Reading a policy: deductibles, limits and exclusions

How a policy is put together

An insurance policy is not one document. It is a small stack that only means something in combination, and knowing which part is which removes most of the confusion. There are three parts.

The wording — labelled the policy booklet, terms, or product disclosure document — is the standard contract, identical for every customer on that product. It holds the definitions, the insuring clauses, the exclusions, the conditions and the claims procedure. It is long because it is generic: it describes every version of the product, including sections nobody bought.

The schedule — also called the certificate, statement of insurance, or cover summary — is the page or two specific to one contract. It carries the sums insured, the deductibles, the period of cover, the insured address or vehicle, any items listed individually, and codes for endorsements. It is short because it is only the variables.

Endorsements are amendments. They add cover, remove cover, or attach a requirement, and usually appear as codes on the schedule with the text held elsewhere. An endorsement overrides the standard wording wherever the two conflict, which is why a policy can look generous in the booklet and be narrower in practice.

What a deductible actually does

A deductible — the same thing as an excess in British usage — is the amount subtracted from a settlement before it is paid. It is not a fee and is not usually invoiced. It simply does not arrive. On a covered loss of $3,000 with a $500 deductible, the payment is $2,500.

The shape matters more than the figure, because it changes the arithmetic:

• Fixed amount. A flat figure per claim. Household and motor policies commonly show something in the range of $100 to $2,500, with voluntary top-ups taking it higher. That is an illustrative band, not a market average.
• Percentage deductible. A share of the insured value rather than of the loss. On a $400,000 sum insured, a 2% deductible is $8,000 whether the damage is $10,000 or $300,000. These are usually attached to named perils.
• Per-claim versus per-item. One storm damaging a roof, a fence and a wall may attract one deductible or three, depending on wording.
• Time deductibles. In income protection and business interruption cover the deductible is a waiting period in days rather than a sum of money.
• Stacked deductibles. Some schedules show a base deductible plus a peril-specific amount. Where the wording says they are cumulative, both apply to one loss.

The awkward interaction is with small claims. Because the deductible comes off the front of every settlement, there is a band of losses just above it where a claim returns very little, and a band below it where it returns nothing.

The deductible is proportionally brutal at the bottom and nearly irrelevant at the top: moving from $250 to $2,500 costs the whole claim at $1,500 and about 6% of it at $40,000. Raising it is therefore a decision about small losses. Insurers price it that way, and the premium reduction from moving up a band is typically modest — often a single-digit or low double-digit percentage, illustratively.

How limits stack in layers

A limit is the most the insurer will pay. Policies rarely have one. They have a hierarchy, and a claim can be cut by a limit several levels down while the headline figure sits untouched.

The outermost layer is the sum insured for each section — buildings, contents, liability — shown on the schedule. Inside it sit narrower caps:

• Per-claim limit. The cap on a single event. On liability sections this is often the only limit that matters, and it can dwarf the property sums insured.
• Annual aggregate. A cap on everything paid in one period of cover. A $30,000 claim against a $50,000 aggregate leaves $20,000 for the rest of the year, whatever the per-claim limit says.
• Per-item or single-article limit. A cap on any one object, applied to valuables not listed individually — commonly a flat figure in the low thousands, or 5% to 10% of the contents sum insured, both illustrative. A ring worth $9,000 inside a $60,000 contents sum insured can settle at a fraction of its value purely because it was never named.
• Category sub-limits. Totals for a class of property: jewellery in aggregate, cash, bicycles, contents in outbuildings. A claim can hit the category cap while the item cap and the sum insured are both intact.
• Inner limits on extras. Alternative accommodation, debris removal and professional fees are usually capped as a percentage of the main sum insured — illustratively 10% to 25% — rather than being open-ended.

Per-item and per-claim limits are often confused. A per-claim limit asks how much this event can cost the insurer. A per-item limit asks how much any single object inside it can contribute. A burglary taking six unlisted items can sit well inside the per-claim limit and still settle low, because the caps were added rather than the values.

Beneath all of it is the basis of settlement, which is not a limit but behaves like one. Replacement cost, or new-for-old, pays what a new equivalent costs. Indemnity, or actual cash value, deducts for age and wear first. The same sofa settles very differently under the two.

