Territory Loss Costs per Exposure

Why Measure Losses per Exposure?

A territory with more insured vehicles will generally produce more total losses simply because it has more exposure to risk. Comparing aggregate losses alone can therefore confuse territory size with loss cost per insured unit. For territorial analysis, a useful measure is the amount of loss associated with one exposure, with losses and exposures placed on appropriate, consistent bases.

This tutorial focuses on calculating that measure from adjusted losses and exposures. The adjustments are treated as supplied inputs: they might place losses on a common period or coverage basis. The exposure count must correspond to the same experience and unit, such as earned vehicle-years. Where experience has already been standardized for exposure mix, use that standardized basis consistently rather than combining it with unstandardized figures.

Definition: A territory's loss cost per exposure is its selected losses, adjusted to the intended common basis, divided by the corresponding number of exposures. It estimates the loss dollars associated with one exposure unit. The measure is also commonly called a pure premium when stated per exposure.

Formula and Interpretation

Let $L_j^{\text{adj}}$ denote adjusted losses for territory $j$, and let $E_j$ denote its earned exposures on the selected basis. The territorial loss cost is

LC_j = \frac{L_j^{\text{adj}}}{E_j}.

The units matter. If adjusted losses are dollars and exposures are earned vehicle-years, the result is dollars per earned vehicle-year. Do not divide by policy count, written exposures, or another quantity unless that is the exposure basis specified for the analysis.

A loss cost is a measure of expected loss per exposure, not a complete premium rate. A premium rate may also need provisions for expenses, profit, contingencies, and other elements. Nor does a loss-cost comparison, by itself, determine an approved territorial relativity. Its role here is to provide a comparable loss-dollar indication per exposure.

Conditions for a meaningful comparison: Use the same definition of loss across territories, the same exposure unit, and losses adjusted to a common basis. The loss measure might be paid, incurred, or ultimate losses, but it must be selected consistently. If territories have materially different risk mixes, a raw comparison may reflect mix as well as territory; use experience standardized to a comparable mix when that is the intended comparison.

Why the Ratio of Totals Is the Right Pooled Measure

Suppose there are $E$ exposures, each with expected loss $\mu$ on the selected basis. If $Z_k$ is the loss associated with exposure $k$, then $E[Z_k]=\mu$. By additivity of expectation,

E\left[\sum_{k=1}^{E} Z_k\right] = \sum_{k=1}^{E} E[Z_k] = E\mu.

Thus expected aggregate loss is exposure count times expected loss per exposure. Dividing aggregate losses by exposures puts the observed total on a per-exposure scale and estimates $\mu$. This justification assumes the exposures represent the intended risk population and that losses are measured on a consistent basis; it does not make an unadjusted or mismatched dataset comparable automatically.

The same reasoning explains how to pool several periods or groups. Add their adjusted losses and exposures, then divide the totals. Equivalently, average the group loss costs using each group's exposure as its weight. A simple unweighted average gives a small group the same influence as a much larger group, which generally does not estimate the combined loss per exposure.

LC_{\text{pooled}} = \frac{\sum_j L_j^{\text{adj}}}{\sum_j E_j} = \sum_j \left(\frac{E_j}{\sum_r E_r}\right)LC_j.

Calculating Adjusted Loss Costs

Worked Example: Applying a Supplied Loss Adjustment

A hypothetical territory has $735,000 in losses before a supplied adjustment of 1.06 and 2,500 earned vehicle-years. The adjustment is to be applied to losses, and the resulting losses are on the required basis. Calculate the adjusted loss cost per earned vehicle-year.

Method. Adjust the loss total first, then divide by the matching exposure count.

Step 1: Adjust losses.

L^{\text{adj}} = \$735{,}000 \times 1.06 = \$779{,}100.

Step 2: Divide by earned exposures.

LC = \frac{\$779{,}100}{2{,}500} = \$311.64 \text{ per earned vehicle-year}.

Check. Multiplying the calculated cost by exposures returns the adjusted losses: $\$311.64 \times 2{,}500 = \$779{,}100$. The answer is $311.64 per earned vehicle-year.

Common mistake avoided. Dividing the unadjusted $735,000 by 2,500 gives $294.00, which is not the requested adjusted loss cost. Applying the adjustment in the wrong direction is also incorrect: the problem explicitly says to multiply losses by 1.06.

Pooling Experience Across Periods

A territory's experience may span several periods with different exposure volumes and supplied adjustment factors. Calculate each period's adjusted loss total, add the adjusted totals, and divide by the total exposures. The exposure-weighted average of period loss costs gives the same result; an unweighted average generally does not.

Worked Example: Combining Two Adjusted Periods

A hypothetical territory has the following experience. Period A recorded $420,000 in losses over 1,800 earned vehicle-years, with a supplied loss adjustment factor of 1.10. Period B recorded $510,000 over 2,200 earned vehicle-years, with a supplied factor of 1.04. Apply each factor to its period's losses, then calculate the pooled loss cost.

Method. Adjust the periods separately, combine the adjusted losses, and divide by combined exposures.

Step 1: Adjust Period A.

\$420{,}000 \times 1.10 = \$462{,}000, \qquad LC_A = \frac{\$462{,}000}{1{,}800} = \$256.6667.

Step 2: Adjust Period B.

\$510{,}000 \times 1.04 = \$530{,}400, \qquad LC_B = \frac{\$530{,}400}{2{,}200} = \$241.0909.

