Why the Type of Limit Matters
An experience rating plan may restrict the losses used in a modification so that a single severe claim does not have unlimited influence. A second kind of restriction may limit the total losses included across the experience period. Although both restrictions reduce losses in some cases, they operate on different units: one applies claim by claim, while the other applies to the combined total.
The distinction matters when calculating an experience modification. A per-claim limit responds to the size and distribution of individual claims. An aggregate cap responds to the total eligible losses for the experience period. If a plan uses both, the calculation must apply each restriction to the quantity it governs before putting the result into the modification formula.
Set Up the Loss Calculation
Let \(X_j\) be the loss amount for claim \(j\), for \(j=1,\ldots,n\). In this tutorial, assume every \(X_j\) is nonnegative, the amounts are already on the plan’s required valuation basis, and the stated limits apply directly to those amounts. No loss exclusions, claim adjustments, or other plan-specific changes are made. The expected losses used in the modification denominator are also given on the appropriate basis.
With a per-claim limit \(L\), the eligible total is the sum of the limited amounts for the individual claims. With an aggregate cap \(C\), first sum the claims and then cap that total. If a hypothetical plan specifies both, this tutorial assumes the per-claim limit is applied first and the aggregate cap is then applied to the resulting total.
Here, \(A_L\), \(A_C\), and \(A_{L,C}\) are eligible actual losses under a per-claim limit, an aggregate cap, or both, respectively. Do not assume that the same cap amount applies in each formula: \(L\) is a per-claim threshold, while \(C\) is a total-period threshold. A plan’s wording controls whether it uses one restriction, the other, or both, and in what sequence.
Using Eligible Losses in a Modification
For the calculations below, use the following credibility-weighted modification model. Let \(E>0\) be expected losses for the experience period, \(A^*\) be the eligible actual losses after applying the stated plan restrictions, and \(Z\) be the credibility assigned to the experience, with \(0\leq Z\leq1\). Assume the plan blends the actual experience with expected losses and has no additional factor adjustment.
The expected-loss denominator is not automatically reduced just because actual losses are limited. Use a modified denominator only if the plan explicitly specifies one. Also, this formula is being used under its stated assumptions: it does not include minimum or maximum modification factors, separate treatment of loss components, or other plan adjustments.
Why the Credibility Formula Has This Form
The formula follows directly from the assumed blend. The plan assigns weight \(Z\) to eligible actual losses and weight \(1-Z\) to expected losses. Divide the weighted amount by \(E\) to express it as a factor relative to expected losses:
This form also provides useful checks. If \(Z=0\), then \(M=1\), because the result is entirely based on expected losses. If \(Z=1\), then \(M=A^*/E\), because the result is entirely based on eligible actual losses. When \(A^*=E\), the factor is 1 for any admissible \(Z\). These checks do not replace the calculation, but can reveal a sign or denominator error.
Worked Calculations
Worked Example: One Per-Claim Limit Versus One Aggregate Cap
A hypothetical insured has three claims of $18,000, $52,000, and $90,000. Compare the eligible loss totals under a $40,000 per-claim limit and a $100,000 aggregate cap, applied separately. For each result, calculate the modification using \(E=\$120{,}000\) and \(Z=0.60\). Assume no other adjustments.
Method. Calculate each eligible total under its own restriction, then substitute that total as \(A^*\). The raw sum is:
$18,000 + $52,000 + $90,000 = $160,000.
Per-claim limit. Limit each claim separately to $40,000:
min($18,000, $40,000) = $18,000; min($52,000, $40,000) = $40,000; min($90,000, $40,000) = $40,000.
Thus \(A_L=\$18{,}000+\$40{,}000+\$40{,}000=\$98{,}000\). The modification is:
\(M_L=1+0.60(\$98{,}000/\$120{,}000-1)=1+0.60(0.8166667-1)=1-0.11=0.8900.\)
Aggregate cap. Add the claims first, then cap the total: \(A_C=\min(\$160{,}000,\$100{,}000)=\$100{,}000\). Therefore:
\(M_C=1+0.60(\$100{,}000/\$120{,}000-1)=1+0.60(0.8333333-1)=1-0.10=0.9000.\)
The aggregate cap produces $2,000 more eligible losses than the per-claim limit, so it produces the higher factor in this example. As a check, the raw, uncapped factor would be \(1+0.60(\$160{,}000/\$120{,}000-1)=1.2000\), higher than either restricted result. Common mistake avoided: applying the $100,000 aggregate cap separately to each claim would not cap the combined total.
Worked Example: The Claim Distribution Changes the Per-Claim Result
Two hypothetical experience periods each have $160,000 of losses. Period A has claims of $20,000, $20,000, $20,000, and $100,000. Period B has four claims of $40,000 each. A plan applies a $50,000 per-claim limit and then a $140,000 aggregate cap. Calculate eligible losses and modifications for both periods using \(E=\$150{,}000\) and \(Z=0.50\).
Method. Apply the per-claim limit to each claim, sum those amounts, and then apply the aggregate cap. For Period A, the per-claim amounts are $20,000, $20,000, $20,000, and $50,000, for a total of $110,000. Applying the aggregate cap gives \(A_{L,C}=\min(\$110{,}000,\$140{,}000)=\$110{,}000\).
Its modification is \(M_A=1+0.50(\$110{,}000/\$150{,}000-1)=1+0.50(0.7333333-1)=1-0.1333333=0.8667\), rounded to four decimal places.
For Period B, each $40,000 claim is below the per-claim limit. The limited total is \(4(\$40{,}000)=\$160{,}000\). The aggregate cap reduces it to \(A_{L,C}=\min(\$160{,}000,\$140{,}000)=\$140{,}000\).
