Zero-Claim Policies In Target Data

Why Zero-Claim Policies Matter

A policy can earn exposure during a coverage period and have no eligible claims in that period. It still represents insured risk during the time it was covered. For an exposure-based frequency or pure-premium target, its exposure belongs in the denominator even though its claim count and eligible claim loss are zero. That combination is easy to mishandle: dropping the policy changes the denominator, while treating it as a claim with zero severity changes the severity observations.

The distinction follows from the unit being measured. Frequency relates claim counts to exposure; pure premium relates eligible loss to exposure. Severity describes loss amounts for claims, so a policy with no claims supplies no claim amount to average. The earlier material on frequency responses, severity observations, and aggregate pure premium provides the terminology used here.

Definition: For a policy $i$ over a defined coverage period, let $e_i$ be its earned exposure, $N_i$ its number of eligible claims, and $L_i$ the total eligible loss amount for those claims. A zero-claim policy has $e_i>0$ and $N_i=0$. When the loss target includes only amounts attached to eligible claims, a correctly constructed zero-claim policy also has $L_i=0$. Its exposure is retained in exposure-based targets, but it creates no claim-level severity observation.

These statements assume that exposure, counts, and losses refer to the same coverage, period, eligibility rules, and loss basis. In particular, a record with no matched claim is not automatically a zero-claim policy: the claim search and target definition must support that classification. If a purported zero-claim policy has positive eligible loss, or a claim is associated with zero exposure, the data need reconciliation before calculating targets.

How Zero-Claim Policies Enter the Targets

For a group of policies with positive earned exposure, the basic aggregate targets are calculated from the totals. The formulas below assume the total eligible claim count is positive when calculating severity.

E=\sum_i e_i,\qquad N=\sum_i N_i,\qquad L=\sum_i L_i

\text{Frequency}=\frac{N}{E},\qquad \text{Severity}=\frac{L}{N}\quad(N>0),\qquad \text{Pure premium}=\frac{L}{E}

A zero-claim policy contributes $e_i$ to $E$, zero to $N$, and zero to $L$. Thus it lowers the aggregate frequency and pure premium relative to calculations that wrongly omit its exposure, but it does not directly add an amount to the severity numerator or denominator. The claim-level severity records are the eligible claim amounts—not one record for every policy.

Important condition: If $N>0$, aggregate severity is defined and the identity $\text{Pure premium}=\text{Frequency}\times\text{Severity}$ holds when all three targets use the same claims, loss amounts, and exposure. If $N=0$ and $L=0$, frequency and pure premium are zero, but severity $L/N$ is undefined; do not describe it as a zero-dollar severity.

Why the Frequency–Severity Identity Still Holds

The identity is an algebraic consequence of using consistent totals. It does not require every policy to have a claim. Suppose $N>0$, and let the individual eligible claim amounts be $Y_1,\ldots,Y_N$. Then $L=\sum_{j=1}^{N}Y_j$, so aggregate severity is the average of those claim amounts.

\text{Frequency}\times\text{Severity} =\frac{N}{E}\times\frac{L}{N} =\frac{L}{E} =\text{Pure premium}

The $N$ in the numerator of frequency cancels the $N$ in the denominator of severity. A zero-claim policy changes $E$, but not $N$ or $L$; therefore it changes frequency and pure premium together while leaving claim-level severity unchanged. This is why retaining its exposure is necessary for a consistent portfolio target.

Fix the target definition.
Specify the coverage period, eligible claims, loss amount basis, and exposure unit. Use the same definitions for every policy in the calculation.

Classify each policy record.
For positive exposure, a verified eligible count of zero indicates a zero-claim policy; its eligible claim loss should also be zero under the stated target.

Build exposure-based totals.
Sum exposure over all included policies, including zero-claim policies. Sum counts and eligible losses using the same population.

Build severity from claims.
Use the amount for each eligible claim. Do not create an additional claim amount, such as zero, for a policy with no eligible claims.

Check the targets.
When total count is positive, verify that frequency multiplied by severity equals pure premium before rounding.

