This is general education rather than legal or tax advice. A building sale asks two insurance policies to meet at a single instant. Nothing in the transaction assigns that job to anyone, so it is the piece that most often goes wrong — and both failure modes are avoidable in a phone call.
Two policies, one building, one instant
The mechanics are simpler than the coordination. On one side is a policy the seller has held, insuring the seller’s interest. On the other is a policy the buyer has yet to buy, insuring an interest that does not exist until the transfer completes.
Neither document knows about the other, and nothing in either one is triggered by a closing. The seller’s coverage runs until the seller ends it. The buyer’s runs from whenever the buyer told the broker to start it. If those two instants are not the same instant, one of two things is true: the building is covered twice, or it is covered by nobody.
The first situation costs a little money. The second is the one worth an hour of planning.
The seller should let it run, not cancel in advance
The single most common seller-side error is canceling on the strength of a date in a contract.
Closings move. Funding is delayed, a lien turns up, a signature is missing, the recording office is closed. A seller who instructed the broker to terminate coverage effective on the scheduled day, and then does not close that day, owns an uninsured commercial building and generally does not realize it until the following week.
The safe instruction is conditional rather than calendared: cancel on written confirmation that the transaction actually completed, with the termination effective at that moment where the carrier will permit backdating. Say it in writing to the broker, and say who will send the confirmation.
Return premium is arithmetic, and it comes in two shapes
Canceling part-way through a term normally produces money back, and it is worth knowing which calculation your policy uses before you count on a figure.
A pro rata return gives back the unearned portion as written. A short rate return keeps something additional to reflect what the insurer spent putting the policy on the books, so the refund is smaller than the calendar alone suggests. Which one applies is a term of the contract rather than a matter for negotiation, and it is stated in the cancellation condition.
Two practical notes. The refund is processed after cancellation is confirmed rather than at the closing table, so nobody should be treating it as closing funds. And where the building sat inside a schedule with other properties, removing it is an endorsement rather than a cancellation, which is a different calculation and a different conversation.
The buyer binds to a moment, and the lender wants proof of it
On the other side, the buyer’s job is to have coverage attach at the same instant risk transfers rather than on the day it does.
Funding, recording and the handover of keys frequently occur at different hours, and a binder written to a date leaves the earliest of them ambiguous. The buyer should name the instant. The full submission sequence that gets a buyer to that point — and why it starts long before closing week — is the insurance onboarding sequence for a first building.
The lender then wants evidence in its own format, and the provision that gives it standing under the policy is an endorsement rather than a line on a summary form. What that document proves is narrower than most parties assume — see the certificate versus the policy. Where the financing is a note carried by the seller, the whole arrangement is thinner than a bank’s, and what each kind of lender wants sets out why.
The bare instant at midnight
Here is the failure itself, stripped to its shape.
The seller’s broker is told the sale completes on a given day and cancels effective that morning. The buyer’s broker is told the same thing and binds effective the following day, because the buyer will not have the keys until the afternoon. Between those two instructions sits an unbroken stretch during which the building has an owner, has tenants, has a boiler and has nobody insuring it.
Nothing in the transaction reveals this. Both brokers did what they were told, both files look complete, and the only event that surfaces the gap is a loss inside it.
The fix is not clever. Somebody has to hold both instructions in view at the same time and check that the end of one is not later than the start of the other.
Real-World Scenario: A seller disposes of a small multi-tenant building and instructs the broker to cancel as of the closing date, expecting the refund to arrive with the sale proceeds. The buyer, whose lender is funding in the afternoon, asks for coverage to attach on receipt of the wire. Overnight, before the wire, a supply line lets go on an upper floor and water travels into two occupied units. The seller still owns the building. The seller’s policy has already been terminated at the start of that day. The buyer’s has not yet begun and would not have responded anyway, because the buyer had no interest in the building when the damage occurred. Two competent brokers, two correct instructions, and a stretch of hours nobody was assigned to think about.
Overlap is the cheap error, and the forms already handle it
Given a choice between a gap and a duplication, choose the duplication every time.
Property forms contain a condition addressing what happens when other coverage applies to the same loss, and it exists precisely so that two policies covering one event become an allocation problem for the insurers rather than a coverage problem for the owner. A deliberate overlap therefore costs an unearned slice of premium and nothing else.
Two things make the overlap work as intended. Both carriers should be told, because a claim presented to two insurers that each believed themselves to be the only one on the risk is a slower claim. And the overlap should be short and intentional — a stated window either side of the transfer — rather than an indefinite state nobody closes.
What travels with the building, and what does not
Almost nothing travels, and the exceptions are the ones worth asking about specifically.
The policy does not travel. The claims history does, in the sense that it stays attached to the address and will be read by the buyer’s markets at the first renewal whether or not anyone disclosed it — the loss run is that document, and reading a commercial building’s insurance history is how a buyer should have obtained it during diligence.
Flood coverage written through the federal program follows its own transfer rules rather than the ordinary ones, and it is a question to raise early rather than a detail to discover at funding; the National Flood Insurance Program publishes the consumer-facing material, and the mapped designation for the parcel is public at the FEMA flood map service.
The tenant evidence file should travel as a matter of contract, because rebuilding it from scratch after closing means chasing people who have no relationship with the new owner yet. And the entity name changes, which is not a detail — the policy protects the party printed on it, a subject worked through in named insured versus additional insured.
The seller’s exposure does not end at the wire
This is the part sellers are most surprised by, and it argues against clearing the file.
Liability written on an occurrence basis responds to bodily injury or property damage taking place during its own policy period, regardless of when a claim eventually arrives. A slip in a common area during the seller’s ownership remains the seller’s policy’s problem long after the building has changed hands, and the general liability page describes what that coverage answers.
So keep the declarations, the forms and endorsements schedule, and the claims record. Note the carrier, the policy numbers and the periods somewhere durable. And where the sale is of an entity rather than of a building, the position is different again and belongs with your attorney rather than with a general article.
If the building sits empty between owners
A sale that completes with units already dark introduces a second clock, and it belongs to the buyer immediately.
Vacancy provisions measure the state of the building rather than anyone’s intentions for it, and a stretch of emptiness that began under the seller does not reset because title moved. Raise it in the submission rather than discovering it at a claim — the vacancy clause and when it starts running explains what the wording actually measures, and the income consequences sit with business income and loss of rents.
The cutover, party by party
For the seller: instruct the broker conditionally rather than by date, confirm the transfer in writing before anything is canceled, understand which return-premium basis applies, and keep the policy record rather than shredding it.
For the buyer: bind to an instant, confirm the seller’s coverage runs to at least that instant, get the lender its evidence in the format asked for, and collect the tenant file while the seller is still motivated.
For whoever is coordinating: hold both instructions side by side and check the seam. Then decide, before closing week, who tells whom when the date moves — because it will. Once the building is yours, the standing question of who handles evidence, renewals and first notice is answered in what your property manager should do about insurance, and the underlying form is described on the commercial property page.
Cancellation mechanics and refund practice are governed where the policy is written, so the regulator to ask is your own; the NAIC directory will find it. Wider reading on the lines themselves sits with the Insurance Information Institute.
If you are on either side of a transfer and want the seam checked before the closing week, send us the closing date, the current declarations and the lender’s requirements and we will tell you where the two instructions do not meet.