Occurrence Limits With Multiple Claims

Why Multiple Claims Change the Limit Calculation

A single event can produce several covered claims. A storm might damage multiple insured buildings, or one accident might injure several people. If the policy has a limit per occurrence, the insurer’s total payment for all covered claims arising from that event is capped collectively. It is not generally correct to apply the occurrence limit separately to every claim and then add the results.

The key calculation is therefore at the event level: first determine the total amount that would be payable for the claims from an occurrence under any applicable claim-level terms, then apply the shared occurrence cap. This distinction matters when calculating expected losses and increased limits factors (ILFs). A claim-level ILF describes how a limit changes payment for an individual claim; it does not, by itself, describe the effect of a cap shared across claims.

Definition: An occurrence is the policy-defined event or cause that groups claims for purposes of a shared limit. For one occurrence with $N$ claims and ground-up claim losses $X_1,\ldots,X_N$, write $S=\sum_{j=1}^{N}X_j$. With an occurrence limit $U$, no deductible, and no separate per-claim limits, the total payment for that occurrence is $Y_U^{\mathrm{occ}}=\min(S,U)$. The limit applies once to the combined covered loss from that occurrence.

The formula assumes all the losses are covered, are assigned to the same occurrence under the policy wording, and are measured on the same loss basis. It does not account for deductibles, coinsurance, separate per-claim limits, or exclusions. If a per-claim limit $c$ also applies, and the contract applies both limits to the same claims, the payment is typically $\min\bigl(\sum_{j=1}^{N}\min(X_j,c),U\bigr)$. Always follow the specified policy terms; a different contract structure can produce a different calculation.

For an annual portfolio with multiple occurrences, calculate the payment for each occurrence using its own shared limit and then add those occurrence payments. A per-occurrence limit is not automatically an annual aggregate limit.

Calculating a Shared Cap

For a known set of claims, the method is direct: add the covered amounts subject to any claim-level terms, then cap that combined amount at the occurrence limit. The order prevents the same occurrence cap from being mistakenly reused for each claim.

Worked Example: Applying One Limit to Three Claims

A hypothetical liability event produces three covered claims of $38,000, $47,000, and $29,000. The policy has a $100,000 occurrence limit, with no deductible or separate per-claim limit. Find the insurer’s total payment for the occurrence. Treat the claim amounts as deterministic point masses at the stated values; no interest or mortality assumption is relevant.

Step 1: Add the claim amounts.

$$ S=38{,}000+47{,}000+29{,}000=114{,}000. $$

Step 2: Apply the occurrence limit once. The total payment is the smaller of the combined loss and $100,000:

Y_{100{,}000}^{\mathrm{occ}}=\min(114{,}000,100{,}000)=100{,}000.

The insurer pays $100,000 in total. The amount above the cap is $114,000 − $100,000 = $14,000. Method and mistake avoided: aggregate the claims before applying the shared limit. Applying the $100,000 cap to each claim separately would incorrectly produce a payment of $114,000.

When claim amounts are random, the same cap is applied to the random total. The expected payment is $E[\min(S,U)]$, not generally the sum of the expected payments obtained by applying $U$ to each claim separately. The cap operates on the combined amount, so the distribution of the total matters.

Result: For claims grouped into one occurrence, the expected payment under a shared limit $U$ is $E[Y_U^{\mathrm{occ}}]=E[\min(S,U)]$, where $S$ is the total covered loss subject to any applicable claim-level terms. The distribution of $S$, including dependence among claims when relevant, is needed to evaluate this expectation.

Why the Expected-Loss Formula Holds

The formula follows by evaluating the contract payment in each possible outcome. If the occurrence total is below the limit, the insurer pays that total; if it exceeds the limit, the insurer pays the limit. Thus the payment in every outcome is exactly the smaller of $S$ and $U$. Taking expectations gives the expected payment. For a discrete total loss with possible values $s_k$ and probabilities $p_k$, this becomes:

E[Y_U^{\mathrm{occ}}] =\sum_k p_k\min(s_k,U), \qquad \sum_k p_k=1.