Its modification is \(M_B=1+0.50(\$140{,}000/\$150{,}000-1)=1+0.50(0.9333333-1)=1-0.0333333=0.9667\), rounded to four decimal places. Both periods have the same raw losses, but their eligible totals differ because the per-claim limit has a different effect on their claim distributions. Common mistake avoided: assuming that equal raw totals necessarily produce equal modified totals.
Worked Example: Applying Both Restrictions in the Specified Order
A hypothetical experience period has claims of $35,000, $80,000, and $140,000. The plan applies a $60,000 per-claim limit and then a $145,000 aggregate cap. Calculate eligible losses and the modification using \(E=\$170{,}000\) and \(Z=0.40\). Also identify the result if only the per-claim limit applied.
Method. First apply the $60,000 limit to each claim: the eligible amounts are $35,000, $60,000, and $60,000. Their sum is \(\$35{,}000+\$60{,}000+\$60{,}000=\$155{,}000\). Apply the aggregate cap to that sum: \(A_{L,C}=\min(\$155{,}000,\$145{,}000)=\$145{,}000\).
The modification under both restrictions is \(M=1+0.40(\$145{,}000/\$170{,}000-1)=1+0.40(0.8529412-1)=1-0.0588235=0.9412\), rounded to four decimal places.
If only the per-claim limit applied, eligible losses would be $155,000 and the factor would be \(1+0.40(\$155{,}000/\$170{,}000-1)=1+0.40(0.9117647-1)=1-0.0352941=0.9647\), rounded to four decimal places. The aggregate cap therefore reduces eligible losses by a further $10,000 and lowers the factor. Common mistake avoided: treating the aggregate cap as a limit on each claim. The aggregate cap applies once to the combined post-limit total, not separately to each amount.
Interpreting the Difference
A per-claim limit changes the contribution of each claim that exceeds its threshold. Adding another claim below the threshold adds its full amount, while adding another large claim adds at most \(L\). An aggregate cap instead restricts the total eligible losses for the period. Once the total reaches \(C\), additional losses do not increase the eligible total under that cap.
When both restrictions apply in the stated order, the aggregate cap can bind even after the per-claim limits have reduced the claim total. It can also fail to bind if the post-limit total is below \(C\). Therefore, calculate the post-limit sum before deciding whether the aggregate cap changes the result. For the model in this tutorial, a smaller eligible actual loss produces a smaller modification when \(Z>0\), \(E>0\), and all other inputs remain fixed.
| Plan provision | Quantity restricted | Calculation | What drives the result |
|---|---|---|---|
| Per-claim limit | Each individual claim | Sum min(\(X_j,L\)) | The number and sizes of claims above \(L\) |
| Aggregate cap | Total losses for the period | min(sum \(X_j,C\)) | The raw combined total relative to \(C\) |
| Both, in the stated order | Each claim, then the resulting total | min(sum min(\(X_j,L\)), \(C\)) | Both the claim distribution and capped total |
Common Errors to Avoid
- Using an aggregate cap as a per-claim limit. The expression \(\sum_j\min(X_j,C)\) does not enforce an aggregate cap of \(C\). Apply the cap once to the total it governs.
- Applying a per-claim limit to the raw total. A per-claim provision requires limiting each \(X_j\) before adding the claims. It is not \(\min(\sum_jX_j,L)\).
- Changing the expected-loss denominator without authority. In the specified model, use the stated \(E\) even when eligible actual losses are capped. Alter \(E\) only when the plan says to do so.
- Using the wrong actual amount in the factor. The formula uses \(A^*\), the eligible amount after the stated restrictions, not the raw sum if the plan restricts losses.
- Forgetting to check whether a restriction binds. If a total is already below an aggregate cap, that cap does not change the total. If every claim is below a per-claim limit, that limit does not change any claim.
A reliable calculation keeps three quantities separate: raw losses, eligible losses after the plan’s specified restrictions, and the final credibility-weighted factor. Labeling each amount before substituting it into the formula makes it easier to spot an incorrect operation or denominator.
Check Your Understanding
For each problem, use the stated credibility-weighted model and assume no adjustments beyond those specified.
- A hypothetical account has claims of $12,000, $48,000, and $75,000. A $40,000 per-claim limit applies. If \(E=\$100{,}000\) and \(Z=0.50\), calculate eligible losses and the modification. (Answer: eligible losses $92,000; modification 0.9600.)
- Claims total $200,000, and the plan imposes a $150,000 aggregate cap. If \(E=\$180{,}000\) and \(Z=0.40\), calculate the modification. (Answer: eligible losses $150,000; modification 0.9333.)
- A period has claims of $30,000, $70,000, and $95,000. Apply a $60,000 per-claim limit followed by a $145,000 aggregate cap. If \(E=\$160{,}000\) and \(Z=0.80\), calculate eligible losses and the modification. (Answer: eligible losses $145,000; modification 0.9250.)
- Compare two periods under a $50,000 per-claim limit followed by a $90,000 aggregate cap. Period A has claims of $10,000, $10,000, and $100,000; Period B has three claims of $40,000. For each period, calculate eligible losses and the modification if \(E=\$100{,}000\) and \(Z=0.50\). (Answer: Period A: $70,000 and 0.8500; Period B: $90,000 and 0.9500.)
- A period has claims of $55,000, $75,000, $130,000, and $20,000. Apply an $80,000 per-claim limit followed by a $210,000 aggregate cap. If \(E=\$240{,}000\) and \(Z=0.60\), calculate the post-limit total, eligible losses after the aggregate cap, and the modification. (Answer: post-limit total $230,000; eligible losses $210,000; modification 0.9250.)