Converting Pure Premiums Between Exposure Units

Why the Exposure Unit Matters

A pure premium is not just a dollar amount. It is a dollar amount per a stated exposure unit. A value of $60 per vehicle-year and a value of $120 per policy-year can describe the same expected loss experience if each policy-year represents two vehicle-years. Comparing or applying those figures without their units can lead to a factor-of-two error.

Converting between units is an algebraic adjustment to the denominator, not a change to the losses. The conversion is valid only when the relationship between the units is specified and applies to the experience being analyzed. The loss definition, coverage, experience period, and other relevant bases must remain consistent.

Definition: Let $E_A$ be the number of exposure units of type $A$, and let $E_B$ be the number of exposure units of type $B$, for the same portfolio and period. If the stated relationship is $E_B=cE_A$, where $c$ is a positive conversion factor, then the pure premiums for the same total losses $L$ are $PP_A=L/E_A$ and $PP_B=L/E_B$. Therefore, $PP_B=PP_A/c$, and $PP_A=cPP_B$. The units must be included when reporting either result.

For example, if one policy-year corresponds to an average of 2.4 vehicle-years, then there are 2.4 vehicle-years per policy-year. The pure premium per vehicle-year is the pure premium per policy-year divided by 2.4. To convert in the opposite direction, multiply the per-vehicle-year amount by 2.4.

The conversion factor follows the exposure relationship, not an intuition about whether the new unit sounds larger or smaller. A unit that produces more exposure units for the same portfolio will have a smaller pure premium per unit, all else equal.

Deriving the Conversion

The conversion comes directly from calculating the same loss amount over two denominators. Suppose the recorded losses are $L$, the experience has $E_A$ units of type $A$, and the corresponding number of units of type $B$ is $E_B=cE_A$. Then:

PP_B=\frac{L}{E_B} =\frac{L}{cE_A} =\frac{1}{c}\frac{L}{E_A} =\frac{PP_A}{c}.

This also gives a practical check: after converting, multiplying the pure premium by its exposure count should recover the original total losses. If it does not, either the conversion factor has been reversed or the exposure relationship has been applied inconsistently.

When the relationship varies across segments, do not assume there is one common factor for every segment. Calculate the total number of each exposure unit for the portfolio, then use the ratio of those totals. This keeps the conversion tied to the actual mix rather than to an unweighted average of segment-level relationships.

Worked Example: Converting Fleet-Year Experience to Vehicle-Year

A hypothetical fleet program has 7,500 fleet-year exposures. The stated average is 2.4 covered vehicles per fleet-year, so the relevant exposure basis is vehicle-years. Covered losses for the same period and coverage are $1,080,000. Calculate the pure premium per fleet-year and per vehicle-year, then verify the conversion.

Method and unit relationship. There are 2.4 vehicle-years per fleet-year, so $E_{\text{vehicle}}=2.4E_{\text{fleet}}$. The vehicle-year exposure count is:

E_{\text{vehicle}}=7{,}500(2.4)=18{,}000\text{ vehicle-years}.

Pure premium per fleet-year. Divide the covered losses by fleet-year exposures:

PP_{\text{fleet}}=\frac{\$1{,}080{,}000}{7{,}500} =\$144\text{ per fleet-year}.

Convert to vehicle-year. Since each fleet-year contains 2.4 vehicle-years, divide the per-fleet-year pure premium by 2.4:

PP_{\text{vehicle}}=\frac{\$144}{2.4} =\$60\text{ per vehicle-year}.

Check. The converted value gives the original loss total: $18{,}000(\$60)=\$1{,}080{,}000$. The reverse conversion also works: $2.4(\$60)=\$144$ per fleet-year.

Common mistake avoided. Multiplying $144 by 2.4 would produce $345.60 per vehicle-year, which uses the factor in the wrong direction. That amount would imply $18{,}000(\$345.60)=\$6{,}220{,}800$, not the observed $1,080,000.

Conversion When the Unit Is a Payroll or Value Increment

Exposure units often represent increments of a continuous base, such as $1,000 of payroll, $100 of insured value, or a specified amount of sales. To convert between two increments, count how many of the new units represent the same total base. A smaller increment produces more exposure units and therefore a smaller pure premium per unit.

The exposure basis and the dollar amount used to define a unit must both be stated. “$12 per unit” is incomplete if the reader cannot tell whether a unit is $100 or $1,000 of payroll.

Worked Example: Converting a Payroll Pure Premium

A hypothetical service portfolio has $24,000,000 of payroll and $3,600,000 of covered losses. Calculate the pure premium per $1,000 of payroll and per $100 of payroll. Then check that both amounts reproduce the same losses.

Method and exposure counts. The number of $1,000 payroll units is $24{,}000{,}000/1{,}000=24{,}000$. The number of $100 payroll units is $24{,}000{,}000/100=240{,}000$. Thus, there are 10 units of $100 payroll for every unit of $1,000 payroll.

Pure premium per $1,000.

PP_{\$1{,}000}=\frac{\$3{,}600{,}000}{24{,}000} =\$150\text{ per \$1,000 of payroll}.

Convert to $100 units. Since the same payroll is represented by 10 times as many $100 units, divide by 10:

PP_{\$100}=\frac{\$150}{10} =\$15\text{ per \$100 of payroll}.

Check. On the $1,000 basis, $24{,}000(\$150)=\$3{,}600{,}000$. On the $100 basis, $240{,}000(\$15)=\$3{,}600{,}000$. The two pure premiums are equivalent because they use different unit sizes.

Common mistake avoided. Multiplying $150 by 10 when converting to $100 units would give $1,500 per $100. That would imply total losses of $240{,}000(\$1{,}500)=\$360{,}000{,}000$, ten times the given losses. More units in the denominator require a lower amount per unit.