Worked Calculations

Worked Example: One Zero-Claim Policy in a Small Portfolio

A hypothetical home coverage portfolio has three policies over the same annual target period. Policy A has $0.50$ policy-years of exposure and no eligible claims. Policy B has $1.00$ policy-year, one eligible claim of $\$600.00$, and Policy C has $1.50$ policy-years, two eligible claims of $\$400.00$ and $\$1{,}000.00$. Calculate aggregate frequency, severity, and pure premium, and identify Policy A's contribution to each target.

Method. Sum policy exposure, claim counts, and eligible claim amounts under the same target definition. Policy A has $N_A=0$ and $L_A=0$; the two claim amounts for Policy C are both individual claim observations.

Total exposure. $E=0.50+1.00+1.50=3.00$ policy-years.

Total count. $N=0+1+2=3$ claims.

Total eligible loss. $L=\$600.00+\$400.00+\$1{,}000.00=\$2{,}000.00$.

Frequency. $N/E=3/3.00=1.00$ claim per policy-year.

Severity. $L/N=\$2{,}000.00/3=\$666.666\ldots$, or $\$666.67$ per claim to the nearest cent.

Pure premium. $L/E=\$2{,}000.00/3.00=\$666.666\ldots$, or $\$666.67$ per policy-year to the nearest cent.

Contribution and check. Policy A contributes $0.50$ policy-years to exposure, zero claims, and $\$0.00$ eligible loss. It supplies no severity observation. Before rounding, frequency times severity is $(3/3.00)(\$2{,}000.00/3)=\$2{,}000.00/3.00$, which equals pure premium. The common mistake avoided is excluding Policy A's exposure because it had no claims.

Worked Example: Measuring the Effect of Dropping Zero-Claim Exposure

A hypothetical equipment policy group has three policies with exposures of $0.25$, $0.75$, and $1.00$ equipment-years. The first has zero eligible claims. The second has one claim of $\$900.00$; the third has one claim of $\$300.00$. Calculate the correct portfolio targets, then show what happens to frequency and pure premium if the first policy is incorrectly omitted.

Correct totals. Total exposure is $E=0.25+0.75+1.00=2.00$ equipment-years. Total count is $N=0+1+1=2$, and total loss is $L=\$900.00+\$300.00=\$1{,}200.00$.

Correct frequency. $N/E=2/2.00=1.00$ claim per equipment-year.

Correct severity. $L/N=\$1{,}200.00/2=\$600.00$ per claim.

Correct pure premium. $L/E=\$1{,}200.00/2.00=\$600.00$ per equipment-year. The identity check gives $1.00\times \$600.00=\$600.00$.

Incorrectly omitting the zero-claim policy. The remaining exposure is $0.75+1.00=1.75$ equipment-years, while count and loss remain 2 and $\$1{,}200.00$. The resulting frequency is $2/1.75=1.142857\ldots$, or $1.143$ claims per equipment-year to three decimals. The resulting pure premium is $\$1{,}200.00/1.75=\$685.714\ldots$, or $\$685.71$ per equipment-year to the nearest cent. Severity remains $\$600.00$.

Check and interpretation. The incorrect frequency times severity is $1.142857\ldots\times\$600.00=\$685.714\ldots$, matching the incorrect pure premium. The arithmetic identity still holds, but both exposure-based targets are biased upward relative to the correct population because a covered policy's exposure was removed. The common mistake is assuming that a policy without a claim is irrelevant to an exposure-based rate.

Worked Example: Converting Exposure Units Without Creating Severity Records

A hypothetical fleet target is recorded in vehicle-months. Vehicle A has 9 vehicle-months and zero eligible claims. Vehicle B has 15 vehicle-months and two claims, of $\$500.00$ and $\$700.00$. Vehicle C has 12 vehicle-months and one claim of $\$1{,}200.00$. Calculate frequency and pure premium per vehicle-year and severity per claim.

Convert exposure. There are 12 vehicle-months per vehicle-year. Total exposure is $9+15+12=36$ vehicle-months, so $E=36/12=3.00$ vehicle-years.

Sum eligible counts and losses. The total count is $N=0+2+1=3$. The loss total is $L=\$500.00+\$700.00+\$1{,}200.00=\$2{,}400.00$.

Frequency. $N/E=3/3.00=1.00$ claim per vehicle-year.

Severity. $L/N=\$2{,}400.00/3=\$800.00$ per claim.