This calculation also explains why the expected ground-up total alone is not enough: in general, knowing $E[S]$ does not determine $E[\min(S,U)]$. The cap makes the payment depend on how the total loss is distributed around $U$.

Worked Example: Expected Payment for Two Random Claims

A hypothetical event produces exactly two claims. Each claim is independently $40,000 or $100,000, each with probability 0.5. There is a $120,000 occurrence limit and no other coverage terms. Find the expected payment. The claim severity distribution is the stated two-point distribution; independence is assumed. No interest or mortality assumption is relevant.

Step 1: List the possible claim totals and probabilities. Independence gives probability $0.5\times0.5=0.25$ for each equal-valued pair and probability $2(0.5\times0.5)=0.50$ for a mixed pair.

\begin{array}{c|c|c} \text{Claim amounts} & S & P(S)\\ (40{,}000,40{,}000) & 80{,}000 & 0.25\\ (40{,}000,100{,}000)\text{ or }(100{,}000,40{,}000) & 140{,}000 & 0.50\\ (100{,}000,100{,}000) & 200{,}000 & 0.25 \end{array}

Step 2: Apply the occurrence cap to each total. The corresponding payments are $80,000, $120,000, and $120,000.

Step 3: Weight the payments by their probabilities.

E[Y_{120{,}000}^{\mathrm{occ}}] =0.25(80{,}000)+0.50(120{,}000)+0.25(120{,}000) =20{,}000+60{,}000+30{,}000 =110{,}000.

The expected ground-up total is $2[0.5(40{,}000)+0.5(100{,}000)]=140{,}000$, which is greater than the expected capped payment, as it must be. Method and mistake avoided: build the distribution of the occurrence total before applying the limit. The result is not obtained by treating the $120,000 limit as a separate cap for each claim.

Occurrence-Level ILFs

An ILF compares expected limited losses at two limits on a consistent basis. With multiple claims per occurrence, the relevant loss basis for a shared cap is the total loss per occurrence. For a base occurrence limit $U_0$ and a higher occurrence limit $U_1$, define the occurrence-level ILF as the ratio of expected payments per occurrence, provided the denominator is positive:

\operatorname{ILF}_{\mathrm{occ}}(U_1;U_0) = \frac{E[\min(S,U_1)]}{E[\min(S,U_0)]}.

The numerator and denominator must use the same occurrence definition, claim grouping, coverage basis, and assumptions about the claim process. If the expected number and mix of occurrences are unchanged between the two limits, multiplying both expected payments by that common occurrence frequency leaves the ratio unchanged. If the exposure or occurrence frequency differs between the compared amounts, it cannot simply be assumed to cancel.

A claim-level ILF instead uses the distribution of an individual claim. It may be appropriate for a limit that applies independently to each claim, but it generally cannot represent a cap shared by multiple claims. Two portfolios can have the same individual-claim severity distribution and different numbers of claims per occurrence; their occurrence-limited expected payments can differ. Dependence between claim amounts can also affect the distribution of the occurrence total, even when individual claim distributions are the same.

Worked Example: Comparing Claim-Level and Occurrence-Level Factors

A hypothetical event always produces exactly two independent claims. Each claim is $20,000 or $80,000 with equal probability. Compare the ILF from a $50,000 to a $100,000 limit when each limit applies separately to each claim with the ILF when it is a shared occurrence limit. Assume the same claim distribution at both limits, with no other policy terms. No interest or mortality assumption is relevant.

Claim-level calculation. At a $50,000 per-claim limit, expected payment per claim is:

E[\min(X,50{,}000)] =0.5(20{,}000)+0.5(50{,}000) =35{,}000.

At a $100,000 per-claim limit, both possible claims are below the limit:

E[\min(X,100{,}000)] =0.5(20{,}000)+0.5(80{,}000) =50{,}000.

The claim-level ILF is $50{,}000/35{,}000=1.428571\ldots$.

Occurrence-level calculation. The total $S$ is $40,000 with probability 0.25, $100,000 with probability 0.50, and $160,000 with probability 0.25. At the $50,000 occurrence limit:

E[\min(S,50{,}000)] =0.25(40{,}000)+0.50(50{,}000)+0.25(50{,}000) =47{,}500.