Checking That a Rate Uses the Intended Unit

A calculated pure premium can be carried into a rate indication only if its unit matches the unit used when applying or communicating the indicated rate. The arithmetic can be correct while the result is mislabeled. A strong check is to apply the rate to the exposure count that its label calls for and confirm that the resulting expected loss is consistent with the underlying amount.

The same principle applies when the target unit is defined by insured value. If a rate is stated per $100 of value, divide total insured value by $100 to obtain the number of units before applying the rate. If it is stated per $1,000, divide by $1,000 instead. Do not apply a per-$100 rate to a count of $1,000 units.

Worked Example: Verifying a Rate per Insured-Value Unit

A hypothetical equipment portfolio has total insured value of $52,000,000 and covered losses of $624,000. An analyst needs to express the pure premium both per $1,000 of insured value and per $100 of insured value. The analyst will then check the amount for a separate account with $250,000 of insured value.

Method and exposure counts. The original portfolio contains $52{,}000{,}000/1{,}000=52{,}000$ units of $1,000 and $52{,}000{,}000/100=520{,}000$ units of $100. There are 10 times as many $100 units.

Calculate the pure premium per $1,000.

PP_{\$1{,}000}=\frac{\$624{,}000}{52{,}000} =\$12\text{ per \$1,000 of insured value}.

Convert to a $100 basis. Divide by 10 because there are 10 times as many units:

PP_{\$100}=\frac{\$12}{10} =\$1.20\text{ per \$100 of insured value}.

Check the separate account. Its $250,000 value corresponds to $250{,}000/100=$ 2,500 units of $100, so the expected amount on that basis is $2{,}500(\$1.20)=\$3{,}000$. On a $1,000 basis, it has 250 units, giving $250(\$12)=\$3{,}000$. Both units produce the same result.

Common mistake avoided. Applying $1.20 to the 250 units of $1,000 would give $300. The $1.20 amount is per $100, not per $1,000. The exposure count and the quoted unit must match.

Different Relationships Across Segments

A single conversion factor is not appropriate if the number of target units per source unit differs among segments. In that case, total the target exposure units across the segments and calculate the pure premium using that total. Equivalently, for the full portfolio, the conversion factor is the total target exposures divided by total source exposures.

c_{\text{portfolio}}=\frac{\sum_j E_{B,j}}{\sum_j E_{A,j}}, \qquad PP_B=\frac{\sum_j L_j}{\sum_j E_{B,j}}.

This method avoids treating all segments as if they had the same relationship. It also avoids averaging segment-level conversion factors without weighting them by their source exposure volumes.

Identify the two units.
Write each unit in full, including the time period or base increment where relevant.

State the relationship.
Determine how many target units correspond to one source unit, or calculate the ratio of total target exposures to total source exposures.

Convert the pure premium.
If the target exposure count is $c$ times the source count, divide the source-basis pure premium by $c$. Multiply by $c$ to convert back.

Verify the basis.
Multiply the converted pure premium by its exposure count. Confirm that it reproduces the same loss amount and that the rate label matches the exposure count used.

Common Exam Traps and Practical Checks

A unit conversion does not repair a mismatch in what the losses represent. Before applying the conversion, confirm that the numerator and both exposure counts refer to the same coverage, experience period, and intended loss basis. If one calculation includes a different loss component or uses exposures from a different period, equality of the unit labels alone does not make the comparison valid.

Reversing the factor. If the target basis has more units, the pure premium per target unit is smaller. Use $PP_B=PP_A/c$ when $E_B=cE_A$.
Dropping the unit label. Report “$15 per $100 of payroll,” not simply “$15.” The amount cannot be interpreted without its denominator.
Using the wrong exposure count. A rate per $100 of value requires a count of $100 units, not a count of $1,000 units.
Assuming a fixed relationship that does not hold. If segment-level relationships differ, use total exposures for the portfolio rather than an unweighted average factor.
Changing the loss basis during conversion. Conversion changes the denominator only; it does not change the losses being divided.

Key takeaway: First establish the exact relationship between the exposure units. Then adjust the pure premium inversely to the exposure count, keep the loss basis unchanged, and verify the result by reconstructing the same total loss amount.

Check Your Understanding

For each problem, identify the exposure relationship, convert the pure premium, and check that the reported unit is the intended one.

A hypothetical delivery program has 7,800 fleet-year exposures, averages 3 covered vehicles per fleet-year, and has $936,000 of covered losses. Find the pure premium per vehicle-year and per fleet-year. (Answer: $40 per vehicle-year and $120 per fleet-year.)
A hypothetical contractor portfolio has $1,500,000 of payroll and $270,000 of covered losses. Calculate the pure premium per $1,000 of payroll and per $100 of payroll. (Answer: $180 per $1,000 and $18 per $100 of payroll.)
A hypothetical property portfolio has $48,000,000 of insured value and $960,000 of covered losses. Convert the pure premium per $1,000 of value to a pure premium per $500 of value. (Answer: $20 per $1,000 and $10 per $500 of insured value.)
A hypothetical equipment insurer has $156,000 of covered losses and $2,400,000 of sales exposure. First calculate the pure premium per $10,000 of sales, then convert it to a per-$1,000 basis. (Answer: $650 per $10,000 and $65 per $1,000 of sales.)
A hypothetical fleet book has two segments. Segment A has 4,000 policy-years, averages 1.5 vehicles per policy-year, and has $360,000 of losses. Segment B has 2,000 policy-years, averages 3 vehicles per policy-year, and has $540,000 of losses. Calculate the aggregate pure premium per vehicle-year and per policy-year. (Answer: $75 per vehicle-year and $150 per policy-year.)

What You'll Learn