Event-based aggregation

Category: Claims and policy principles · Reviewed by the Apex broking team · Last reviewed 2026-08-22 · ~6 min read

Where this entry sits

Aggregation wordings come in families, and the family chosen changes the answer more than any other drafting decision in the clause. For the general mechanics of aggregation and how it interacts with limits and deductibles, see the aggregation clause entry. For the professional indemnity treatment, including the “series of related matters or transactions” formulations common in that market, see aggregation of claims in professional indemnity.

This entry addresses one question: what happens when the aggregating word is event or occurrence.

What English law treats as an event

The starting point is that an event is a happening — something that occurs at a particular time, at a particular place, in a particular way. That is a demanding definition, and it excludes a good deal of what commercial parties intuitively want to aggregate.

In Caudle v Sharp [1995] LRLR 433 the underwriter’s failure to appreciate the exposure he was writing, resulting in the negligent writing of thirty-two separate reinsurance contracts, was held not to be one event: the court required a more definite happening of something at some time. A continuing state of blindness or ignorance was not an event.

The contrast was drawn sharply in Axa Reinsurance (UK) plc v Field [1996] 3 All ER 517, where the House of Lords distinguished “event” from “originating cause”. A cause is something altogether less constricted than an event: it can be a continuing state of affairs, and it can be the absence of something happening. That single distinction explains most disputes in this area. A wording keyed to originating cause aggregates widely; a wording keyed to event aggregates narrowly.

The four unities

Where a wording aggregates losses arising out of one event, the courts have used a working test drawn from the Dawson’s Field arbitration: the unities of cause, locality, time and the intentions of the human agents involved. The Court of Appeal approved that approach in Scott v Copenhagen Reinsurance (UK) Ltd [2003] EWCA Civ 688, while making clear that the unities are only a template and not the answer — the question remains one of construing the particular words in their commercial context.

Scott also addressed the causal link required by “arising from”: more than a weak causal connection is needed; there must be a significant causal link between the event and the losses said to arise from it.

In Mann v Lexington Insurance Co [2001] LRLR 179 losses arising from rioting spread across multiple locations and several days were held to lack sufficient unity to constitute one occurrence, notwithstanding a degree of central coordination. Geographic and temporal spread will usually defeat an event aggregator even where there is a common instigator.

What this means when the wording is drafted

The consequences run both ways, and which way an insured wants them to run depends entirely on the structure of the limits and the deductible.

Where the limit is expressed per event, wide aggregation reduces the cover available: many losses collapsed into one event share one limit. Where the deductible is expressed per event, wide aggregation reduces the number of deductibles the insured pays. On a policy with a substantial deductible and a generous limit, the insured may want aggregation; on a policy with a tight limit, it may not. This is why “does this clause aggregate?” is the wrong question and “what does aggregation do to my net position on the losses I actually expect?” is the right one.

Because event wordings aggregate narrowly, a series of similar losses with a common root cause — a repeated design flaw, a systemic process failure, a defective batch delivered over months — will usually not be one event. Where the commercial intention is that they should be treated together, the wording has to say so using cause or series language, not event language.

A related trap is mixing families within one policy. It is common to find an event aggregator in one section and an originating cause or series aggregator in another, or an event-based deductible sitting under a cause-based limit. That combination produces results neither party expected and is worth spotting at placement.

Reinsurance and the aggregation chain

Event aggregation matters most where excess of loss reinsurance sits above a direct book. The reinsurance is usually written per event or per occurrence, and whether the cedant can present a group of underlying claims as one event decides whether the reinsurance responds at all.

Much of the case law in this area is reinsurance case law for that reason, and the underlying wordings in the reported cases are reinsurance wordings. The construction principles carry across to direct policies, but the commercial pressure to aggregate or to disaggregate can point in opposite directions at different levels of the same programme. Buyers of large direct programmes with their own reinsurance arrangements should check that the aggregation language is consistent up the chain.

Why it matters

Aggregation is the least-read clause with the largest financial effect in a commercial policy. On a bad year it decides whether a series of losses is one claim with one deductible against one limit, or forty claims with forty deductibles against forty limits. Where the wording uses event or occurrence, the honest expectation is narrow aggregation, and any expectation of wider grouping needs to be built into the wording deliberately rather than hoped for at claim stage.

Frequently asked questions

What is the difference between aggregating by event and by cause?

An event is a happening at a particular time, place and in a particular way. A cause, as the House of Lords put it in Axa Reinsurance v Field, is altogether less constricted: it can be a continuing state of affairs or the absence of something happening. Cause wordings therefore aggregate far more widely than event wordings.

What are the four unities?

Cause, locality, time and the intentions of the human agents involved. The test comes from the Dawson's Field arbitration and was approved by the Court of Appeal in Scott v Copenhagen Reinsurance in 2003, which stressed that the unities are only a template and not the answer, the question remaining one of construing the actual words used.

Do a series of similar losses count as one event?

Usually not. In Caudle v Sharp the negligent writing of thirty-two separate contracts was held not to be one event, and in Mann v Lexington rioting across several locations and days lacked the unity needed for one occurrence. Where losses share a root cause rather than a happening, event wordings tend not to aggregate them.

Is aggregation good or bad for the policyholder?

It depends on the structure. Where the limit applies per event, aggregation reduces the cover available. Where the deductible applies per event, aggregation reduces how many deductibles are paid. The useful question is what aggregation does to your net position on the pattern of losses you actually expect.

Related entries

• Aggregation clause
• Aggregation of claims in professional indemnity
• Aggregate limit
• Proximate cause
• Deductible

This entry is part of the Apex Insurance Wiki. This entry is insurance information, not legal advice. It describes UK insurance law and market practice as at August 2026 and does not address the terms of any particular policy. Take advice on your own wording and your own facts before acting. Last reviewed 2026-08-22. Next review: 2027-02-22.

Apex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority (FRN 724952). This page is general information, not advice on a specific policy.