Business Interruption Cannot Be Calculated from the Property-Loss Timestamp

The damage date is a useful fact.

It is not a calculation.

A fire at an insured premises starts a property claim, and, if the policy responds, a business interruption claim.

The shared cause is real enough.

The urge to give them one timeline, one reserve view, one closing event is understandable.

It is also wrong.

The property loss triggers the BI claim. It is not the number that settles it.

We ask what happened to building, equipment, and stock. Whether the peril is insured. How the damage is valued. Which limit applies.

BI asks something else entirely.

What the business would have earned. Which costs stopped. How long a competent restoration takes. What earnings survived. Whether a supplier problem sits inside an extension.

Those questions meet at one claim event.

We should not fuse them into one calculation because a claims screen prefers a tidy story. None of which argues for two rival claims platforms.

It argues against a false shared state in our systems.

A repair can be complete while the period of indemnity is still open. The period can end while the business has not got its customers back. A property payment can be final while the accountants are still arguing about the sales projection. Our system should be able to say each of those without treating any of them as a contradiction.

The trigger is deliberately narrow, and the door is physical

A BI claim clears four gates, in order.

An insured peril has to operate.

It has to cause direct physical loss or damage. The property has to be insured property at a described premises, unless an extension reaches out. And the damage has to cause a necessary suspension of operations, with the loss flowing from it. That sequence is not a workflow an adjuster invented.

It is the coverage grant we wrote.

Revenue can vanish for good commercial reasons and never enter the calculation.

Physical damage isn’t a technicality at the edge.

It is the front door of the product.

The pandemic demonstrated that brutally.

The March 2020 collapse in foodservice did not by itself make a property-linked BI claim. The UK Supreme Court’s FCA test case concerned non-damage disease and prevention-of-access extensions. It didn’t turn a physical-damage form into a revenue guarantee. Canadians who inherit American civil-authority shorthand are exposed here. The issued wording, its stated radius, and its own time cap are the rules that matter.

Civil authority is not an escape hatch.

It needs covered damage to somebody else’s property, an order prohibiting access, and it pays for a capped period.

Ingress and egress has a different access condition.

Service interruption needs damage to utility property, and transmission lines are commonly excluded unless bought back.

Those are separate grants.

They are not friendly labels for one event.

The same discipline covers contingent BI.

A supplier, customer, leader, or attraction property has to suffer damage the policy insures. Many wordings name the dependent property, and set a sublimit below the main BI limit.

A supplier’s outage is central to our insured.

It still isn’t our insured’s property loss.

The model needs the dependency, the extension, the supplier evidence, and the sublimit.

“Related to loss” is not enough.

Property has a valuation problem, and BI has a counterfactual one

Commercial property schedules in Canada commonly split the risk into Building, Equipment, and Stock.

Already a warning against a generic loss field.

A building settles on actual cash value, or on replacement cost subject to repair with due diligence and dispatch. Stock values at cost to replace rather than selling price, unless a selling-price clause says otherwise. Equipment breakdown needs its own form where the property wording excludes the mechanical event.

The property limit structure has real consequences too.

Under a scheduled policy, each location and item has its own ceiling and coinsurance test.

Under a blanket policy, values test in aggregate.

The signed statement of values still matters, and a margin clause caps a location just above its reported value. The property coinsurance bargain wants eighty to a hundred per cent of value at the time of loss.

A formula for a covered property loss.

It is not a BI forecast.

None of it tells us what the business would have earned while the plant stood still.

BI asks us for a counterfactual reconstruction.

Canadian gross earnings forms start with net sales plus other earnings. Then they deduct merchandise sold and materials consumed in production. So payroll and fixed overhead stay inside by default.

That is a financial definition.

Not a property valuation basis in a different hat.

The trends clause then makes the task harder.

The projection reflects the pre-loss trend, and the conditions that would have hit the business anyway. A newly won contract. A lost customer. A market decline. Sales preserved through a third-party producer.

Each changes the answer.

A property estimate can be excellent and irrelevant here.

We treat the form family as a visible input, not an inference.

IBC publishes Canadian Gross Earnings and Profits forms.

A proprietary package can depart from either, and layered placements may use CP 00 30 or a London wording.

Those names are not filing furniture.

They identify different clocks and different measurement bases.

A migration therefore needs the issued form and edition before mapping a field.

