Non-Proportional CoversCore

Per-Risk vs Per-Occurrence XoL

One risk, or one event

The distinction between excess-of-loss cover that attaches separately to each individual risk, versus cover that attaches once per event no matter how many risks it damages.

Open the interactive version → definition, quiz, structure diagram and progress tracking

Definition

Per-risk excess-of-loss protects a cedent against one severe loss to a single insured risk, testing each risk's own loss against the retention independently. Per-occurrence (or per-event) excess-of-loss instead aggregates all losses arising from a single event — a hurricane, an earthquake, a large liability incident touching many insureds — and tests that combined total against the retention and limit just once. The two use the same layer mechanics but answer a completely different question: how big was this one risk's loss, versus how big was this one event's total damage.

Worked example

A hurricane damages 500 insured properties. A per-risk XoL layer responds separately to any single property whose own loss exceeds the retention; a per-occurrence cat layer instead responds once to the combined total loss from all 500 properties together.

Scenario · figures in USD

The same storm, tested two different ways

A cedent holds a $1M xs $500,000 per-risk layer and a separate $50M xs $10M per-occurrence cat layer. A hurricane produces many property losses at once.
Largest single property loss from the storm$700,000
Per-risk layer test for that propertyExceeds $500,000 retention — layer responds
Combined losses from the storm across all affected properties$18M
Per-occurrence layer test for the whole stormExceeds $10M retention — layer responds once to the $8M excess
Key differencePer-risk tested one property in isolation; per-occurrence tested the entire storm's combined total
So whatThe same event can trigger both layers for entirely different reasons — one because a single risk's loss was large, the other because the event's combined total was large.

Check your understanding

What does a per-risk excess-of-loss layer test against its retention?

One individual risk's own loss, independent of what happened to other risks. Per-risk XoL looks at each individual risk's own loss in isolation — it does not care whether other risks were also damaged by the same event.

Why would a cedent buy both per-risk and per-occurrence excess-of-loss protection?

Per-risk protects against one severe single loss, while per-occurrence protects against an accumulation from one widespread event — different exposures entirely. The two covers protect against fundamentally different exposures — a single catastrophic risk loss versus an accumulation of many moderate losses from one event — so most cedents need both.

Word problem

An earthquake causes losses to 200 insured buildings. The single largest building loss is $600,000. The cedent's per-risk XoL retention is $500,000 and its per-occurrence cat layer retention is $15M. Combined losses across all 200 buildings total $22M. Which layer or layers respond, and to what amounts?

Show a hint
Test the single largest building loss against the per-risk retention, and separately test the combined total against the per-occurrence retention.
Reveal the worked answer
  1. Per-risk test: the largest single building loss is $600,000, which exceeds the $500,000 per-risk retention by $100,000 — the per-risk layer responds to that one building.
  2. Per-occurrence test: the combined total across all 200 buildings is $22M, which exceeds the $15M per-occurrence retention by $7M — the cat layer responds to the event as a whole.
  3. These are two separate, independent tests — the $600,000 building loss is also part of the $22M combined total, but each layer is tested on its own basis.
  4. Both layers can respond to the same underlying earthquake for entirely different reasons.
Both layers respond: the per-risk XoL pays $100,000 for the one building whose individual loss exceeded its $500,000 retention, and the per-occurrence cat layer separately pays $7M for the combined $22M event total exceeding its $15M retention — one testing a single risk, the other testing the whole event.

Related terms

Part of the Treaty Reinsurance guide, where this term is explained alongside every other treaty metric.