambiguous element while nothing is broken is a short exchange of letters. The cost after a chiller fails is a negotiation about a six-figure number in which both sides read the same clause differently and neither is disinterested.
Elements worth separating explicitly
Certain components generate a disproportionate share of disputes because they genuinely sit across the boundary rather than because anyone drafted carelessly. Five are worth naming individually in any abstraction, whatever taxonomy is otherwise used, because the standard category language will not resolve them.
- Rooftop plant serving a single tenant. Located in landlord territory, serving demised space, and claimed by whichever party is not paying.
- Shared building systems. Risers, loops and controls serving both demised and common areas, where apportionment is rarely stated.
- Glazing and curtain walling. Simultaneously structure, exterior and part of the tenant's enclosure under most category language.
- Lifts and escalators. Usually landlord responsibility, but with modernisation programmes that sit awkwardly between repair and capital improvement.
- Supplementary cooling installed by the tenant. Tenant equipment in landlord plant space, with reinstatement duties that surface at lease end.
A five-step abstraction roadmap
Step 1, Build the element list once per building type
Create a standard element taxonomy covering structure, envelope, mechanical, electrical, vertical transportation and interior finishes. Reuse it across the portfolio so leases become comparable rather than each being abstracted into its own private vocabulary.
Step 2, Map each lease onto the taxonomy
Work through the repair, service charge and alterations provisions, recording the responsible party, replacement treatment, recoverability and standard for each element, with the clause reference. Expect gaps; recording them is the point of the exercise.
Step 3, Flag and triage the ambiguities
Sort the unresolved elements by replacement cost and remaining useful life. A silent clause on an element with twenty years of life left is a low priority. The same silence on a chiller approaching end of life needs resolving this year.
Step 4, Resolve the priority gaps in writing
Approach the counterparty on the high-priority items while the position is still abstract. A short side letter or recorded agreement is usually achievable, because neither side knows yet which way the exposure will fall.
Step 5, Connect the matrix to condition data
Link the responsibility matrix to asset condition and remaining-life information from the property team. The combination turns the matrix from a reference document into a forecast of who will be asked to spend what, and when.
Know who pays before it breaks
Element-level responsibility, replacement treatment and recoverability held per lease.
Why this matters more to the landlord side
Tenants typically hold a handful of leases and can afford to read each repair clause closely. Landlords and managing agents hold hundreds, frequently on several different standard forms acquired through portfolio transactions, and cannot. The result is that service charge budgets are built on assumptions about recoverability that are correct for the dominant lease form and wrong for a subset, which surfaces at reconciliation as a set of challenges with no quick answer. An element-level matrix across the portfolio turns that into a query, and it is the same control that makes vendor cost classification reliable in CRE vendor service contracts.
Elements and the questions to answer for each
The table shows a short extract from a typical matrix. A working version runs to several dozen rows and should follow the taxonomy used across the portfolio rather than the wording of any single lease.
| Element | Repair | Replacement | Recoverable |
|---|---|---|---|
| Roof membrane | Landlord | Often silent, check clause | Usually, subject to capital exclusion |
| Rooftop unit serving one tenant | Frequently ambiguous | Frequently ambiguous | Rarely, if exclusively serving |
| Shared chilled water loop | Landlord | Landlord, check amortisation | Usually, apportioned |
| Glazing to demised space | Split by lease form | Split by lease form | Depends on allocation |
| Tenant supplementary cooling | Tenant | Tenant | No, plus reinstatement at expiry |
The matrix is not glamorous work and it does not produce a saving anyone can point to in the quarter it is built. Its value appears later, in the disputes that do not happen, the service charge challenges that are answered in a day, and the capital plan that reflects what the leases actually say rather than what the property team assumed they said. For a portfolio of any size, the first full pass through the matrix usually identifies at least one element where the assumed recovery position has been wrong for years.
For related guidance, see Lease abstraction that holds up under ASC 842, ASC 842 and IFRS 16, the lease data auditors actually test and Property vendor service contracts, a controls datasheet.
