Accounting for Changes in Coverage and Business Mix

Why Coverage and Mix Comparability Matters

An indication compares historical loss experience with a premium or exposure basis intended to represent a prospective book. That comparison is meaningful only when the losses and the prospective risks refer to a sufficiently consistent set of insured coverages and risk characteristics. A deductible increase can reduce insurer-paid losses without changing the underlying ground-up claims. A shift toward higher-risk territories can raise aggregate losses even if each territory’s loss cost is unchanged. If these differences go unrecognized, the experience may suggest a change in underlying risk that is actually due to changed coverage or business composition.

This tutorial focuses on identifying and accounting for those differences. It assumes the experience period, premium or exposure records, loss measure, and any required development or trend adjustments have already been aligned as appropriate. The adjustment discussed here is not a substitute for those steps.

Definition: A coverage change changes the protection provided by the policy, such as its deductible, limit, or definition of covered loss. A business-mix change changes the distribution of insured risks among groups with different expected loss costs, such as territories, classes, or coverage selections. To compare historical experience with a prospective book, express both on a consistent coverage and mix basis, or explain why a reliable adjustment cannot be made.

Restating Losses for Changed Policy Terms

When claim-level ground-up losses are available, it may be possible to calculate what the insurer would have paid under different policy terms. For a ground-up loss $X$, a per-claim deductible $d$, and a maximum insurer payment $u$, define the insurer payment as

Y^L=\min\bigl((X-d)_+,u\bigr),\qquad (z)_+=\max(z,0).

This expression assumes the deductible is applied first and the limit $u$ caps the insurer’s payment after the deductible. Policy wording can define limits differently, so the formula must match the actual contract. With no payment limit, the payment is $(X-d)_+$. With no deductible, it is $\min(X,u)$.

Recalculate each claim under the terms being compared, then sum the payments. The difference between totals measures the effect of those terms on this particular set of ground-up claims. It does not, by itself, establish the future expected cost: the observed claims may not represent the future severity distribution, and claim frequency or mix may also change.

Worked Example: Recalculating Claims Under a Higher Deductible

A hypothetical commercial policy has no payment limit. Three ground-up claims are $1,500, $5,000, and $10,000. The historical deductible was $500, and the prospective deductible is $2,000. Calculate the total insurer payments under each deductible and the ratio of prospective to historical payments. Assume the same three ground-up losses would occur under either contract.

Method. Recalculate each claim using $(X-d)_+$, first with the historical deductible and then with the prospective deductible. This avoids treating the old insurer payments as if they were ground-up losses.

Under the $500 deductible, the claim payments are:

(1{,}500-500)_+=1{,}000,\quad (5{,}000-500)_+=4{,}500,\quad (10{,}000-500)_+=9{,}500.

The historical total is $1{,}000+4{,}500+9{,}500=\$15{,}000$.

Under the $2,000 deductible, the payments are:

(1{,}500-2{,}000)_+=0,\quad (5{,}000-2{,}000)_+=3{,}000,\quad (10{,}000-2{,}000)_+=8{,}000.

The prospective total is $0+3{,}000+8{,}000=\$11{,}000$. The ratio of prospective to historical payments is $11{,}000/15{,}000=0.7333$, or about 73.33%. For this claim set, payments under the higher deductible are 26.67% lower.

Check. Each recalculated payment is nonnegative and no larger than its ground-up claim. The higher deductible reduces each payment or leaves it at zero, so the prospective total is below the historical total. A common mistake is to subtract the new deductible from the old insurer-paid amounts. That is wrong because those amounts have already been reduced by the old deductible; the calculation must start from ground-up losses.

Worked Example: Accounting for a Higher Per-Claim Limit

A hypothetical liability policy has no deductible. Its historical per-claim limit was $50,000, and its prospective limit is $100,000. Four ground-up claims are $20,000, $60,000, $120,000, and $40,000. Calculate total payments at each limit and the ratio of prospective to historical payments.

Method. With no deductible, payment for each claim is $\min(X,u)$. Apply each limit separately to every ground-up claim.

At the $50,000 limit, the payments are $20,000, $50,000, $50,000, and $40,000. The total is $20{,}000+50{,}000+50{,}000+40{,}000=\$160{,}000$.

At the $100,000 limit, the payments are $20,000, $60,000, $100,000, and $40,000. The total is $20{,}000+60{,}000+100{,}000+40{,}000=\$220{,}000$.

The ratio is $220{,}000/160{,}000=1.375$. Thus, payments on these claims under the higher limit are 37.5% greater than under the historical limit.

Check. The claim of $120,000 reaches both limits, while the $60,000 claim reaches only the higher limit. The other two claims are below both limits. The calculation therefore increases payments only for the two claims affected by the higher limit. Applying the limit to the total of all ground-up claims, rather than to each claim, would be a common error.

Standardizing for a Change in Business Mix

A mix comparison requires groups, or cells, that are defined consistently across historical and prospective business. For example, cells might distinguish territory and class together. Let $w_j$ be the share of exposure in cell $j$, where the shares sum to 1, and let $L_j$ be the loss cost per exposure for that cell on a common coverage and time basis. The weighted loss cost for a book with those shares is

L=\sum_j w_j L_j,\qquad \sum_j w_j=1.

This is a weighted average: each cell’s loss cost contributes according to its share of the total exposure. If cell-level loss costs are held constant, a mix-only adjustment factor comparing prospective with historical business is

F_{\text{mix}}= \frac{\sum_j w_{j,\text{prospective}}L_j} {\sum_j w_{j,\text{historical}}L_j}.

The factor is meaningful only if the cell definitions and exposure bases are consistent, and the cell loss costs represent the same coverage and time basis. This calculation isolates the effect of changed shares under those assumptions; it does not establish that future cell-level costs will remain unchanged.