The main families of exclusion

Exclusions look arbitrary read one at a time. Grouped, they fall into a handful of recognisable families, and most policies carry a version of each.

• Gradual causes. Wear and tear, deterioration, rust, damp, settlement, insect or vermin damage. Insurance covers sudden and accidental events; slow decline is treated as maintenance.
• Inherent defect. Faulty design, workmanship or materials. Often the defective part is excluded while resulting damage to other property is covered, and the distinction turns on precise wording.
• Conduct. Deliberate acts, fraud, illegal activity, and the broad requirement to take reasonable care to prevent loss.
• Perils sold separately. Flood, earthquake, subsidence and similar. Their absence is not always in the exclusions list; sometimes the peril was never in the insuring clause at all.
• Systemic risks. War, terrorism, nuclear events, communicable disease, cyber events, government seizure — excluded as standard because they hit every policyholder at once.
• Known circumstances. Anything that had already happened, or was already anticipated, when cover started.
• Use and occupancy. Business use of a private item, letting to third parties, and property left unoccupied beyond a stated number of consecutive days — illustratively 30 to 60 — after which cover narrows automatically.
• Consequential loss. Lost income, lost value, inconvenience, and the cost of making a claim, unless a section buys them back.

Three structural points help. Exclusions appear at two levels: general exclusions cover the whole policy, section exclusions one part of it, so both are in play for any claim. Exclusions frequently contain an exception written back in — the phrase to look for is "this exclusion does not apply to" — and that write-back can restore most of what was removed. And a condition is not an exclusion: it is a requirement, such as fitting particular locks, whose breach can affect a claim even though no exclusion was triggered.

Why the schedule outranks the wording

The wording describes a product. The schedule describes the contract. Where they differ, the schedule and its endorsements normally govern, and that asymmetry produces most of the surprises.

Several things exist only on the schedule. The sums insured are there and nowhere else — no booklet can say whether a building is insured for $200,000 or $2,000,000. So are the deductibles, the period of cover, named drivers and locations, and individually listed valuables; an item absent from that list falls back to the single-article limit.

Endorsement codes are the part most easily skimmed. A line reading "subject to clause 14" is doing real work: clause 14 might raise a deductible for one peril, exclude a category of property, or impose a security requirement. A common pattern in disputes is a policyholder quoting accurate wording that was amended, on their own schedule, by a clause they never opened.

A reading order that finds problems

Reading a policy front to back buries the important parts in definitions. Working from the specific to the general surfaces problems faster:

1. The schedule in full, every endorsement code included, with the numbers written down.
2. The text of each endorsement named there, since these override anything read afterwards.
3. The insuring clause for the relevant section — the short paragraph stating what is covered.
4. The definitions of the few capitalised or bolded words in that clause. Defined terms rarely mean what they mean in ordinary speech; "accidental damage", "flood" and "contents" are all narrower than they sound.
5. The section exclusions, then the general exclusions, watching for write-backs.
6. The limits table, frequently a separate schedule of sub-limits rather than part of the wording.
7. The conditions and the claims procedure, including notification timeframes.

Anything still ambiguous after that sequence is genuinely ambiguous, and the insurer or a qualified professional who knows the situation is the place to resolve it rather than a second reading of the same paragraph.

What this means in practice

A policy is a filter with four stages, not a promise with fine print attached. The insuring clause decides whether a loss is in scope. The exclusions decide whether it stays in scope. The limits set the ceiling. The deductible sets the floor. Any single stage can reduce a settlement to nothing, and each is documented in a different place.

A few considerations follow from that structure, none specific to any one policy:

• The deductible governs the small end of the range, deciding which losses are worth claiming at all. Comparing quotes without comparing deductibles compares different products.
• Limits are layered, so the headline sum insured is a weak guide to what a realistic claim would pay. Sub-limits and single-article limits are where claims are most often cut.
• Naming valuable items on the schedule is the mechanism that lifts them above the single-article limit, and the schedule is the only place naming counts.
• Exclusions cluster into a few families. Recognising the family is faster than reading the list.
• Where a clause is unclear or a claim sits on a boundary, the reading is a matter for the insurer, an ombudsman scheme, or a qualified professional who knows the circumstances.

None of this requires legal training. It requires holding the schedule next to the wording, and adding up the numbers in the order the contract applies them.