Insurance Maths

Covered versus uncovered claim

The choice is between two claims processes rather than between two prices. Both routes want photographs, both expect the packaging to be kept, and both run against a window that closes on a date nobody announces twice. What cover changes is who carries the loss and how much of it survives the paperwork. A decision taken on the premium alone ignores the half of the comparison that decides the outcome in the cases where the decision actually matters. Comparing two routes on the premium alone compares the visible half of the decision and calls it the whole of it.

Without cover the claim runs against the carrier's own liability, which is limited and documented rather than open-ended. Those terms usually cap what can be recovered, exclude damage that the packaging should have prevented, and treat cosmetic marking differently from damage that stops the item working. The desk models recovery on that route at about 60 per cent of the loss. That figure is an editorial estimate rather than a measurement, and it is labelled as one because the entire comparison below rests on it.

Cover changes who pays without changing what has to be proved. A premium band of around three per cent of declared value buys a claim against an insurer, and the claim still needs the same evidence plus whatever the policy adds: a damage report, photographs of the outer packaging before it is opened, sometimes a statement about how the parcel was handled after delivery. Some categories are excluded outright, and a parcel packed against the carrier's own requirements can fail a claim without any argument about the loss itself ever taking place.

Two features of that comparison drive everything below. The premium is a certain, small payment while the loss is an uncertain and occasionally very large one, which makes the decision an expected-value problem with a variance caveat attached to it. The desk therefore publishes the arithmetic and then states where the arithmetic stops being the answer: a single total loss on a high-value parcel does not care about an average, and no expected value describes how that outcome lands when it happens to the one parcel that mattered that month.

Itemised scoring

Four inputs drive the calculation, and each one comes from the desk's parameter layer with a check date attached. The declared value, which is also the ceiling on what any claim can recover. The premium band, quoted as a share of that value. The recorded loss and damage rate for the service level, which runs from 3.4 per cent on economy lanes across 23 recorded parcels, through 2.1 per cent on standard across 47, to 1.2 per cent on express across 52. And the share of a loss typically recovered without cover, modelled at about 60 per cent.

Premium arithmetic is a single multiplication. Three per cent of a 120 dollar declared value is 3.60, and the same band on a 400 dollar parcel is 12.00. Two details matter inside that line. The share applies to the declared value rather than to the item price on its own, so a declared figure that includes freight raises the premium along with it. And the declared value is the ceiling on the payout, so a parcel declared below its replacement cost can be covered on paper and still leave a real gap.

Expected value is the second half of the calculation. On a 120 dollar parcel moving on a standard lane, the recorded loss and damage rate of 2.1 per cent gives an expected loss near 2.52, and recovering 60 per cent of it leaves roughly 1.00 exposed. The same value on an economy lane lifts the exposed figure to about 1.63. Set those against a premium of 3.60 and the arithmetic favours declining cover on both, which is the result most buyers find counter-intuitive and the reason the variance caveat has to travel with the number wherever it is quoted.

Averages hide the shape of the risk, and the shape is what the money is really buying. A loss rate of two per cent describes 98 parcels in 100 losing nothing at all and a small number losing everything, so an expected value is a poor description of any individual order. Cover priced at three per cent is expensive against an average and cheap against a tail. The honest conclusion is that cover is bought for the tail rather than for the average, which means a buyer who can absorb the tail without changing plans is paying for something they do not need.

Scenario conclusions

Cheap goods on slow lanes produce the clearest answer. A low declared value, a premium charged as a share of it, and a claim process consuming hours of photographs and correspondence all point the same way: carry the risk. The recorded loss and damage rate is highest on exactly those lanes, which narrows the gap without closing it, because the recoverable amount is small enough that a fixed handling band plus a few hours of work exceeds whatever the claim would return.

Middle values produce the cases where the answer genuinely depends on the buyer. A parcel between 300 and 500 dollars sits where the premium is a real line in the budget and the exposure is uncomfortable but survivable. The deciding factor there is usually not the probability but the replacement question: whether the same item can be bought again at a similar price, and how long that would take. Where a replacement is quick and easy, cover is largely a convenience purchase; where it is not, the case for cover gets much stronger. Neither answer is wrong on its own terms, and the difference between them is what the buyer would do the following week if the parcel never arrived at all.

High-value items on fast lanes move the ratio the other way. Express lanes carry the lowest recorded loss and damage rate, which argues against cover, but the amount at stake is large enough that one event dominates every average in the table. On a parcel worth several hundred, a premium near three per cent is a rounding error against the cost of replacing the item out of pocket. The argument for cover in that case is a variance argument rather than an expected-value one, and it should be made in those terms rather than through the loss rate.

Consolidated parcels complicate the picture in a way that is easy to miss. Cover is normally bought against a consignment's declared total, so the premium scales with the whole box while a loss usually affects one item inside it. Splitting cover across items duplicates the premium for the same exposure. The question worth answering before paying is which value a claim would use, and whether the loss of one component in a mixed parcel would be assessed against the total or against the component.

Situations to decline cover

Declining cover makes sense wherever the premium exceeds the expected shortfall and the buyer can absorb a total loss without changing plans. The comparison above produces that answer for cheap items on every service level, and for mid-value items on lanes with a low recorded loss rate where a replacement is available at a similar price. The saving is real money inside the order budget, and it is not offset by any obligation a carrier imposes on a shipment that travels without cover.

Eligibility destroys a claim before evidence ever gets a hearing. Items outside the covered categories, parcels packed against the carrier's published packaging requirements, and goods already noted as damaged at the point of delivery are all cases where a premium buys a document rather than a remedy. Reading the exclusions before paying is the cheapest check available on this page, and it takes about as long as reading the premium figure, which is unfortunately the part most buyers read instead of everything else.

Process cost sometimes decides on its own. A claim consumes photographs, messages, a damage report in some cases, and a wait measured in weeks. On a low-value parcel that effort can exceed the amount at stake, and where the platform's own buyer route already produces a comparable outcome, paying for a second route buys duplication rather than protection. The desk prices the hours alongside the premium, because the hours are the one part of a loss that cannot be recovered from anybody afterwards. A claim that demands more attention than the item is worth is not protection, whatever the premium was called on the page.