Pure premium. $L/E=\$2{,}400.00/3.00=\$800.00$ per vehicle-year. The unrounded check is $1.00\times\$800.00=\$800.00$.

Common mistake avoided. Vehicle A's 9 vehicle-months are $9/12=0.75$ vehicle-years, not $9$ vehicle-years; they remain part of the total $3.00$ vehicle-years. Vehicle A also does not generate a fourth severity observation of $\$0.00$. There are exactly three eligible claim amounts and hence three severity observations.

Common Mistakes and Data Checks

Dropping zero-claim exposure. A policy with positive earned exposure remains in the denominator when it belongs to the target population. Removing it generally raises the calculated frequency and pure premium.
Adding zero severity observations. A zero-claim policy has no claim amount. Adding a zero to a claim-level severity average changes the unit being averaged and understates severity.
Confusing a zero count with a missing count. A zero is a value supported by the defined eligible-claim search; a blank, unmatched record, or incomplete claim listing is not necessarily zero. Establish the data status before assigning the target.
Mixing target definitions. Counts, claim amounts, and exposure must describe the same coverage and period. A claim amount that is outside the eligible loss basis cannot be used to contradict a zero eligible count without first resolving the mismatch.
Using severity when there are no claims. If $N=0$, there are no claim amounts to average. With no eligible loss, frequency and pure premium are zero, but severity is undefined—not zero.
Rounding before checking. Calculate with unrounded totals and round reported results only at the end. Rounded frequency times rounded severity may differ slightly from rounded pure premium.

Key takeaway: A verified zero-claim policy with positive exposure contributes exposure, but no eligible count, no eligible claim loss, and no claim-level severity observation. Retain it in frequency and pure-premium denominators; never turn it into a zero-dollar claim.

Check Your Understanding

For each problem, use consistent exposure, count, and eligible-loss totals. Report money to the nearest cent where needed.

A hypothetical liability group has two policies: one with $0.20$ policy-years and zero eligible claims, and one with $0.80$ policy-years and one eligible claim of $\$450.00$. Find frequency, severity, and pure premium, and state whether the first policy supplies a severity observation. (Answer: Frequency $=1.00$ claim per policy-year; severity $=\$450.00$ per claim; pure premium $=\$450.00$ per policy-year; the first policy supplies no severity observation.)
A hypothetical property portfolio has exposures of $0.50$, $1.00$, and $1.50$ property-years. Their eligible counts are $0$, $2$, and $1$. The two claims on the second policy are $\$300.00$ and $\$500.00$; the claim on the third is $\$900.00$. Calculate all three aggregate targets. (Answer: Frequency $=1.00$ claim per property-year; severity $=\$566.67$ per claim; pure premium $=\$566.67$ per property-year.)
A hypothetical vehicle target has exposures of 9, 15, and 12 vehicle-months, with eligible counts $0$, $1$, and $2$. The claims are $\$600.00$ on the second vehicle and $\$300.00$ and $\$900.00$ on the third. Calculate frequency per vehicle-year, severity, and pure premium per vehicle-year. (Answer: Frequency $=1.00$ claim per vehicle-year; severity $=\$600.00$ per claim; pure premium $=\$600.00$ per vehicle-year.)
A hypothetical segment has 2.00 policy-years with zero eligible claims and $\$0.00$ eligible loss, plus 1.50 policy-years with three claims totaling $\$1{,}000.00$. Find frequency, severity, and pure premium; then state the effect on frequency and pure premium of wrongly dropping the zero-claim exposure. (Answer: Correct frequency $=0.8571$ claims per policy-year, severity $=\$333.33$ per claim, and pure premium $=\$285.71$ per policy-year; dropping the zero-claim exposure gives frequency $=2.000$ and pure premium $=\$666.67$, while severity is unchanged.)
A hypothetical target has 1.25 earned policy-years, a recorded eligible claim count of zero, and a positive eligible claim loss total of $\$250.00$. Can the record be used as a zero-claim policy as stated? Explain what must be done before calculating targets. (Answer: No. A zero eligible count and positive eligible claim loss are inconsistent under the stated target; reconcile claim eligibility, count, loss assignment, and records before calculating the targets.)

What You'll Learn