Why the Count Base Matters
A fixed expense total does not determine how much expense should be assigned to each rate unit by itself. The calculation also needs a denominator: the count of the items receiving the allocation. In insurance, a policy, an insured unit, and an earned exposure are not interchangeable. A policy may insure several vehicles, while an earned exposure measures the portion of a coverage unit that was actually in force during a period.
The distinction matters whenever the book contains policies of different sizes, policies that are not in force for a full year, or expenses driven by different activities. Dividing an expense total by the wrong count can produce a calculation that adds up arithmetically but assigns cost to the wrong basis. The objective is not to select the largest or most convenient denominator. It is to match each expense allocation to the activity or rate unit it is intended to represent.
Three Denominators, Three Interpretations
Suppose a hypothetical insurer writes personal watercraft coverage. One policy might cover two watercraft. If the policy is in force for a full year, it contributes one policy-year but may contribute two watercraft-years. If it is in force for half a year, it may contribute one-half earned policy exposure and one earned watercraft exposure, assuming each watercraft was covered for that half-year. Actual exposure rules depend on the product's rating basis and data definitions, so the counts must be interpreted in context.
These measures answer different questions:
- Policies are relevant when the expense is driven by contract-level work, such as issuing or renewing a contract.
- Insured units are relevant when the expense is driven by the number of covered objects, such as maintaining item-level records or inspecting each object.
- Earned exposures are relevant when the allocation is intended to correspond to coverage earned over time. Earned exposures may be fractional and are not necessarily equal to policies in force at a particular date.
These are useful indicators, not universal rules. For instance, an expense recorded as “policy administration” may in practice vary with the number of insured units or endorsements. The analyst should understand the expense's cause and the intended rate unit before selecting the denominator.
| Count measure | What is counted | Example interpretation | Common mismatch |
|---|---|---|---|
| Policy count | Contracts | One contract for a specified coverage period | Treating one multi-unit policy as one insured object |
| Insured-unit count | Covered objects or units | Three vehicles insured under one contract | Assuming each unit was covered for a full year |
| Earned exposures | Coverage earned, measured on the defined exposure basis | Six months of coverage may earn one-half of an annual exposure | Using a policy count as if it were a time-adjusted exposure total |
Choosing and Applying an Allocation Base
Let \(F\) denote the fixed expenses to be allocated, and let \(N\) denote the count on the selected base. If the allocation is uniform across items counted by that base, the amount assigned per counted item is \(F/N\), provided \(N>0\). The implied allocation to a group containing \(N_g\) items is \(N_g(F/N)\). The result is an allocation, not evidence that every item actually caused exactly the same expense.
The most important control is consistency. The expense total, denominator, and intended allocation period should describe compatible business. If \(F\) covers a full year of expense, a full-year exposure denominator may be appropriate when the selected base is earned exposure. If \(F\) covers a shorter accounting period, the count and scope should be aligned with that period or adjusted under a clearly stated method. Do not silently combine an annual expense total with a count for a different time period.
State which fixed expenses are included, the period they cover, and the business segment to which they relate.
Decide whether the allocation is meant to follow contracts, insured objects, or earned coverage. Document the reason for the choice.
Check its definition, period, segment, and unit. Do not substitute policy count for earned exposure merely because it is easier to obtain.
Allocate the expense to the relevant groups, then confirm that the group allocations sum to the original expense pool.
Worked Allocations
Worked Example: Choosing Among Policies, Units, and Earned Exposures
A hypothetical insurer has $420,000 of annual fixed service expense for a small commercial equipment product. During the year it has 14,000 policies, 21,000 insured pieces of equipment, and 17,500 earned equipment exposures. Calculate the uniform allocation per item under each possible base. Which base is appropriate if the expense is primarily driven by servicing each insured piece of equipment? Use the method of dividing the expense pool by the matching count, and avoid the common mistake of treating the three counts as equivalent.
Step 1: Calculate the policy-based amount. If the expense were allocated uniformly per contract, the calculation would be:
Step 2: Calculate the insured-unit-based amount. If the expense is driven by servicing each piece of equipment, the matching base is the 21,000 insured units:
Step 3: Calculate the earned-exposure-based amount. If the intended allocation is per earned equipment exposure, use the 17,500 earned exposures:
Decision and check. Because the stated expense is primarily driven by servicing each insured piece of equipment, the unit-based allocation is the best match: $20.00 per insured unit. Its aggregate allocation is \(21{,}000 \times \$20.00=\$420{,}000\), which reconciles to the pool. The $30.00 and $24.00 figures are valid calculations under different allocation choices, but they do not describe the same cost per item. The common mistake this method avoids is selecting a denominator based only on convenience and then giving the resulting amount the wrong interpretation.
Worked Example: Allocating a Unit-Driven Expense Across Two Groups
A hypothetical insurer has $360,000 of fixed inspection expense for two classes of agricultural equipment. Class A has 8,000 policies, 12,000 insured units, and 6,000 earned exposures. Class B has 4,000 policies, 10,000 insured units, and 4,000 earned exposures. The inspections are performed once per insured unit, regardless of how long coverage was in force. Allocate the expense to the classes using insured units. Then compare the class allocations that would result from policy-count and earned-exposure bases. Use a uniform allocation within each chosen base; the key trap is to confuse the inspection driver with the exposure measure.
Step 1: Total each candidate denominator.
Step 2: Allocate by insured unit. The unit cost is \(\$360{,}000/22{,}000=\$16.363636\ldots\). Applying the unrounded amount to each class gives:
Rounded to cents, the class allocations sum to \(\$196{,}363.64+\$163{,}636.36=\$360{,}000.00\). Since inspections are performed per unit, these are the allocations that follow the stated cost driver.
