An insurance contract is presumed to require that the insured peril must be the proximate cause of the insured’s loss (see Incorporated General Insurance Ltd v A.R. Shooter trading as Shooter’s Fisheries 1987).

Causation involves two distinct enquiries namely factual causation and then legal causation.

The test for factual causation is generally described as the ‘but for’ test.

In the insurance context the enquiry is focused on the occurrence of the insured peril. But for the peril relied on and insured, would the loss event have occurred?

If factual causation is established the second part of the enquiry is whether the loss is sufficiently closely connected to the factual cause to be characterised as the legal cause of the loss.

In that regard, the yardstick in the insurance context is what the parties intended in the policy.

The court in Concord Insurance Co Limited v Oelofsen NO 1992 said that ‘effect must be given to the party’s own perception of causality’.

In an insurance contract, legal causation is assessed with reference to the standard laid down in the insurance contract. It is not a consideration of what public policy demands as in delict or criminal law.

So even if factual causation is established, an insured must also establish that the insured peril and the loss are sufficiently closely connected to satisfy the contractual standard that must be met to establish legal causation (see Napier NO v Collett & Another 1995).

In determining the proximate cause of the loss insured against, regard must be had to both factual and legal causation.

The proximate cause of the loss does not necessarily mean the latest cause in time but refers to the direct, dominant, operative and efficient cause.

Each case must be decided on its own wording and facts.