I would keep the inputs visibly separate.

Property valuation basis and insured value. BI form family. Selected limit. Historical accounts. Trend assumptions. Expense treatment. Actual sales during interruption. Mitigation evidence. A system deriving the last five from a repair-completion date has confused an event with an explanation.

There are three clocks here, and only one of them is the repair

The first clock we run is physical work.

It records emergency mitigation, scope, repair, holdback, and the return of damaged assets. A replacement-cost settlement advances actual cash value first, then releases the difference as repairs proceed.

Ontario fire statutory conditions make payment due sixty days after a completed proof of loss.

Quebec follows the Civil Code of Québec instead.

Those are our property controls, and they stay provincial.

Commercial BI wording isn’t filed for approval the way personal auto is.

OSFI, FSRA, and the AMF hold prudential or conduct roles.

None supplies a pre-issue calculation for a manuscript placement.

The issued policy, the endorsements, and the provincial overlay are the evidence.

We cannot fill that absence with a default rule and call the result Canadian.

The second clock we run is the BI waiting period.

It is a time deductible, a day to three.

Not a dollar deductible copied off the property file.

Some wordings work on a franchise basis, paying from hour zero once it crosses.

Subtract the property deductible from BI and we changed the mechanism.

So has treating the waiting period as the first unpaid slice of a long loss.

The third clock is the period of indemnity we owe.

Under a gross earnings form it starts at damage and ends when the property could have been repaired or replaced with reasonable speed and similar quality.

It is a restoration period we assume, not one we observe.

The business may wait, argue with a contractor, never rebuild.

Neither lengthens the period.

Partial resumption reduces the loss without ending the period.

This is where a tidy screen turns dangerous.

“Repair complete” means clock one reached a state.

It does not establish the second, and it does not necessarily establish the third.

An extended period of indemnity can cover the recovery ramp for a month, a quarter, or a year after restoration.

Without one, sales that haven’t returned sit outside gross earnings coverage while the building stands newly repaired.

The form family moves the clock again.

A profits form may use an indemnity period selected at placement, one to three years from damage, whatever the repair date.

That selection isn’t an adjuster’s discretion, and it isn’t a field a property workflow can infer.

Ordinary payroll may also be limited to a stated number of days unless bought back in full.

The policy text is a first-class input.

The two coinsurance clauses are not cousins

Property coinsurance looks backward, to value at the time of loss.

It tests the carried limit against the required percentage of building, equipment, or stock value.

A scheduled item can fail while the account looks adequate.

A blanket programme can look generous and be constrained by its statement of values.

Gross earnings contribution looks at a different exposure.

It measures the BI limit against actual gross earnings for the twelve months after inception.

The result can reduce an earnings recovery even where the BI limit is never exhausted.

Not an intuitive outcome.

Which is exactly why it should be a separately versioned calculation, and not a property subroutine.

The two clauses share a family resemblance the way a building estimate and an income statement do.

Both can be wrong expensively.

They use different declarations, dates, evidence, and reviewers.

Combine them into one “underinsurance” flag and the dashboard gets pleasant and the control goes useless.

Give BI its own timeline, even when the interface wants one card

A useful BI file does not wait for every issue to resolve before it becomes legible.

We can appoint an adjuster and a forensic accountant while origin and cause continues under a reservation of rights.

The insured may retain an accountant too.

The live disputes are rarely about one magic number.

They are the projected sales, the expenses that genuinely ceased, and the reasonable restoration period.

We keep those inputs as inputs, with sources and versions.

Expenses to reduce the loss are recoverable only up to the loss they actually avoid.

So a third-party production arrangement can be sensible mitigation without becoming a blank cheque.

Advances bridge the gap before settlement, once both sides agree a floor.

Neither belongs inside a property payment ledger.

Most forms also carry a claim-preparation sublimit for the insured’s accounting work.

Another quiet sign the property scope was never meant to answer the earnings claim alone.

The file needs room for the insured’s financial evidence and our challenge to it.

Not a document attachment against the repair reserve.

So link the files by peril and location.

Don’t collapse their status into one green tick.

The board conversation isn’t about closing files faster.

It is whether we can tell repaired property from restored operations, and restored operations from recovered earnings.

Three different statements. Three different evidentiary burdens.

That distinction is our protection against false closure.

A repaired building can still be a stopped business.

We should see the difference before the claimant does.