Pure Premium as Loss Cost per Exposure

Introduction

A ratemaking analysis needs a loss measure that can be compared on a common basis across the risks being rated. Pure premium expresses projected covered losses for that risk group as a dollar amount per exposure unit. It is a loss-cost estimate, not automatically the amount an insurer should charge.

The distinction is important: a quoted rate may need to fund expenses and other provisions in addition to losses. Pure premium isolates the projected covered-loss component, making it a useful basis for a rate indication. This tutorial focuses on its meaning, calculation, and interpretation; the particular choice of exposure base and the full conversion from loss cost to an indicated rate are separate steps.

Definition: For a defined coverage, population, and projection basis, pure premium (also called loss cost per exposure) is projected covered losses divided by the corresponding exposure units. If $L_{\text{proj}}$ is the projected covered loss amount and $E$ is the number of exposure units on the same basis, then

\text{Pure premium}=\frac{L_{\text{proj}}}{E}.

The result is measured in dollars per exposure unit. For example, if the exposure unit is one vehicle-year, a pure premium of $300 means projected covered losses of $300 per vehicle-year for the specified group and period. It does not mean every vehicle will have a $300 loss, nor does it mean the quoted premium for each vehicle must be $300.

What Belongs in the Calculation?

The numerator is a projected amount of covered losses, not necessarily the total originally observed in historical data. It may reflect the selected loss basis and adjustments appropriate to the projection. The denominator is the exposure volume corresponding to that numerator. Both must refer to a compatible population, coverage, and basis.

The word covered matters. The loss amount must reflect the coverage being analyzed and the treatment of policy limits, deductibles, and other coverage terms. A loss amount that includes items outside the defined coverage does not provide the corresponding pure premium. Likewise, expenses should not be silently added to a calculation described as covered losses. If a particular analysis defines its loss measure to include a specified additional component, that definition must be stated and applied consistently.

Exposure units measure the amount of risk represented in the analysis. Their form depends on the selected exposure base. Here, treat the exposure base as already specified: the essential calculation is to divide projected losses by exposure units measured on that same basis. A figure described only as “$300 per unit” is incomplete unless the unit is clear.

Worked Example: Projecting Losses Before Dividing by Exposure

A hypothetical group has 1,200 exposure units. Selected historical covered losses are $486,000. For the stated projection, apply a loss development factor of 1.08 and then a trend factor of 1.05. Calculate the pure premium. Assume both factors apply multiplicatively to the selected historical losses, and that the exposure count is already on the required basis.

Step 1: Develop the selected losses.

\$486{,}000\times 1.08=\$524{,}880.

Step 2: Apply the trend factor.

\$524{,}880\times 1.05=\$551{,}124.

Thus the projected covered losses are $551,124. Step 3: Divide by exposure units.

\frac{\$551{,}124}{1{,}200}=\$459.27 \quad\text{per exposure unit}.

The pure premium is $459.27 per exposure unit. The check is $1{,}200\times \$459.27=\$551{,}124$, which recovers the projected loss amount exactly. The method is to project the numerator first and then divide by its corresponding exposure volume. A common mistake is to divide the historical losses by exposure and then forget to apply the projection factors; that would give $405.00, which is not the requested projected pure premium.

Frequency and Severity Interpretation

Pure premium can also be understood through expected claim count and average covered loss per claim. This interpretation is useful when a claim-count and severity analysis provides the inputs. It does not change the definition: it expresses the same expected covered-loss amount per exposure unit.

Model and assumptions: Let $N$ be the number of covered claims in a fixed group of $E$ exposure units, and let $X_j$ be the covered loss amount for claim $j$. Suppose claim severities have the same finite mean $\mu=E[X_j]$ and are independent of the claim count $N$. Then the expected aggregate covered loss is $E[N]\mu$, and the expected pure premium is $E[N]\mu/E$. This representation does not apply as written if the severity distribution depends on claim count or if the claims do not share the stated mean; in those cases, use the appropriate expected aggregate loss directly.

Define expected claim frequency per exposure unit as $f=E[N]/E$. Under the stated assumptions, pure premium is frequency multiplied by mean covered severity. Frequency is measured in claims per exposure unit and severity in dollars per claim, so their product is dollars per exposure unit.

\frac{E[N]\mu}{E} = \left(\frac{E[N]}{E}\right)\mu = f\mu.

The justification follows from conditioning on the number of claims. Given $N=n$, the aggregate covered loss is $X_1+\cdots+X_n$. Its conditional expected value is $n\mu$, since each claim has mean $\mu$. Averaging over possible values of $N$ gives

E\left[\sum_{j=1}^{N}X_j\right] = E\big[E[\sum_{j=1}^{N}X_j\mid N]\big] = E[N\mu] = E[N]\mu.

Dividing by $E$ exposure units establishes the result. This derivation explains why the frequency-severity calculation and the projected aggregate-loss calculation agree when their assumptions and coverage basis match.

Worked Example: Calculating Pure Premium from Frequency and Severity

A hypothetical portfolio has 2,400 exposure units. A model projects 300 covered claims, with an average covered loss of $2,800 per claim. Assume claim severities have the stated mean and are independent of claim count. Find the pure premium using both the aggregate-loss and frequency-severity calculations.

Step 1: Find projected aggregate covered losses.

300\text{ claims}\times \$2{,}800\text{ per claim}=\$840{,}000.

Step 2: Divide by exposure units.

\frac{\$840{,}000}{2{,}400}=\$350 \quad\text{per exposure unit}.

As a check using frequency and severity, the projected frequency is

\frac{300}{2{,}400}=0.125 \quad\text{claims per exposure unit}, \qquad 0.125\times \$2{,}800=\$350 \quad\text{per exposure unit}.