Worked Example: Standardizing a Shift Between Territories

A hypothetical insurer has two territory cells with loss costs of $900 per exposure in Territory A and $400 per exposure in Territory B. Historical business was 75% A and 25% B. The prospective book is expected to be 40% A and 60% B. Assume these cell loss costs are on the same coverage and time basis and remain constant. Calculate the historical and prospective weighted loss costs and the mix factor.

Method. Use the exposure shares as weights in each book’s weighted average, then divide prospective by historical.

Historical weighted loss cost: $0.75(900)+0.25(400)=675+100=\$775$ per exposure.

Prospective weighted loss cost: $0.40(900)+0.60(400)=360+240=\$600$ per exposure.

The mix factor is $600/775=0.7741935$, or 0.7742 to four decimal places. On these assumptions, the changed territory distribution reduces weighted loss cost by about 22.58%.

Check. The prospective book has a smaller share of the higher-cost territory, so its weighted loss cost should be lower. The calculated $600 lies between the two cell costs, as does the historical $775. Do not take a simple average of the two territory costs: the shares are unequal and determine the weights.

Coverage Definitions and the Limits of Aggregate Data

Not every coverage change can be handled by applying a deductible or limit formula. A revised definition of covered loss may add or remove types of claims, change an exclusion, or alter conditions for coverage. In such cases, a defensible restatement may require claim-level information identifying which losses fall under the old and new definitions. If historical aggregate losses do not separate affected from unaffected claims, the amount to remove or add may be unknown.

A useful approach is to identify the component of historical loss that would not be covered prospectively, or the component that would be newly covered, and adjust only when it can be measured on a supportable basis. If the data do not support that measurement, report the limitation rather than inventing a factor. Any comparison should also avoid adjusting the same effect twice—for example, removing an excluded component from historical losses and then applying another adjustment that already reflects the same exclusion.

Worked Example: Separating a Newly Excluded Coverage Component

A hypothetical policy’s historical loss total is $480,000 across 600 exposures. Claim records show that $120,000 relates to portable devices, a component excluded under the prospective coverage definition. The remaining losses are covered under both definitions. Assume the same exposure count is used for the comparison and no other adjustment is needed. Calculate the restated historical total and loss cost under the prospective definition, and compare it with the unadjusted historical loss cost.

Method. Remove the identifiable losses that are covered historically but excluded prospectively, then divide the restated total by the consistent exposure count.

The restated historical losses are $480{,}000-120{,}000=\$360{,}000$. The restated loss cost is $360{,}000/600=\$600$ per exposure. The unadjusted historical loss cost is $480{,}000/600=\$800$ per exposure. The restated cost is $600/800=0.75$, or 25% below the unadjusted cost.

Check. The removed component is one quarter of the original total: $120{,}000/480{,}000=0.25$. Removing it leaves 75% of the original losses, consistent with the ratio $600/800=0.75$. This adjustment is supported here because the affected component is identified. If only the $480,000 aggregate were known, the $120,000 removal could not be justified from that total alone.

A Practical Comparability Review

Before using historical experience to represent the prospective book, document what changed and whether the available data permit a consistent comparison. The following sequence keeps coverage and mix effects distinct:

Describe the historical and prospective basis.
List relevant deductibles, limits, coverage definitions, and risk cells for each book.

Locate the supporting data.
Determine whether ground-up claim amounts, claim-level coverage details, and exposure counts by comparable cell are available.

Restate only measurable differences.
Recalculate claim payments from ground-up losses when policy terms change, and standardize cell weights when the business mix changes.

Check the result and its assumptions.
Confirm that exposure bases and coverage definitions match, and that no effect has been adjusted twice.

State any limitation.
If aggregate data do not identify the affected coverage or cell, explain why a precise adjustment is not supportable.

The most important exam trap is to treat every change in aggregate losses as a change in risk cost. A lower total after a deductible increase, for example, may reflect only the reduced insurer payment per claim. Conversely, greater total losses may reflect a move toward higher-cost cells rather than higher costs within each cell. Keep the comparison on a common basis before interpreting it.

Key takeaway: Coverage adjustments require enough claim detail to apply the changed contract terms; mix adjustments require comparable cell-level costs and exposure shares. If the required information is unavailable, an unsupported adjustment is less reliable than clearly stating the limitation.

Check Your Understanding

For each problem, state the calculation on a consistent basis and check that the result matches the direction implied by the data.

A hypothetical policy has no payment limit. Ground-up claims are $800, $2,000, and $4,000. The deductible changes from $250 to $1,000. Find the total payments under each deductible. (Answer: Old total $6,050; new total $4,000.)
A policy has no deductible. Three ground-up claims are $10,000, $30,000, and $50,000. The per-claim limit changes from $25,000 to $40,000. Find the payment totals and the prospective-to-historical ratio. (Answer: Old total $60,000; new total $80,000; ratio 1.3333.)
Two cells have loss costs of $500 and $1,200 per exposure. Historical shares are 60% and 40%; prospective shares are 30% and 70%. Assuming cell costs are unchanged and comparable, find both weighted costs and the mix factor. (Answer: Historical $780; prospective $990; factor 1.2692.)
Historical losses are $800,000 across 500 exposures. Claim-level records show that 15% of losses are for a component excluded prospectively. Find the restated historical losses and loss cost under the prospective definition. (Answer: Restated losses $680,000; loss cost $1,360 per exposure.)
Two risk cells have loss costs of $200 and $500. Historical exposure shares are 80% and 20%; prospective shares are 50% and 50%. Calculate the weighted costs and the mix factor. (Answer: Historical $260; prospective $350; factor 1.3462.)

What You'll Learn