At the $100,000 occurrence limit:

E[\min(S,100{,}000)] =0.25(40{,}000)+0.50(100{,}000)+0.25(100{,}000) =85{,}000.

The occurrence-level ILF is $85{,}000/47{,}500=1.789473\ldots$. Rounded only at the final step, to three decimal places, the claim-level factor is 1.429 and the occurrence-level factor is 1.789. The occurrence factor is larger in this example because increasing the shared cap allows more payment when the combined event loss exceeds the lower limit. Method and mistake avoided: calculate each ILF on the loss basis to which its limit applies. Substituting the claim-level ratio for the occurrence-level ratio would misstate the effect of this shared cap.

Claim Counts and Practical Checks

If the number of claims varies, the occurrence total is a random sum. One standard way to calculate the expected limited payment is to condition on the claim count:

E[\min(S,U)] =\sum_{n=0}^{\infty}P(N=n)\, E\!\left[\min\!\left(\sum_{j=1}^{n}X_j,U\right)\middle|N=n\right].

For $n=0$, the sum of claim amounts is zero and the payment is zero. The conditional calculation for positive $n$ must reflect the severity distribution and any relevant dependence among the claims. Assuming independent claim amounts is a modeling choice, not a consequence of having a known claim count.

A useful check is that the limited payment is nonnegative and cannot exceed the occurrence limit. Therefore $0\leq E[\min(S,U)]\leq U$. If the limit is increased while all other terms and the loss distribution remain fixed, the payment in each outcome cannot decrease, so expected payment cannot decrease either. These checks can reveal an incorrectly applied cap or inconsistent limit basis.

Identify the occurrence.
Determine which claims are grouped under the stated policy definition of one event.

Apply other stated terms.
For example, if a per-claim limit applies, apply it to each claim before summing, subject to the contract wording.

Aggregate the occurrence loss.
Add the claim amounts that remain payable before the shared occurrence cap.

Apply the occurrence cap once.
Take the lesser of the combined amount and the occurrence limit; for expected loss, average these capped payments over the occurrence distribution.

Common Mistakes

Using the occurrence limit once per claim. A shared limit caps the combined payment for an occurrence, not each claim individually.
Applying the cap before combining claims. With no per-claim term, combine the covered losses first and then apply the occurrence limit.
Using only the mean ground-up loss. The expected capped payment depends on the distribution of the occurrence total, not just its mean.
Using a claim-level ILF for a shared cap. The claim-level severity distribution alone does not describe how many claims arise per occurrence or their joint loss distribution.
Mixing loss bases. The numerator and denominator of an ILF must reflect the same occurrence grouping and coverage terms, changing only the limit being compared.

Key takeaway: A shared occurrence limit applies to the combined covered loss from one occurrence. Calculate or model that total first, cap it once, and use expected occurrence payments—not automatically claim-level limited losses—when the limit is per occurrence.

Check Your Understanding

For each problem, treat all stated claims as belonging to the same occurrence. Answers are rounded only where indicated.

Three deterministic claims are $25,000, $45,000, and $50,000. There is a $100,000 occurrence limit and no other terms. What is the total payment? (Answer: $100,000)
An occurrence has exactly two independent claims, each $20,000 or $80,000 with probability 0.5. With a $100,000 occurrence limit, what is the expected payment? (Answer: $85,000)
The number of claims $N$ has probabilities $P(N=0)=0.10$, $P(N=1)=0.60$, and $P(N=2)=0.30$. Each claim is a deterministic $40,000, and the occurrence limit is $70,000. Find expected payment per occurrence. (Answer: $45,000)
An occurrence always has two independent claims, each $30,000 or $90,000 with probability 0.5. Find the expected payment under a $100,000 occurrence limit, and compare it with the expected total if the $100,000 limit instead applied separately to each claim. (Answer: $90,000 under the occurrence limit; $120,000 under separate per-claim limits)
The claim count has probabilities $P(N=0)=0.20$, $P(N=1)=0.50$, and $P(N=2)=0.30$. Each claim is deterministically $40,000. Calculate the occurrence-level ILF from a $40,000 base occurrence limit to an $80,000 occurrence limit. (Answer: 1.375)

What You'll Learn