Both methods give a pure premium of $350 per exposure unit. The calculation is consistent because $2{,}400\times \$350=\$840{,}000$. The common mistake is to divide the projected losses by the number of claims: $\$840{,}000/300=\$2{,}800$, which is average severity per claim, not loss cost per exposure. The correct denominator for pure premium is exposure units.

Pure Premium Is Not the Quoted Rate

A quoted rate is the amount charged per rating unit under the applicable rating structure. Pure premium describes projected covered losses per exposure unit. The two amounts may use the same denominator, but they represent different components. A rate generally must account for provisions beyond covered losses, such as expenses and other selected provisions. Therefore, pure premium alone does not determine a complete quoted rate.

The comparison is meaningful only when the amounts are expressed on compatible bases. If the rate is per exposure unit, compare it with loss cost per that same unit. If one amount is per policy and the other is per vehicle-year, the numerical comparison is not interpretable until the bases are aligned.

Measure	What it represents	Typical units
Pure premium (loss cost)	Projected covered losses divided by exposure units	Dollars per exposure unit
Quoted rate	Amount charged per rating unit under the applicable rate structure	Dollars per rating unit

Worked Example: Comparing Loss Cost with a Quoted Rate

A hypothetical book has 1,600 exposure units and projected covered losses of $496,000. Its quoted rate is a constant $420 per exposure unit. Calculate the pure premium, the premium at the quoted rate, the dollar difference per exposure unit, and the ratio of projected losses to premium at that rate. Assume the exposure basis is identical for all quantities.

Step 1: Calculate pure premium.

\frac{\$496{,}000}{1{,}600}=\$310 \quad\text{per exposure unit}.

Step 2: Calculate premium at the quoted rate.

1{,}600\times \$420=\$672{,}000.

Step 3: Compare the two amounts per exposure unit.

\$420-\$310=\$110 \quad\text{per exposure unit}.

Step 4: Calculate projected losses as a share of premium.

\frac{\$496{,}000}{\$672{,}000}=0.738095\ldots=73.81\%.

The pure premium is $310 per exposure unit; premium at the quoted rate is $672,000; and the quoted rate exceeds the loss cost by $110 per exposure unit. The projected loss-to-premium ratio is 73.81%. The remaining difference is not automatically profit: it must be interpreted in light of the other provisions and assumptions underlying the rate. The check is that $1{,}600\times \$310=\$496{,}000$, matching projected losses. The method avoids the common error of calling the quoted $420 rate the pure premium simply because both amounts are stated per exposure unit.

Using Pure Premium as a Loss-Cost Basis

Pure premium provides a compact way to express the projected loss burden of a defined risk group. In a rate indication, it supplies the loss-cost basis that can then be considered alongside the other components needed to determine a rate. That role is distinct from deciding the appropriate exposure base, combining experience periods, or calculating a final indicated rate.

A useful calculation sequence is:

Define the loss measure.
Specify the coverage and the covered-loss amount being projected. Keep any included components explicit.

Establish projected losses.
Use the selected projection basis to obtain a loss amount for the target analysis.

Confirm matching exposure units.
Ensure the exposure volume corresponds to the same population, coverage, and analysis basis as the losses.

Divide and label the result.
Report dollars per exposure unit, rather than presenting the result as an unlabeled dollar amount or as the quoted rate.

The ratio answers a specific question: What projected covered-loss amount corresponds to one unit of this exposure measure? It does not say that each exposure will incur that exact loss. Individual claims vary, and some exposure units may have no claim while others may produce substantial losses. Pure premium is an average loss-cost measure for the defined group and basis.

Common Errors and Interpretation Traps

Using claim count as the denominator. Dividing aggregate losses by claims gives average severity per claim. Pure premium divides by exposure units.
Leaving the exposure unit unstated. A result such as “$350” is ambiguous; report “$350 per exposure unit” and identify the unit where possible.
Mixing bases. The exposure denominator must correspond to the loss numerator in population, coverage, and period. Otherwise the ratio may not represent the intended loss cost.
Confusing loss cost with price. A pure premium is the covered-loss component per exposure unit, not automatically the complete amount charged.
Applying a projection adjustment inconsistently. If the numerator is meant to represent projected losses, do not report an unadjusted historical loss amount as though it were already projected.

Key takeaway: Pure premium is projected covered losses divided by matching exposure units. It is a loss-cost basis for ratemaking, not a quoted rate; frequency multiplied by mean covered severity gives the same result when the stated model assumptions hold.

Check Your Understanding

For each problem, identify the relevant loss-cost calculation and keep the units explicit.

A hypothetical group has 900 exposure units and projected covered losses of $252,000. What is its pure premium? (Answer: $280 per exposure unit.)
A selected covered-loss amount of $360,000 is to be developed by a factor of 1.10 and then trended by a factor of 1.10. The group has 1,500 matching exposure units. What is the projected pure premium? (Answer: $290.40 per exposure unit.)
A portfolio has 4,200 exposure units and 525 expected covered claims. Mean covered severity is $2,400 per claim. Under the frequency-severity assumptions in this tutorial, calculate the pure premium. (Answer: $300 per exposure unit.)
A group has 800 exposure units, projected covered losses of $328,000, and a quoted rate of $515 per exposure unit. Calculate the pure premium, total premium at the quoted rate, and projected losses as a percentage of that premium. (Answer: $410 per exposure unit; $412,000 total premium; 79.61%.)
A portfolio has 1,600 exposure units and a pure premium of $275 per exposure unit. A revised projection increases covered losses by 4%, with exposure unchanged. What is the revised projected loss amount and pure premium? (Answer: $457,600 and $286 per exposure unit.)

What You'll Learn