Step 3: Pool totals. Adjusted losses total $\$462{,}000+\$530{,}400=\$992{,}400$, and exposures total $1{,}800+2{,}200=4{,}000$.

LC_{\text{pooled}} = \frac{\$992{,}400}{4{,}000} = \$248.10 \text{ per earned vehicle-year}.

Check. The exposure weights are $1{,}800/4{,}000=0.45$ and $2{,}200/4{,}000=0.55$. The weighted cost is $0.45(\$256.6667)+0.55(\$241.0909)=\$248.10$, subject to rounding in the displayed intermediate costs.

Common mistake avoided. The simple average of the two period costs is approximately $248.88. It is not the pooled result because it gives each period equal weight despite their different exposure counts.

Aggregate Losses and Per-Exposure Indications

Aggregate adjusted losses describe the total dollars in the observed experience; loss cost describes those dollars relative to exposure. These answer different questions. A large territory can have higher aggregate losses and a lower loss cost if it has many more exposures. Conversely, a smaller territory can have lower aggregate losses and a higher loss cost.

Worked Example: Separating Territory Size from Loss Cost

Two hypothetical territories have already been placed on their respective required loss bases. North has $840,000 before a supplied adjustment factor of 1.10 and 6,600 earned vehicle-years. South has $480,000 before a supplied factor of 1.05 and 2,400 earned vehicle-years. Calculate each adjusted loss cost and the combined loss cost, then interpret the aggregate-loss comparison.

Method. Adjust each territory's losses, divide by its own exposures, and use total adjusted losses over total exposures for the combined measure.

Step 1: North. Adjusted losses are $\$840{,}000 \times 1.10=\$924{,}000$. Therefore,

LC_N = \frac{\$924{,}000}{6{,}600} = \$140.00 \text{ per earned vehicle-year}.

Step 2: South. Adjusted losses are $\$480{,}000 \times 1.05=\$504{,}000$. Therefore,

LC_S = \frac{\$504{,}000}{2{,}400} = \$210.00 \text{ per earned vehicle-year}.

Step 3: Combined measure. Total adjusted losses are $\$924{,}000+\$504{,}000=\$1{,}428{,}000$, and total exposures are $6{,}600+2{,}400=9{,}000$.

LC_{\text{combined}} = \frac{\$1{,}428{,}000}{9{,}000} = \$158.67 \text{ per earned vehicle-year, rounded to cents}.

Check and interpretation. North has greater aggregate adjusted losses ($924,000 versus $504,000), but its loss cost is lower ($140.00 versus $210.00). North's larger total reflects, in part, its larger exposure volume. The combined loss cost lies between the two territory costs, as expected for an exposure-weighted average.

Common mistake avoided. Ranking North as the higher-cost territory based only on aggregate losses would reverse the per-exposure comparison. The aggregate total is not the loss cost.

Common Traps and Practical Checks

Mixing units. Confirm whether exposures are earned vehicle-years, policies, or another specified unit. The answer inherits the denominator's unit.
Adjusting the wrong quantity. Apply a supplied loss adjustment to losses when instructed; do not automatically apply it to exposures as well.
Averaging period costs equally. Pool by dividing combined adjusted losses by combined exposures, or use exposure weights.
Comparing inconsistent loss bases. A paid-loss measure in one territory and an ultimate-loss measure in another is not a like-for-like comparison without appropriate treatment.
Treating loss cost as a final premium rate. Loss cost reflects loss dollars per exposure; it does not, by itself, include all other components of a premium rate or establish a final territorial factor.
Ignoring mix. A difference in raw loss costs may reflect differences in the types of risks insured. Use an appropriate standardized basis when the objective is to compare territories at a common mix.

Key takeaway: Calculate territorial loss cost as adjusted losses divided by corresponding exposures. For combined experience, use the ratio of totals; interpret the result per exposure and separately from aggregate loss volume or a complete premium rate.

Check Your Understanding

For each problem, use the specified adjustment and exposure basis. Round dollar loss costs to the nearest cent when needed.

A hypothetical territory has $312,000 in losses and 1,800 earned vehicle-years. A supplied factor of 1.08 is to be applied to losses. What is the adjusted loss cost per earned vehicle-year? (Answer: $187.20 per earned vehicle-year)
Period A has $260,000 in losses, 1,200 exposures, and a supplied loss factor of 1.05. Period B has $340,000 in losses, 1,800 exposures, and a factor of 0.95. Calculate the pooled adjusted loss cost. (Answer: $198.67 per exposure)
Territory A has $720,000 in adjusted losses and 3,600 exposures; Territory B has $510,000 in adjusted losses and 1,700 exposures. Which has the greater aggregate loss, which has the greater loss cost, and what are the two costs? (Answer: A has greater aggregate losses; B has greater loss cost; A is $200.00 per exposure and B is $300.00 per exposure)
A combined book contains 800 exposures with an adjusted cost of $180 each and 1,200 exposures with an adjusted cost of $240 each. Calculate the combined adjusted loss cost and explain why the unweighted average is not the answer. (Answer: $216.00 per exposure; the unweighted average is $210.00 and gives the two groups equal weight despite their different exposure counts)
A territory has $125,000 in losses before a supplied 1.12 adjustment and 500 earned exposures. The factor applies to losses only. Calculate the adjusted loss cost, and state the unit if exposures are earned vehicle-years. (Answer: $280.00 per earned vehicle-year)

What You'll Learn