The performance test grades the operator against a comp set. On some deals, the owner owns part of that comp set — and the benchmark stops being independent.
An HMA performance test grades the operator on RevPAR Index — the subject hotel’s RevPAR against its comp set’s average. Clear the threshold, usually 85% to 92%, and the operator passes. The comp set is the benchmark, and measuring against it is sound: a relative benchmark is how you separate operator skill from a market-wide move.
Here’s the structural problem. The owner holds the subject hotel and another hotel in the same market. Both sit in the comp set. The party setting the grade owns part of the benchmark. The benchmark isn’t independent.
The Mechanism
RevPAR Index measures the subject’s RevPAR against the comp set’s average, as a percentage. The comp set average is the denominator. Raise the denominator, and the index falls — even if the subject’s own RevPAR didn’t move.
When the owner owns a hotel in that comp set, the owner influences one part of the benchmark: that hotel’s pricing, positioning, occupancy, and RevPAR. Not the whole comp set — just one piece of it. But that piece moves the average.
Take the case directly. The comp set includes a hotel the owner also holds. The owner’s comp hotel runs strong — higher ADR, stronger occupancy, corporate accounts the owner steers to that property. The comp average rises. The subject’s index falls. The operator’s test gets harder through the owner’s own asset.
An owner runs its portfolio to win. That’s normal business. An owner with several hotels in a market will price them competitively, position them for different segments, and move demand across properties to maximize total return. None of that is improper. It just shouldn’t also set the operator’s performance grade.
The Conflict
A fair performance test needs an independent benchmark. A benchmark partly owned by the grading party isn’t one. This is the informativeness principle in its plainest form: a performance measure should reflect the agent’s own contribution and filter out what the agent doesn’t control (Holmström, 1979). A benchmark the grading party partly owns fails that standard at the source.
And comp sets aren’t neutral to begin with. STR Analytics’ study of roughly 30,000 comp sets found hotels name each other reciprocally only 45.1% of the time. Comp-set membership is already subjective and asymmetric in practice — not the clean, mutual benchmark the test structure assumes. Put an owner-affiliated hotel inside that already-soft construct and the independence problem compounds.
The conflict holds regardless of intent. Even with no active steering — no deliberate demand allocation, no pricing moves aimed at lifting the average — the independence is compromised. The owner’s ordinary decisions about its own comp hotel move the operator’s test outcome. The grader can move the curve.
The owner could steer demand to its comp hotel. Whether it does is beside the point. What matters is that it can. This is Goodhart’s Law with the roles reversed: once the comp average is the number the operator is judged on, the party who can move that average has every reason to. And the performance test is already an owner-held option — the right to act on a failed test, analyzed here. Controlling part of the comp set lets the owner tilt whether the test fails at all. The trigger and the option sit on the same side of the table.
The Remedy
The time to fix this is bid stage, before signing. Comp-set composition is a recognized flashpoint in owner-operator disputes (ISHC). Once the HMA is signed, the comp set is locked, and the operator has no leverage to reopen it.
So the operator negotiates comp-set composition during the bid. The ask: exclude owner-affiliated or related-party hotels from the set. The benchmark should be independent third-party hotels only — hotels the owner doesn’t own, doesn’t operate, and doesn’t price. That benchmark is clean.
Where exclusion isn’t granted, the operator seeks an adjustment. One version: the related-party hotel stays in the set for market coverage but drops out of the index calculation. Another: it stays in, but the index threshold moves down to reflect the conflict. At a minimum, the operator negotiates disclosure of related-party holdings in the comp set. Disclosure doesn’t fix the conflict, but it surfaces it.
Whether a given related-party conflict breaches a good-faith obligation, or is actionable under contract law, is a question for counsel. The operator’s job is to spot the related-party hotel in the comp set and negotiate it out at bid stage.
The Benchmark Defines the Test
The benchmark defines the test. When the grader owns part of the benchmark, the test isn’t independent. The operator is measured against a standard the owner partly controls.
The move is to negotiate benchmark independence before signing. Exclude related-party hotels. Get the adjustment in writing. Make the benchmark clean while the operator still has leverage — because after signing, it’s gone.
So before signing, one question is worth running down: does the comp set the operator will be graded against include a hotel the owner also holds — and if it does, is that in the bid analysis, or buried in an exhibit no one modeled?
Dash Decisions provides vendor-side financial deal desk services for hotel operators bidding on HMAs.