Step 3: Compare alternative bases. A policy-count allocation would be \(\$360{,}000/12{,}000=\$30.00\) per policy. It gives Class A \(8{,}000 \times \$30.00=\$240{,}000\) and Class B \(4{,}000 \times \$30.00=\$120{,}000\). An earned-exposure allocation would be \(\$360{,}000/10{,}000=\$36.00\) per earned exposure. It gives Class A \(6{,}000 \times \$36.00=\$216{,}000\) and Class B \(4{,}000 \times \$36.00=\$144{,}000\). Each method reconciles to $360,000, but the allocations differ because the class shares of policies, units, and earned exposures differ. The mistake to avoid is assuming that reconciliation alone proves the denominator is appropriate.
Why Allocations Reconcile
For a uniform allocation on one base, the expense assigned to group \(g\) is its share of the total count multiplied by the total expense. This gives a simple conservation check: if all groups are included and their counts sum to the total denominator, their assigned expenses must sum to the original pool.
The derivation depends on \(N=\sum_g N_g\), on all groups being included, and on the same uniform rate \(F/N\) being applied to every counted item. If groups have different cost drivers or intentionally different allocation rules, this particular uniform allocation calculation does not apply without modification. Reconciliation is necessary, but it does not decide whether the chosen base reflects the expense being allocated.
Worked Example: Separating Contract and Unit Expenses
A hypothetical insurer has $270,000 of fixed expense for a marine policy portfolio: $180,000 is contract issuance expense, and $90,000 is equipment-record maintenance expense. The portfolio contains 9,000 policies, 15,000 insured watercraft, and 12,000 earned watercraft exposures. Allocate each expense component to its matching base. Then calculate the portfolio's combined expense per policy and its blended expense per earned exposure. The common trap is to divide the whole expense pool by one denominator without recognizing that the two components have different drivers.
Step 1: Allocate issuance expense by policy. Contract issuance is driven by policies, so:
The total allocated issuance expense is \(9{,}000 \times \$20.00=\$180{,}000\).
Step 2: Allocate record-maintenance expense by insured unit. The maintenance work is driven by the number of watercraft:
The total allocated maintenance expense is \(15{,}000 \times \$6.00=\$90{,}000\). Together the components allocate \(\$180{,}000+\$90{,}000=\$270{,}000\), matching the expense pool.
Step 3: Calculate the portfolio averages requested. The combined portfolio expense per policy is \(\$270{,}000/9{,}000=\$30.00\). The blended expense per earned watercraft exposure is \(\$270{,}000/12{,}000=\$22.50\). These are portfolio averages; neither replaces the component allocations. In particular, $22.50 is not evidence that the entire expense was caused by earned exposure. It is a blended equivalent obtained by dividing the total by that exposure count.
Check. The two component allocations total $270,000, while the two portfolio averages each use the full $270,000 numerator and their own stated denominator. This method keeps the allocation drivers separate and avoids treating an average across the portfolio as a causal per-unit expense.
Common Mistakes and Practical Checks
- Equating a policy with an insured unit. A contract may cover several objects. Count the units that the allocation base actually calls for.
- Equating a count with earned exposure. Exposure reflects earned coverage under the defined exposure basis; a count of policies or objects at a date may not reflect time in force.
- Changing the denominator without changing the interpretation. A result in dollars per policy is not dollars per unit or dollars per earned exposure, even if all three calculations use the same expense total.
- Combining distinct expense drivers too early. When one component is driven by policies and another by insured units, allocate the components separately before calculating any portfolio average.
- Using a reconciliation check as the only test. A calculation can sum to the correct total and still distribute expense to the wrong groups. Check both the arithmetic and the rationale for the base.
These distinctions support consistent expense provisions in ratemaking: the analyst can connect an expense pool to the correct unit of business and explain how the allocation changes across segments. They also make data questions easier to detect. If a portfolio reports more earned exposures than insured units, or an unexpected relationship between policy and unit counts, investigate definitions, timing, and scope before proceeding. The appropriate denominator depends on the expense purpose and rating basis; it is not determined by the labels alone. This treatment is aligned with the CAS Exam 5 (Basic Ratemaking and Estimated Claim Liabilities) syllabus.
Check Your Understanding
For each problem, identify the relevant denominator before calculating. Amounts are hypothetical.
- A specialty insurer has $96,000 of expense intended to be allocated per insured laboratory instrument. Its book has 3,200 policies, 4,800 instruments, and 4,000 earned instrument exposures. What is the allocation per instrument? (Answer: $20.00 per insured instrument)
- A fixed service pool of $210,000 is allocated by insured unit. Class A has 5,000 units and Class B has 7,000 units. What amount is allocated to each class, and what is the uniform amount per unit? (Answer: $87,500 to A; $122,500 to B; $17.50 per unit)
- A portfolio has $120,000 of policy-driven expense and $80,000 of unit-driven expense, 4,000 policies, 10,000 insured units, and 5,000 earned exposures. Find the expense per policy for the policy-driven component, the expense per unit for the unit-driven component, and the blended total per earned exposure. (Answer: $30.00 per policy; $8.00 per unit; $40.00 per earned exposure)
- A $150,000 pool is allocated by earned exposure. Segment A has 5,000 earned exposures and 4,000 policies; Segment B has 2,500 earned exposures and 4,000 policies. Find each segment's allocation and the allocation per policy within each segment. (Answer: A receives $100,000, or $25.00 per policy; B receives $50,000, or $12.50 per policy)
- A fixed expense pool of $240,000 is split into $144,000 of policy issuance expense and $96,000 of unit maintenance expense. There are 6,000 policies, 12,000 insured units, and 9,000 earned exposures. Calculate each component's amount per matching base, then calculate the blended total per earned exposure. (Answer: $24.00 per policy; $8.00 per unit; $26.67 per earned exposure)